liu.seSök publikationer i DiVA
Ändra sökning
Avgränsa sökresultatet
1234 101 - 150 av 159
RefereraExporteraLänk till träfflistan
Permanent länk
Referera
Referensformat
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • oxford
  • Annat format
Fler format
Språk
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Annat språk
Fler språk
Utmatningsformat
  • html
  • text
  • asciidoc
  • rtf
Träffar per sida
  • 5
  • 10
  • 20
  • 50
  • 100
  • 250
Sortering
  • Standard (Relevans)
  • Författare A-Ö
  • Författare Ö-A
  • Titel A-Ö
  • Titel Ö-A
  • Publikationstyp A-Ö
  • Publikationstyp Ö-A
  • Äldst först
  • Nyast först
  • Skapad (Äldst först)
  • Skapad (Nyast först)
  • Senast uppdaterad (Äldst först)
  • Senast uppdaterad (Nyast först)
  • Disputationsdatum (tidigaste först)
  • Disputationsdatum (senaste först)
  • Standard (Relevans)
  • Författare A-Ö
  • Författare Ö-A
  • Titel A-Ö
  • Titel Ö-A
  • Publikationstyp A-Ö
  • Publikationstyp Ö-A
  • Äldst först
  • Nyast först
  • Skapad (Äldst först)
  • Skapad (Nyast först)
  • Senast uppdaterad (Äldst först)
  • Senast uppdaterad (Nyast först)
  • Disputationsdatum (tidigaste först)
  • Disputationsdatum (senaste först)
Markera
Maxantalet träffar du kan exportera från sökgränssnittet är 250. Vid större uttag använd dig av utsökningar.
  • 101.
    Lundquist, Christian
    et al.
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Schön, Thomas
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Gustafsson, Fredrik
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Situational Awareness and Road Prediction for Trajectory Control Applications2012Ingår i: Handbook of Intelligent Vehicles / [ed] Azim Eskandarian, Springer London, 2012, s. 365-396Kapitel i bok, del av antologi (Refereegranskat)
    Abstract [en]

    The Handbook of Intelligent Vehicles provides a complete coverage of the fundamentals, new technologies, and sub-areas essential to the development of intelligent vehicles; it also includes advances made to date, challenges, and future trends. Significant strides in the field have been made to date; however, so far there has been no single book or volume which captures these advances in a comprehensive format, addressing all essential components and subspecialties of intelligent vehicles, as this book does. Since the intended users are engineering practitioners, as well as researchers and graduate students, the book chapters do not only cover fundamentals, methods, and algorithms but also include how software/hardware are implemented, and demonstrate the advances along with their present challenges. Research at both component and systems levels are required to advance the functionality of intelligent vehicles. This volume covers both of these aspects in addition to the fundamentals listed above.

    Ladda ner fulltext (pdf)
    fulltext
  • 102.
    Nilsson, Emil
    et al.
    Autoliv Electronics AB, Sweden.
    Lundquist, Christian
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Schön, Thomas
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Forslund, David
    Autoliv Electronics AB, Sweden.
    Roll, Jacob
    Autoliv Electronics AB, Sweden.
    Vehicle Motion Estimation Using an Infrared Camera2011Ingår i: Proceedings of the 18th IFAC World Congress / [ed] Bittanti, Sergio; Cenedese, Angelo; Zampieri, Sandro, Elsevier, 2011, s. 12952-12957Konferensbidrag (Refereegranskat)
    Abstract [en]

    This paper is concerned with vehicle motion estimation. The problem is formulated as a sensor fusion problem, where the vehicle motion is estimated based on the information from a far infrared camera, inertial sensors and the vehicle speed. This information is already present in premium cars. We are concerned with the off-line situation and the approach taken is to formulate the problem as a nonlinear least squares problem. In order to illustrate the performance of the proposed method experiments on rural roads in Sweden during night time driving have been performed. The results clearly indicates the efficacy of the approach.

    Ladda ner fulltext (pdf)
    fulltext
  • 103.
    Ninness, Brett
    et al.
    University of Newcastle, Australia.
    Wills, Adrian
    University of Newcastle, Australia.
    Schön, Thomas
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Estimation of General Nonlinear State-Space Systems2011Rapport (Övrigt vetenskapligt)
    Abstract [en]

    This paper presents a novel approach to the estimation of a general class of dynamic nonlinear system models. The main contribution is the use of a tool from mathematical statistics, known as Fishers’ identity, to establish how so-called “particle smoothing” methods may be employed to compute gradients of maximum-likelihood and associated prediction error cost criteria.

    Ladda ner fulltext (pdf)
    FULLTEXT01
    Ladda ner fulltext (pdf)
    FULLTEXT03
  • 104.
    Ninness, Brett
    et al.
    University of Newcastle, Australia.
    Wills, Adrian
    University of Newcastle, Australia.
    Schön, Thomas
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Estimation of General Nonlinear State-Space Systems2010Ingår i: Proceedings of the 49th IEEE Conference on Decision and Control, 2010, s. 6371-6376Konferensbidrag (Refereegranskat)
    Abstract [en]

    This paper presents a novel approach to the estimation of a general class of dynamic nonlinear system models. The main contribution is the use of a tool from mathematical statistics, known as Fishers’ identity, to establish how so-called “particle smoothing” methods may be employed to compute gradients of maximum-likelihood and associated prediction error cost criteria.

  • 105.
    Norrlöf, Mikael
    et al.
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Henriksson, Robert
    McKinsey & Company, Sweden.
    Moberg, Stig
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Wernholt, Erik
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Schön, Thomas
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Experimental Comparison of Observers for Tool Position Estimation of Industrial Robots2009Ingår i: Proceedings of the 48th IEEE Conference on Decision and Control, 2009, s. 8065-8070Konferensbidrag (Refereegranskat)
    Abstract [en]

    This paper investigates methods for tool position estimation of industrial robots. It is assumed that the motor angular position and the tool acceleration are measured. The considered observers are different versions of the extended Kalman filter as well as a deterministic observer. A method for tuning the observers is suggested and the robustness of the methods is investigated. The observers are evaluated experimentally on a commercial industrial robot.

  • 106.
    Orguner, Umut
    et al.
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Schön, Thomas
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Gustafsson, Fredrik
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Improved Target Tracking with Road Network Information2009Rapport (Övrigt vetenskapligt)
    Abstract [en]

    In this paper we consider the problem of tracking targets, which can move both on-road and off-road, with particle filters utilizing the road-network information. It is argued that the constraints like speed-limits and/or one-way roads generally incorporated into on-road motion models make it necessary to consider additional high-bandwidth off-road motion models. This is true even if the targets under consideration are only allowed to move on-road due to the possibility of imperfect road-map information and drivers violating the traffic rules. The particle filters currently used struggles during sharp mode transitions, with poor estimation quality as a result. This is due to the fact the number of particles allocated to each motion mode is varying according to the mode probabilities. A recently proposed interacting multiple model (IMM) particle filtering algorithm, which keeps the number of particles in each mode constant irrespective of the mode probabilities, is applied to this problem and its performance is compared to a previously existing algorithm. The results of the simulations on a challenging bearing-only tracking scenario show that the proposed algorithm, unlike the previously existing algorithm, can achieve good performance even under the sharpest mode transitions.

    Ladda ner fulltext (pdf)
    FULLTEXT01
  • 107.
    Orguner, Umut
    et al.
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Schön, Thomas
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Gustafsson, Fredrik
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Improved Target Tracking with Road Network Information2009Ingår i: Proceedings of the '09 IEEE Aerospace Conference, 2009, s. 1-11Konferensbidrag (Refereegranskat)
    Abstract [en]

    In this paper we consider the problem of tracking targets, which can move both on-road and off-road, with particle filters utilizing the road-network information. It is argued that the constraints like speed-limits and/or one-way roads generally incorporated into on-road motion models make it necessary to consider additional high-bandwidth off-road motion models. This is true even if the targets under consideration are only allowed to move on-road due to the possibility of imperfect road-map information and drivers violating the traffic rules. The particle filters currently used struggles during sharp mode transitions, with poor estimation quality as a result. This is due to the fact the number of particles allocated to each motion mode is varying according to the mode probabilities. A recently proposed interacting multiple model (IMM) particle filtering algorithm, which keeps the number of particles in each mode constant irrespective of the mode probabilities, is applied to this problem and its performance is compared to a previously existing algorithm. The results of the simulations on a challenging bearing-only tracking scenario show that the proposed algorithm, unlike the previously existing algorithm, can achieve good performance even under the sharpest mode transitions.

  • 108.
    Schön, Thomas
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    An Explanation of the Expectation Maximization Algorithm2009Rapport (Övrigt vetenskapligt)
    Abstract [en]

    The expectation maximization (EM) algorithm computes maximum like-lihood estimates of unknown parameters in probabilistic models involvinglatent variables. More pragmatically speaking, the EM algorithm is an iter-ative method that alternates between computing a conditional expectationand solving a maximization problem, hence the name expectation maxi-mization. We will in this work derive the EM algorithm and show that itprovides a maximum likelihood estimate. The aim of the work is to showhow the EM algorithm can be used in the context of dynamic systems andwe will provide a worked example showing how the EM algorithm can beused to solve a simple system identification problem.

    Ladda ner fulltext (pdf)
    FULLTEXT01
  • 109.
    Schön, Thomas
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Fusion of Data from Different Sources2009Rapport (Övrigt vetenskapligt)
    Abstract [en]

    The use of data from different, often complementary sources in order to obtain a better estimate of the state of the system under consideration has recently become very popular within many scientific areas. We will in this talk provide a framework, including the popular Kalman and particle filters for fusing data from different, complementary sources. The theory will be illustrated using several application examples from the automotive and the aerospace industry. Possible applications for 3D analysis of human motion will be discussed.

    Ladda ner fulltext (pdf)
    FULLTEXT01
  • 110.
    Schön, Thomas
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Fusion of Data from Different Sources2008Ingår i: Proceedings of the 10th International Symposium on 3D Analysis of Human Movement, 2008Konferensbidrag (Övrigt vetenskapligt)
    Abstract [en]

    The use of data from different, often complementary sources in order to obtain a better estimate of the state of the system under consideration has recently become very popular within many scientific areas. We will in this talk provide a framework, including the popular Kalman and particle filters for fusing data from different, complementary sources. The theory will be illustrated using several application examples from the automotive and the aerospace industry. Possible applications for 3D analysis of human motion will be discussed.

  • 111.
    Schön, Thomas
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    On computational methods for nonlinear estimation2003Licentiatavhandling, sammanläggning (Övrigt vetenskapligt)
    Abstract [en]

    The Bayesian approach provides a rather powerful framework for handling nonlinear, as well as linear, estimation problems. We can in fact pose a general solution to the nonlinear estimation problem. However, in the general case there does not exist any closed-form solution and we are forced to use approximate techniques. In this thesis we will study one such technique, the sequential Monte Carlo method, commonly referred to as the particle filter. Some work on linear stochastic differential-algebraic equations and constrained estimation using convex optimization will also be presented.

    The sequential Monte Carlo method offers a systematic framework for handling estimation of nonlinear systems subject to non-Gaussian noise. Its main drawback is that it requires a lot of computational power. We will use the particle filter both for the nonlinear state estimation problem and the nonlinear system identification problem. The details for the marginalized (Rao-Blackwellized) particle filter applied to a general nonlinear state-space model will also be given.

    General approaches to modeling, for instance using object-oriented software, lead to differential-algebraic equations. One of the topics in this thesis is to extend the standard Kalman filtering theory to the class of linear differential-algebraic equations, by showing how to incorporate white noise in this type of equations.

    There will also be a discussion on how to use convex optimization for solving the estimation problem. For linear state-space models with Gaussian noise the Kalman filter computes the maximum a posteriori estimate. We interpret the Kalman filter as the solution to a convex optimization problem, and show that we can generalize the maximum a posteriori state estimator to any noise with log-concave probability density function and any combination of linear equality and convex inequality constraints.

    Delarbeten
    1. A Modeling and Filtering Framework for Linear Differential-Algebraic Equations
    Öppna denna publikation i ny flik eller fönster >>A Modeling and Filtering Framework for Linear Differential-Algebraic Equations
    2003 (Engelska)Ingår i: Proceedings of the 42th IEEE Conference on Decision and Control, 2003, s. 892-897 vol.1Konferensbidrag, Publicerat paper (Refereegranskat)
    Abstract [en]

    General approaches to modeling, for instance using object-oriented software, lead to differential-algebraic equations (DAE). As the name reveals, it is a combination of differential and algebraic equations. For state estimation using observed system inputs and outputs in a stochastic framework similar to Kalman filtering, we need to augment the DAE with stochastic disturbances ("process noise"), whose covariance matrix becomes the tuning parameter. We will determine the subspace of possible causal disturbances based on the linear DAE model. This subspace determines all degrees of freedom in the filter design, and a Kalman filter algorithm is given. We illustrate the design on a system with two interconnected rotating masses.

    Nyckelord
    Implicit systems, Descriptor systems, Singular systems, White noise, Noise, Discretization, Kalman filters
    Nationell ämneskategori
    Teknik och teknologier Reglerteknik
    Identifikatorer
    urn:nbn:se:liu:diva-13917 (URN)10.1109/CDC.2003.1272679 (DOI)000189434100154 ()0-7803-7924-1 (ISBN)
    Konferens
    42nd IEEE Conference on Decision and Control, Maui, HI, USA, December, 2003
    Tillgänglig från: 2006-09-04 Skapad: 2006-09-04 Senast uppdaterad: 2013-11-27
    2. A Note on State Estimation as a Convex Optimization Problem
    Öppna denna publikation i ny flik eller fönster >>A Note on State Estimation as a Convex Optimization Problem
    2003 (Engelska)Ingår i: Proceedings of the 2003 IEEE International Conference on Acoustics, Speech, and Signal Processing, 2003, Vol. 6, nr 6-10, s. 61-64 vol.6Konferensbidrag, Publicerat paper (Refereegranskat)
    Abstract [en]

    The Kalman filter computes the maximum a posteriori (MAP) estimate of the states for linear state space models with Gaussian noise. We interpret the Kalman filter as the solution to a convex optimization problem, and show that we can generalize the MAP state estimator to any noise with a log-concave density function and any combination of linear equality and convex inequality constraints on the states. We illustrate the principle on a hidden Markov model, where the state vector contains probabilities that are positive and sum to one.

    Nyckelord
    State estimation, Kalman filter, Convex optimization, Hidden Markov Models
    Nationell ämneskategori
    Teknik och teknologier Reglerteknik
    Identifikatorer
    urn:nbn:se:liu:diva-13918 (URN)10.1109/ICASSP.2003.1201618 (DOI)0-7803-7663-3 (ISBN)
    Konferens
    2003 IEEE International Conference on Acoustics, Speech, and Signal Processing, Hong Kong, China, April, 2003
    Tillgänglig från: 2006-09-04 Skapad: 2006-09-04 Senast uppdaterad: 2013-11-27
    3. Marginalized Particle Filters for Mixed Linear/Nonlinear State-Space Models
    Öppna denna publikation i ny flik eller fönster >>Marginalized Particle Filters for Mixed Linear/Nonlinear State-Space Models
    2005 (Engelska)Ingår i: IEEE Transactions on Signal Processing, ISSN 1053-587X, E-ISSN 1941-0476, Vol. 53, nr 7, s. 2279-2289Artikel i tidskrift (Refereegranskat) Published
    Abstract [en]

    The particle filter offers a general numerical tool to approximate the posterior density function for the state in nonlinear and non-Gaussian filtering problems. While the particle filter is fairly easy to implement and tune, its main drawback is that it is quite computer intensive, with the computational complexity increasing quickly with the state dimension. One remedy to this problem is to marginalize out the states appearing linearly in the dynamics. The result is that one Kalman filter is associated with each particle. The main contribution in this paper is the derivation of the details for the marginalized particle filter for a general nonlinear state-space model. Several important special cases occurring in typical signal processing applications will also be discussed. The marginalized particle filter is applied to an integrated navigation system for aircraft. It is demonstrated that the complete high-dimensional system can be based on a particle filter using marginalization for all but three states. Excellent performance on real flight data is reported.

    Ort, förlag, år, upplaga, sidor
    IEEE Signal Processing Society, 2005
    Nyckelord
    Kalman filter, Marginalization, Navigation systems, Nonlinear systems, Particle filter, State estimation
    Nationell ämneskategori
    Reglerteknik
    Identifikatorer
    urn:nbn:se:liu:diva-11749 (URN)10.1109/TSP.2005.849151 (DOI)
    Tillgänglig från: 2008-05-07 Skapad: 2008-05-07 Senast uppdaterad: 2017-12-13Bibliografiskt granskad
    4. Particle Filters for System Identification of State-Space Models Linear in Either Parameters or States
    Öppna denna publikation i ny flik eller fönster >>Particle Filters for System Identification of State-Space Models Linear in Either Parameters or States
    2003 (Engelska)Ingår i: Proceedings of the 13th IFAC Symposium on System Identification, 2003, s. 1251-1256 vol.1Konferensbidrag, Publicerat paper (Refereegranskat)
    Abstract [en]

    The potential use of the marginalized particle filter for nonlinear system identification is investigated. The particle filter itself offers a general tool for estimating unknown parameters in non-linear models of moderate complexity, and the basic trick is to model the parameters as a random walk (so called roughening noise) with decaying variance. We derive algorithms for systems which are non-linear in either the parameters or the states, but not both generally. In these cases, marginalization applies to the linear part, which firstly significantly widens the scope of the particle filter to more complex systems, and secondly decreases the variance in the linear parameters/states for fixed filter complexity. This second property is illustrated on an example of chaotic model. The particular case of freely parametrized linear state space models, common in subspace identification approaches, is bi-linear in states and parameters, and thus both cases above are satisfied. One can then choose which one to marginalize.

    Nyckelord
    System identification, Nonlinear estimation, Recursive estimation, Particle filters, Kalman filters, Bayesian estimation
    Nationell ämneskategori
    Teknik och teknologier Reglerteknik
    Identifikatorer
    urn:nbn:se:liu:diva-13919 (URN)978-0080437095 (ISBN)
    Konferens
    13th IFAC Symposium on System Identification, Rotterdam, The Netherlands, August, 2003
    Tillgänglig från: 2006-09-04 Skapad: 2006-09-04 Senast uppdaterad: 2013-11-27
  • 112.
    Schön, Thomas
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    On computational methods for nonlinear estimation2003Licentiatavhandling, monografi (Övrigt vetenskapligt)
  • 113. Beställ onlineKöp publikationen >>
    Schön, Thomas B.
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Estimation of Nonlinear Dynamic Systems: Theory and Applications2006Doktorsavhandling, sammanläggning (Övrigt vetenskapligt)
    Abstract [en]

    This thesis deals with estimation of states and parameters in nonlinear and non-Gaussian dynamic systems. Sequential Monte Carlo methods are mainly used to this end. These methods rely on models of the underlying system, motivating some developments of the model concept. One of the main reasons for the interest in nonlinear estimation is that problems of this kind arise naturally in many important applications. Several applications of nonlinear estimation are studied.

    The models most commonly used for estimation are based on stochastic difference equations, referred to as state-space models. This thesis is mainly concerned with models of this kind. However, there will be a brief digression from this, in the treatment of the mathematically more intricate differential-algebraic equations. Here, the purpose is to write these equations in a form suitable for statistical signal processing.

    The nonlinear state estimation problem is addressed using sequential Monte Carlo methods, commonly referred to as particle methods. When there is a linear sub-structure inherent in the underlying model, this can be exploited by the powerful combination of the particle filter and the Kalman filter, presented by the marginalized particle filter. This algorithm is also known as the Rao-Blackwellized particle filter and it is thoroughly derived and explained in conjunction with a rather general class of mixed linear/nonlinear state-space models. Models of this type are often used in studying positioning and target tracking applications. This is illustrated using several examples from the automotive and the aircraft industry. Furthermore, the computational complexity of the marginalized particle filter is analyzed.

    The parameter estimation problem is addressed for a relatively general class of mixed linear/nonlinear state-space models. The expectation maximization algorithm is used to calculate parameter estimates from batch data. In devising this algorithm, the need to solve a nonlinear smoothing problem arises, which is handled using a particle smoother. The use of the marginalized particle filter for recursive parameterestimation is also investigated.

    The applications considered are the camera positioning problem arising from augmented reality and sensor fusion problems originating from automotive active safety systems. The use of vision measurements in the estimation problem is central to both applications. In augmented reality, the estimates of the camera’s position and orientation are imperative in the process of overlaying computer generated objects onto the live video stream. The objective in the sensor fusion problems arising in automotive safety systems is to provide information about the host vehicle and its surroundings, such as the position of other vehicles and the road geometry. Information of this kind is crucial for many systems, such as adaptive cruise control, collision avoidance and lane guidance.

    Delarbeten
    1. Marginalized Particle Filters for Mixed Linear/Nonlinear State-Space Models
    Öppna denna publikation i ny flik eller fönster >>Marginalized Particle Filters for Mixed Linear/Nonlinear State-Space Models
    2005 (Engelska)Ingår i: IEEE Transactions on Signal Processing, ISSN 1053-587X, E-ISSN 1941-0476, Vol. 53, nr 7, s. 2279-2289Artikel i tidskrift (Refereegranskat) Published
    Abstract [en]

    The particle filter offers a general numerical tool to approximate the posterior density function for the state in nonlinear and non-Gaussian filtering problems. While the particle filter is fairly easy to implement and tune, its main drawback is that it is quite computer intensive, with the computational complexity increasing quickly with the state dimension. One remedy to this problem is to marginalize out the states appearing linearly in the dynamics. The result is that one Kalman filter is associated with each particle. The main contribution in this paper is the derivation of the details for the marginalized particle filter for a general nonlinear state-space model. Several important special cases occurring in typical signal processing applications will also be discussed. The marginalized particle filter is applied to an integrated navigation system for aircraft. It is demonstrated that the complete high-dimensional system can be based on a particle filter using marginalization for all but three states. Excellent performance on real flight data is reported.

    Ort, förlag, år, upplaga, sidor
    IEEE Signal Processing Society, 2005
    Nyckelord
    Kalman filter, Marginalization, Navigation systems, Nonlinear systems, Particle filter, State estimation
    Nationell ämneskategori
    Reglerteknik
    Identifikatorer
    urn:nbn:se:liu:diva-11749 (URN)10.1109/TSP.2005.849151 (DOI)
    Tillgänglig från: 2008-05-07 Skapad: 2008-05-07 Senast uppdaterad: 2017-12-13Bibliografiskt granskad
    2. Complexity Analysis of the Marginalized Particle Filter
    Öppna denna publikation i ny flik eller fönster >>Complexity Analysis of the Marginalized Particle Filter
    2005 (Engelska)Ingår i: IEEE Transactions on Signal Processing, ISSN 1053-587X, E-ISSN 1941-0476, Vol. 53, nr 11, s. 4408-4411Artikel i tidskrift (Refereegranskat) Published
    Abstract [en]

    In this paper, the computational complexity of the marginalized particle filter is analyzed and a general method to perform this analysis is given. The key is the introduction of the equivalent flop measure. In an extensive Monte Carlo simulation, different computational aspects are studied and compared with the derived theoretical results.

    Ort, förlag, år, upplaga, sidor
    IEEE Signal Processing Society, 2005
    Nyckelord
    Complexity analysis, Kalman filter, Equivalent flop, Marginalized particle filter, Nonlinear estimation
    Nationell ämneskategori
    Reglerteknik
    Identifikatorer
    urn:nbn:se:liu:diva-11751 (URN)10.1109/TSP.2005.857061 (DOI)
    Tillgänglig från: 2008-05-07 Skapad: 2008-05-07 Senast uppdaterad: 2017-12-13
    3. A Modeling and Filtering Framework for Linear Differential-Algebraic Equations
    Öppna denna publikation i ny flik eller fönster >>A Modeling and Filtering Framework for Linear Differential-Algebraic Equations
    2003 (Engelska)Ingår i: Proceedings of the 42th IEEE Conference on Decision and Control, 2003, s. 892-897 vol.1Konferensbidrag, Publicerat paper (Refereegranskat)
    Abstract [en]

    General approaches to modeling, for instance using object-oriented software, lead to differential-algebraic equations (DAE). As the name reveals, it is a combination of differential and algebraic equations. For state estimation using observed system inputs and outputs in a stochastic framework similar to Kalman filtering, we need to augment the DAE with stochastic disturbances ("process noise"), whose covariance matrix becomes the tuning parameter. We will determine the subspace of possible causal disturbances based on the linear DAE model. This subspace determines all degrees of freedom in the filter design, and a Kalman filter algorithm is given. We illustrate the design on a system with two interconnected rotating masses.

    Nyckelord
    Implicit systems, Descriptor systems, Singular systems, White noise, Noise, Discretization, Kalman filters
    Nationell ämneskategori
    Teknik och teknologier Reglerteknik
    Identifikatorer
    urn:nbn:se:liu:diva-13917 (URN)10.1109/CDC.2003.1272679 (DOI)000189434100154 ()0-7803-7924-1 (ISBN)
    Konferens
    42nd IEEE Conference on Decision and Control, Maui, HI, USA, December, 2003
    Tillgänglig från: 2006-09-04 Skapad: 2006-09-04 Senast uppdaterad: 2013-11-27
    4. A Note on State Estimation as a Convex Optimization Problem
    Öppna denna publikation i ny flik eller fönster >>A Note on State Estimation as a Convex Optimization Problem
    2003 (Engelska)Ingår i: Proceedings of the 2003 IEEE International Conference on Acoustics, Speech, and Signal Processing, 2003, Vol. 6, nr 6-10, s. 61-64 vol.6Konferensbidrag, Publicerat paper (Refereegranskat)
    Abstract [en]

    The Kalman filter computes the maximum a posteriori (MAP) estimate of the states for linear state space models with Gaussian noise. We interpret the Kalman filter as the solution to a convex optimization problem, and show that we can generalize the MAP state estimator to any noise with a log-concave density function and any combination of linear equality and convex inequality constraints on the states. We illustrate the principle on a hidden Markov model, where the state vector contains probabilities that are positive and sum to one.

    Nyckelord
    State estimation, Kalman filter, Convex optimization, Hidden Markov Models
    Nationell ämneskategori
    Teknik och teknologier Reglerteknik
    Identifikatorer
    urn:nbn:se:liu:diva-13918 (URN)10.1109/ICASSP.2003.1201618 (DOI)0-7803-7663-3 (ISBN)
    Konferens
    2003 IEEE International Conference on Acoustics, Speech, and Signal Processing, Hong Kong, China, April, 2003
    Tillgänglig från: 2006-09-04 Skapad: 2006-09-04 Senast uppdaterad: 2013-11-27
    5. Particle Filters for System Identification of State-Space Models Linear in Either Parameters or States
    Öppna denna publikation i ny flik eller fönster >>Particle Filters for System Identification of State-Space Models Linear in Either Parameters or States
    2003 (Engelska)Ingår i: Proceedings of the 13th IFAC Symposium on System Identification, 2003, s. 1251-1256 vol.1Konferensbidrag, Publicerat paper (Refereegranskat)
    Abstract [en]

    The potential use of the marginalized particle filter for nonlinear system identification is investigated. The particle filter itself offers a general tool for estimating unknown parameters in non-linear models of moderate complexity, and the basic trick is to model the parameters as a random walk (so called roughening noise) with decaying variance. We derive algorithms for systems which are non-linear in either the parameters or the states, but not both generally. In these cases, marginalization applies to the linear part, which firstly significantly widens the scope of the particle filter to more complex systems, and secondly decreases the variance in the linear parameters/states for fixed filter complexity. This second property is illustrated on an example of chaotic model. The particular case of freely parametrized linear state space models, common in subspace identification approaches, is bi-linear in states and parameters, and thus both cases above are satisfied. One can then choose which one to marginalize.

    Nyckelord
    System identification, Nonlinear estimation, Recursive estimation, Particle filters, Kalman filters, Bayesian estimation
    Nationell ämneskategori
    Teknik och teknologier Reglerteknik
    Identifikatorer
    urn:nbn:se:liu:diva-13919 (URN)978-0080437095 (ISBN)
    Konferens
    13th IFAC Symposium on System Identification, Rotterdam, The Netherlands, August, 2003
    Tillgänglig från: 2006-09-04 Skapad: 2006-09-04 Senast uppdaterad: 2013-11-27
    6. Maximum Likelihood Nonlinear System Estimation
    Öppna denna publikation i ny flik eller fönster >>Maximum Likelihood Nonlinear System Estimation
    2006 (Engelska)Ingår i: Proceedings of the 14th IFAC Symposium on System Identification, Newcastle, Australia, 2006, s. 1003-1008Konferensbidrag, Publicerat paper (Refereegranskat)
    Abstract [en]

    This paper is concerned with the parameter estimation of a relatively general class of nonlinear dynamic systems. A Maximum Likelihood (ML) framework is employed in the interests of statistical efficiency, and it is illustrated how an Expectation Maximisation (EM) algorithm may be used to compute these ML estimates. An essential ingredient is the employment of so-called "particle smoothing" methods to compute required conditional expectations via a Monte Carlo approach. A simulation example demonstrates the efficacy of these techniques.

    Nyckelord
    Nonlinear systems, System identification, Maximum likelihood, Expectation maximisation algorithm, Particle smoother
    Nationell ämneskategori
    Teknik och teknologier Reglerteknik
    Identifikatorer
    urn:nbn:se:liu:diva-13920 (URN)10.3182/20060329-3-AU-2901.00160 (DOI)978-3-902661-02-9 (ISBN)
    Konferens
    14th IFAC Symposium on System Identification, Newcastle, Australia, March, 2006
    Tillgänglig från: 2006-09-04 Skapad: 2006-09-04 Senast uppdaterad: 2013-04-07
    7. Integrated Navigation of Cameras for Augmented Reality
    Öppna denna publikation i ny flik eller fönster >>Integrated Navigation of Cameras for Augmented Reality
    2005 (Engelska)Ingår i: Proceedings of the 16th IFAC World Congress, 2005, s. 187-187Konferensbidrag, Publicerat paper (Refereegranskat)
    Abstract [en]

    In augmented reality, the position and orientation of the camera must be estimated very accurately. This paper will propose a filtering approach, similar to integrated navigation in aircraft, which is based on inertial measurements as primary sensor on which dead-reckoning can be based, and features in the image as supporting information to stabilize the dead-reckoning. The image features are considered to be sensor signals in a Kalman filter framework.

    Nyckelord
    Sensor fusion, Kalman filter, Inertial navigation, Augmented reality, Computer vision, Feature extraction
    Nationell ämneskategori
    Teknik och teknologier Reglerteknik
    Identifikatorer
    urn:nbn:se:liu:diva-13921 (URN)10.3182/20050703-6-CZ-1902.00188 (DOI)978-3-902661-75-3 (ISBN)
    Konferens
    16th IFAC World Congress, Prague, Czech Republic, July, 2005
    Tillgänglig från: 2006-09-04 Skapad: 2006-09-04 Senast uppdaterad: 2013-03-23
    8. The Marginalized Particle Filter in Practice
    Öppna denna publikation i ny flik eller fönster >>The Marginalized Particle Filter in Practice
    2006 (Engelska)Ingår i: Proceedings of the 2006 IEEE Aerospace Conference, 2006Konferensbidrag, Publicerat paper (Refereegranskat)
    Abstract [en]

    The marginalized particle filter is a powerful combination of the particle filter and the Kalman filter, which can be used when the underlying model contains a linear sub-structure, subject to Gaussian noise. This paper will illustrate several positioning and target tracking applications, solved using the marginalized particle filter. Furthermore, we analyze several properties of practical importance, such as its computational complexity and how to cope with quantization effects.

    Nyckelord
    Gaussian noise, Adaptive Kalman filters, Computational complexity, Particle filtering (numerical methods), Position control, Quantisation (signal), Target tracking, Linear sub-structure, Marginalized particle filter, Positioning, Quantization effects
    Nationell ämneskategori
    Teknik och teknologier Reglerteknik
    Identifikatorer
    urn:nbn:se:liu:diva-13922 (URN)10.1109/AERO.2006.1655922 (DOI)9780780395459 (ISBN)
    Konferens
    2006 IEEE Aerospace Conference, Big Sky, MT, USA, March, 2006
    Forskningsfinansiär
    Vinnova
    Tillgänglig från: 2006-09-04 Skapad: 2006-09-04 Senast uppdaterad: 2013-04-07
    9. Lane Departure Detection for Improved Road Geometry Estimation
    Öppna denna publikation i ny flik eller fönster >>Lane Departure Detection for Improved Road Geometry Estimation
    2006 (Engelska)Ingår i: Proceedings of the 2006 IEEE Intelligent Vehicle Symposium, 2006, s. 546-551Konferensbidrag, Publicerat paper (Refereegranskat)
    Abstract [en]

    An essential part of future collision avoidance systems is to be able to predict road curvature. This can be based on vision data, but the lateral movement of leading vehicles can also be used to support road geometry estimation. This paper presents a method for detecting lane departures, including lane changes, of leading vehicles. This information is used to adapt the dynamic models used in the estimation algorithm in order to accommodate for the fact that a lane departure is in progress. The goal is to improve the accuracy of the road geometry estimates, which is affected by the motion of leading vehicles. The significantly improved performance is demonstrated using sensor data from authentic traffic environments.

    Nyckelord
    Automotive tracking, Change detection, State estimation, Kalman filter, CUSUM-test
    Nationell ämneskategori
    Teknik och teknologier Reglerteknik
    Identifikatorer
    urn:nbn:se:liu:diva-13923 (URN)10.1109/IVS.2006.1689685 (DOI)
    Konferens
    2006 IEEE Intelligent Vehicle Symposium, Tokyo, Japan, June, 2006
    Tillgänglig från: 2006-09-04 Skapad: 2006-09-04 Senast uppdaterad: 2013-02-26
    Ladda ner fulltext (pdf)
    FULLTEXT01
  • 114.
    Schön, Thomas B.
    et al.
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Gustafsson, Fredrik
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Karlsson, Rickard
    NIRA Dynamics AB, Linköping.
    Particle Filter in Practice2011Ingår i: The Oxford Handbook of Nonlinear Filtering / [ed] Dan Crisan, Boris Rozovskii, Oxford, UK: Oxford University Press, 2011Kapitel i bok, del av antologi (Refereegranskat)
    Abstract [en]

    In many areas of human endeavor, the systems involved are not available for direct measurement. Instead, by combining mathematical models for a system's evolution with partial observations of its evolving state, we can make reasonable inferences about it. The increasing complexity of the modern world makes this analysis and synthesis of high-volume data an essential feature in many real-world problems. The celebrated Kalman-Bucy filter, designed for linear dynamical systems with linearly structured measurements, is the most famous Bayesian filter. Its generalizations to nonlinear systems and/or observations are collectively referred to as nonlinear filtering (NLF), an extension of the Bayesian framework to the estimation, prediction, and interpolation of nonlinear stochastic dynamics. NLF uses a stochastic model to make inferences about an evolving system and is a theoretically optimal algorithm.The breadth of its applications, firmly established and still emerging, is simply astounding. Early uses such as cryptography, tracking, and guidance were mostly of a military nature. Since then, the scope has exploded. It includes the study of global climate, estimating the state of the economy, identifying tumors using non-invasive methods, and much more.The Oxford Handbook of Nonlinear Filtering is the first comprehensive written resource for the subject. It contains classical and recent results and applications, with contributions from 58 authors. Collated into 10 parts, it covers the foundations of nonlinear filtering, connections to stochastic partial differential equations, stability and asymptotic analysis, estimation and control, approximation theory and numerical methods for solving the nonlinear filtering problem (including particle methods). It also contains a part dedicated to the application of nonlinear filtering to several problems in mathematical finance.

  • 115.
    Schön, Thomas
    et al.
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Eidehall, Andreas
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Gustafsson, Fredrik
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Lane Departure Detection for Improved Road Geometry Estimation2005Rapport (Övrigt vetenskapligt)
    Abstract [en]

    An essential part of future collision avoidance systems is to be able to predict road curvature. This can be based on vision data, but the lateral movement of leading vehicles can also be used to support road geometry estimation. This paper presents a method for detecting lane departures, including lane changes, of leading vehicles. This information is used to adapt the dynamic models used in the estimation algorithm in order to accommodate for the fact that a lane departure is in progress. The goal is to improve the accuracy of the road geometry estimates, which is affected by the motion of leading vehicles. The significantly improved performance is demonstrated using sensor data from authentic traffic environments.

    Ladda ner fulltext (pdf)
    FULLTEXT01
  • 116.
    Schön, Thomas
    et al.
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Eidehall, Andreas
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Gustafsson, Fredrik
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Lane Departure Detection for Improved Road Geometry Estimation2006Ingår i: Proceedings of the 2006 IEEE Intelligent Vehicle Symposium, 2006, s. 546-551Konferensbidrag (Refereegranskat)
    Abstract [en]

    An essential part of future collision avoidance systems is to be able to predict road curvature. This can be based on vision data, but the lateral movement of leading vehicles can also be used to support road geometry estimation. This paper presents a method for detecting lane departures, including lane changes, of leading vehicles. This information is used to adapt the dynamic models used in the estimation algorithm in order to accommodate for the fact that a lane departure is in progress. The goal is to improve the accuracy of the road geometry estimates, which is affected by the motion of leading vehicles. The significantly improved performance is demonstrated using sensor data from authentic traffic environments.

  • 117.
    Schön, Thomas
    et al.
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Gerdin, Markus
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Glad, Torkel
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Gustafsson, Fredrik
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    A Modeling and Filtering Framework for Linear Differential-Algebraic Equations2003Ingår i: Proceedings of the 42th IEEE Conference on Decision and Control, 2003, s. 892-897 vol.1Konferensbidrag (Refereegranskat)
    Abstract [en]

    General approaches to modeling, for instance using object-oriented software, lead to differential-algebraic equations (DAE). As the name reveals, it is a combination of differential and algebraic equations. For state estimation using observed system inputs and outputs in a stochastic framework similar to Kalman filtering, we need to augment the DAE with stochastic disturbances ("process noise"), whose covariance matrix becomes the tuning parameter. We will determine the subspace of possible causal disturbances based on the linear DAE model. This subspace determines all degrees of freedom in the filter design, and a Kalman filter algorithm is given. We illustrate the design on a system with two interconnected rotating masses.

  • 118.
    Schön, Thomas
    et al.
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Gerdin, Markus
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Glad, Torkel
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Gustafsson, Fredrik
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    A Modeling and Filtering Framework for Linear Implicit Systems2003Rapport (Övrigt vetenskapligt)
    Abstract [en]

    General approaches to modeling, for instance using object-oriented software, lead to differential algebraic equations (DAE), also called implicit systems. For state estimation using observed system inputs and outputs in a stochastic framework similar to Kalman filtering, we need to augment the DAE with stochastic disturbances (process noise), whose covariance matrix becomes the tuning parameter. We will determine the subspace of possible causal disturbances based on the linear DAE model. This subspace determines all degrees of freedom in the filter design, and a Kalman filter algorithm is given.We illustrate the design on a system with two interconnected rotating masses.

    Ladda ner fulltext (pdf)
    FULLTEXT01
  • 119.
    Schön, Thomas
    et al.
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Gustafsson, Fredrik
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    A Note on State Estimation as a Convex Optimization Problem2003Rapport (Övrigt vetenskapligt)
    Abstract [en]

    The Kalman filter computes the maximum a posteriori (MAP) estimate of the states for linear state space models with Gaussian noise. We interpret the Kalman filter as the solution to a convex optimization problem, and show that we can generalize the MAP state estimator to any noise with a log-concave density function and any combination of linear equality and convex inequality constraints on the states. We illustrate the principle on a hidden Markov model, where the state vector contains probabilities that are positive and sum to one.

    Ladda ner fulltext (pdf)
    FULLTEXT01
  • 120.
    Schön, Thomas
    et al.
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Gustafsson, Fredrik
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Integrated Navigation of Cameras for Augmented Reality2004Rapport (Övrigt vetenskapligt)
    Abstract [en]

    In augmented reality, the position and orientation of the camera must be estimated very accurately. This paper will propose a filtering approach, similar to integrated navigation in aircraft, which is based on inertial measurements as primary sensor on which dead-reckoning can be based, and features in the image as supporting information to stabilize the dead-reckoning. The image features are considered to be sensor signals in a Kalman filter framework.

    Ladda ner fulltext (pdf)
    FULLTEXT01
  • 121.
    Schön, Thomas
    et al.
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Gustafsson, Fredrik
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Integrated Navigation of Cameras for Augmented Reality2005Ingår i: Proceedings of the 16th IFAC World Congress, 2005, s. 187-187Konferensbidrag (Refereegranskat)
    Abstract [en]

    In augmented reality, the position and orientation of the camera must be estimated very accurately. This paper will propose a filtering approach, similar to integrated navigation in aircraft, which is based on inertial measurements as primary sensor on which dead-reckoning can be based, and features in the image as supporting information to stabilize the dead-reckoning. The image features are considered to be sensor signals in a Kalman filter framework.

  • 122.
    Schön, Thomas
    et al.
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Gustafsson, Fredrik
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Particle Filters for System Identification of State-Space Models Linear in either Parameters or States2004Ingår i: Proceedings of Reglermöte 2004, 2004, s. 1-6Konferensbidrag (Övrigt vetenskapligt)
    Abstract [en]

    The potential use of the marginalized particle filter for nonlinear system identification is investigated. The particle filter itself offers a general tool for estimating unknown parameters in non-linear models of moderate complexity, and the basic trick is to model the parameters as a random walk (so called roughening noise) with decaying variance. We derive algorithms for systems which are non-linear in either the parameters or the states, but not both generally. In these cases, marginalization applies to the linear part, which firstly significantly widens the scope of the particle filter to more complex systems, and secondly decreases the variance in the linear parameters/states for fixed filter complexity. This second property is illustrated on an example of chaotic model. The particular case of freely parametrized linear state space models, common in subspace identification approaches, is bi-linear in states and parameters, and thus both cases above are satisfied. One can then choose which one to marginalize.

  • 123.
    Schön, Thomas
    et al.
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Gustafsson, Fredrik
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Particle Filters for System Identification of State-Space Models Linear in either Parameters or States2003Rapport (Övrigt vetenskapligt)
    Abstract [en]

    The potential use of the marginalized particle filter for nonlinear system identification is investigated. The particle filter itself offers a general tool for estimating unknown parameters in non-linear models of moderate complexity, and the basic trick is to model the parameters as a random walk (so called roughening noise) with decaying variance. We derive algorithms for systems which are non-linear in either the parameters or the states, but not both generally. In these cases, marginalization applies to the linear part, which firstly significantly widens the scope of the particle filter to more complex systems, and secondly decreases the variance in the linear parameters/states for fixed filter complexity. This second property is illustrated on an example of chaotic model. The particular case of freely parametrized linear state space models, common in subspace identification approaches, is bi-linear in states and parameters, and thus both cases above are satisfied. One can then choose which one to marginalize.

    Ladda ner fulltext (pdf)
    FULLTEXT01
  • 124.
    Schön, Thomas
    et al.
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Gustafsson, Fredrik
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Particle Filters for System Identification of State-Space Models Linear in Either Parameters or States2003Ingår i: Proceedings of the 13th IFAC Symposium on System Identification, 2003, s. 1251-1256 vol.1Konferensbidrag (Refereegranskat)
    Abstract [en]

    The potential use of the marginalized particle filter for nonlinear system identification is investigated. The particle filter itself offers a general tool for estimating unknown parameters in non-linear models of moderate complexity, and the basic trick is to model the parameters as a random walk (so called roughening noise) with decaying variance. We derive algorithms for systems which are non-linear in either the parameters or the states, but not both generally. In these cases, marginalization applies to the linear part, which firstly significantly widens the scope of the particle filter to more complex systems, and secondly decreases the variance in the linear parameters/states for fixed filter complexity. This second property is illustrated on an example of chaotic model. The particular case of freely parametrized linear state space models, common in subspace identification approaches, is bi-linear in states and parameters, and thus both cases above are satisfied. One can then choose which one to marginalize.

  • 125.
    Schön, Thomas
    et al.
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Gustafsson, Fredrik
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Hansson, Anders
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    A Note on State Estimation as a Convex Optimization Problem2003Ingår i: Proceedings of the 2003 IEEE International Conference on Acoustics, Speech, and Signal Processing, 2003, Vol. 6, nr 6-10, s. 61-64 vol.6Konferensbidrag (Refereegranskat)
    Abstract [en]

    The Kalman filter computes the maximum a posteriori (MAP) estimate of the states for linear state space models with Gaussian noise. We interpret the Kalman filter as the solution to a convex optimization problem, and show that we can generalize the MAP state estimator to any noise with a log-concave density function and any combination of linear equality and convex inequality constraints on the states. We illustrate the principle on a hidden Markov model, where the state vector contains probabilities that are positive and sum to one.

  • 126.
    Schön, Thomas
    et al.
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Gustafsson, Fredrik
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Nordlund, Per-Johan
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Marginalized Particle Filters for Mixed Linear/Nonlinear State-Space Models2005Ingår i: IEEE Transactions on Signal Processing, ISSN 1053-587X, E-ISSN 1941-0476, Vol. 53, nr 7, s. 2279-2289Artikel i tidskrift (Refereegranskat)
    Abstract [en]

    The particle filter offers a general numerical tool to approximate the posterior density function for the state in nonlinear and non-Gaussian filtering problems. While the particle filter is fairly easy to implement and tune, its main drawback is that it is quite computer intensive, with the computational complexity increasing quickly with the state dimension. One remedy to this problem is to marginalize out the states appearing linearly in the dynamics. The result is that one Kalman filter is associated with each particle. The main contribution in this paper is the derivation of the details for the marginalized particle filter for a general nonlinear state-space model. Several important special cases occurring in typical signal processing applications will also be discussed. The marginalized particle filter is applied to an integrated navigation system for aircraft. It is demonstrated that the complete high-dimensional system can be based on a particle filter using marginalization for all but three states. Excellent performance on real flight data is reported.

    Ladda ner fulltext (pdf)
    fulltext
  • 127.
    Schön, Thomas
    et al.
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Gustafsson, Fredrik
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Nordlund, Per-Johan
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Marginalized Particle Filters for Nonlinear State-space Models2003Rapport (Övrigt vetenskapligt)
    Abstract [en]

    The recently developed particle filter offers a general numerical tool to approximate the state a posteriori density in nonlinear and non-Gaussian filtering problems with arbitrary accuracy. Because the particle filter is fairly easy to implement and tune, it has quickly become a popular tool in signal processing applications. Its main drawback is that it is quite computer intensive. For a given filtering accuracy, the computational complexity increases quickly with the state dimension. One remedy to this problem is what in statistics is called Rao-Blackwellization, where states appearing linearly in the dynamics are marginalized out. This leads to that a Kalman filter is attached to each particle. Our main contribution here is to sort out when marginalization is possible for state space models, and to point out the implications in some typical signal processing applications. The methodology and impact in practice is illustrated on terrain navigation for aircrafts. The marginalized particle filter for a state-space model with nine states is evaluated on real aircraft data, and the result is that very good accuracy is achieved with quite reasonable complexity.

    Ladda ner fulltext (pdf)
    FULLTEXT01
  • 128.
    Schön, Thomas
    et al.
    Linköpings universitet, Tekniska högskolan. Linköpings universitet, Institutionen för systemteknik, Reglerteknik.
    Karlsson, G Rickard
    Linköpings universitet, Tekniska högskolan. Linköpings universitet, Institutionen för systemteknik, Reglerteknik.
    Gustafsson, Fredrik
    Linköpings universitet, Tekniska högskolan. Linköpings universitet, Institutionen för systemteknik, Reglerteknik.
    The Marginalized Particle Filter - Analysis, Applications and Generalizations2006Ingår i: Workshop on Sequential Monte Carlo Methods: filtering and other applications,2006, 2006Konferensbidrag (Refereegranskat)
    Abstract [en]

    The marginalized particle filter is a powerful combination of the particle filter and the Kalman filter, which can be used when the underlying model contains a linear sub-structure, subject to Gaussian noise. This paper will briefly introduce the marginalized particle filter and hint at possible generalizations, giving rise to a larger family of marginalized nonlinear filters. Furthermore, we analyze several properties of the marginalized particle filter, including its ability to reduce variance and its computational complexity. Finally, we provide an introduction to various applications of the marginalized particle filter. 

  • 129.
    Schön, Thomas
    et al.
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Karlsson, Rickard
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Gustafsson, Fredrik
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    The Marginalized Particle Filter: Analysis, Applications and Generalizations2006Ingår i: Proceedings of the 2006 Workshop on Sequential Monte Carlo Methods: filtering and other applications, 2006, s. 53-64Konferensbidrag (Refereegranskat)
    Abstract [en]

    The marginalized particle filter is a powerful combination of the particle filter and the Kalman filter, which can beused when the underlying model contains a linear sub-structure, subject to Gaussian noise. This paper will briefly introduce the marginalized particle filter and hint at possible generalizations, giving rise to a larger family of marginalized nonlinear filters. Furthermore, we analyze several properties of the marginalized particle filter, including its ability to reduce variance and its computational complexity. Finally, we provide an introduction to various applications of the marginalized particle filter.

  • 130.
    Schön, Thomas
    et al.
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Karlsson, Rickard
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Gustafsson, Fredrik
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    The Marginalized Particle Filter in Practice2005Rapport (Övrigt vetenskapligt)
    Abstract [en]

    The marginalized particle filter is a powerful combination of the particle filter and the Kalman filter, which can be used when the underlying model contains a linear sub-structure, subject to Gaussian noise. This paper will illustrate several positioning and target tracking applications, solved using the marginalized particle filter. Furthermore, we analyze several properties of practical importance, such as its computational complexity and how to cope with quantization effects.

    Ladda ner fulltext (pdf)
    FULLTEXT01
  • 131.
    Schön, Thomas
    et al.
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Karlsson, Rickard
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Gustafsson, Fredrik
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    The Marginalized Particle Filter in Practice2006Ingår i: Proceedings of the 2006 IEEE Aerospace Conference, 2006Konferensbidrag (Refereegranskat)
    Abstract [en]

    The marginalized particle filter is a powerful combination of the particle filter and the Kalman filter, which can be used when the underlying model contains a linear sub-structure, subject to Gaussian noise. This paper will illustrate several positioning and target tracking applications, solved using the marginalized particle filter. Furthermore, we analyze several properties of practical importance, such as its computational complexity and how to cope with quantization effects.

  • 132.
    Schön, Thomas
    et al.
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Karlsson, Rickard
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Törnqvist, David
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Gustafsson, Fredrik
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    A Framework for Simultaneous Localization and Mapping Utilizing Model Structure2007Ingår i: Proceedings of the 10th International Conference on Information Fusion, 2007, s. 1-8Konferensbidrag (Refereegranskat)
    Abstract [en]

    This contribution aims at unifying two recent trends in applied particle filtering (PF). The first trend is the major impact in simultaneous localization and mapping (SLAM) applications, utilizing the FastSLAM algorithm. The second one is the implications of the marginalized particle filter (MPF) or the Rao-Blackwellized particle filter (RBPF) in positioning and tracking applications. An algorithm is introduced, which merges FastSLAM and MPF, and the result is an MPF algorithm for SLAM applications, where state vectors of higher dimensions can be used. Results using experimental data from a 3D SLAM development environment, fusing measurements from inertial sensors (accelerometer and gyro) and vision are presented.

  • 133.
    Schön, Thomas
    et al.
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Karlsson, Rickard
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Törnqvist, David
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Gustafsson, Fredrik
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    A Framework for Simultaneous Localization and Mapping Utilizing Model Structure2007Rapport (Övrigt vetenskapligt)
    Abstract [en]

    The basic nonlinear ltering problem for dynamical systems is considered. Approximating the optimal lter estimate by particle lter methods has become perhaps the most common and useful method in recent years. Many variants of particle lters have been suggested, and there is an extensive lit- erature on the theoretical aspects of the quality of the approximation. Still,a clear cut result that the approximate solution, for unbounded functions, converges to the true optimal estimate as the number of particles tends to innity seems to be lacking. It is the purpose of this contribution to give such a basic convergence result.

    Ladda ner fulltext (pdf)
    FULLTEXT01
  • 134.
    Schön, Thomas
    et al.
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Roll, Jacob
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Ego-Motion and Indirect Road Geometry Estimation Using Night Vision2009Rapport (Övrigt vetenskapligt)
    Abstract [en]

    The sensors present in modern premium cars deliver a wealth of information. We will in this work illustrate one way of making better use of the sensor information already present in modern premium cars. More specifically, we will show how a far infrared (FIR) camera can be used to enhance the estimates of the vehicle ego-motion and indirectly the road geometry in 3D. The FIR camera is primarily intended for pedestrian detection. The solution is obtained by solving a suitable sensor fusion problem, where we merge information from proprioceptive sensors with the FIR camera images. In order to illustrate the performance of the proposed method we have made use of measurement sequences recorded during night-time driving on rural roads in Sweden. The results illustrate that the FIR images can be used to improve the ego-motion estimation, especially during night time driving.

    Ladda ner fulltext (pdf)
    FULLTEXT01
  • 135.
    Schön, Thomas
    et al.
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Roll, Jacob
    Linköpings universitet, Tekniska högskolan. Linköpings universitet, Institutionen för systemteknik, Reglerteknik.
    Ego-Motion and Indirect Road Geometry Estimation Using Night Vision2009Ingår i: Proceedings of the '09 IEEE Intelligent Vehicle Symposium, 2009, s. 30-35Konferensbidrag (Refereegranskat)
    Abstract [en]

    The sensors present in modern premium cars deliver a wealth of information. We will in this work illustrate one way of making better use of the sensor information already present in modern premium cars. More specifically, we will show how a far infrared (FIR) camera can be used to enhance the estimates of the vehicle ego-motion and indirectly the road geometry in 3D. The FIR camera is primarily intended for pedestrian detection. The solution is obtained by solving a suitable sensor fusion problem, where we merge information from proprioceptive sensors with the FIR camera images. In order to illustrate the performance of the proposed method we have made use of measurement sequences recorded during night-time driving on rural roads in Sweden. The results illustrate that the FIR images can be used to improve the ego-motion estimation, especially during night time driving.

  • 136.
    Schön, Thomas
    et al.
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Törnqvist, David
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Gustafsson, Fredrik
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Fast Particle Filters for Multi-Rate Sensors2007Rapport (Övrigt vetenskapligt)
    Abstract [en]

    Computational complexity is a major concern for practical use of the versatile particle filter (PF) for nonlinear filtering applications. Previous work to mitigate the inherent complexity includes the marginalized particle filter (MPF), with the fastSLAM algorithm as one important case. MPF utilizes a linear Gaussian sub-structure in the problem, where the Kalman filter (KF) can be applied. While this reduces the state dimension in the PF, the present work aims at reducing the sampling rate of the PF. The algorithm is derived for a class of models with linear Gaussian dynamic model and two multirate sensors, with different sampling rates, one slow with a nonlinear and/or non-Gaussian measurement relation and one fast with a linear Gaussian measurement relation. For this case, the KF is used to process the information from the fast sensor and the information from the slow sensor is processed using the PF. The problem formulation covers the important special case of fast dynamics and one slow sensor, which appears in many navigation and tracking problems.

    Ladda ner fulltext (pdf)
    FULLTEXT01
  • 137.
    Schön, Thomas
    et al.
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Törnqvist, David
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Gustafsson, Fredrik
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Fast Particle Filters for Multi-Rate Sensors2007Ingår i: Proceedings of the 15th European Signal Processing Conference, 2007Konferensbidrag (Refereegranskat)
    Abstract [en]

    Computational complexity is a major concern for practical use of the versatile particle filter (PF) for nonlinear filtering applications. Previous work to mitigate the inherent complexity includes the marginalized particle filter (MPF), with the fastSLAM algorithm as one important case. MPF utilizes a linear Gaussian sub-structure in the problem, where the Kalman filter (KF) can be applied. While this reduces the state dimension in the PF, the present work aims at reducing the sampling rate of the PF. The algorithm is derived for a class of models with linear Gaussian dynamic model and two multirate sensors, with different sampling rates, one slow with a nonlinear and/or non-Gaussian measurement relation and one fast with a linear Gaussian measurement relation. For this case, the KF is used to process the information from the fast sensor and the information from the slow sensor is processed using the PF. The problem formulation covers the important special case of fast dynamics and one slow sensor, which appears in many navigation and tracking problems.

  • 138.
    Schön, Thomas
    et al.
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Wills, Adrian
    University of Newcastle, Australia.
    Ninness, Brett
    University of Newcastle, Australia.
    Maximum Likelihood Nonlinear System Estimation2005Rapport (Övrigt vetenskapligt)
    Abstract [en]

    This paper is concerned with the parameter estimation of a relatively general class of nonlinear dynamic systems. A Maximum Likelihood (ML) framework is employed in the interests of statistical efficiency, and it is illustrated how an Expectation Maximisation (EM) algorithm may be used to compute these ML estimates. An essential ingredient is the employment of so-called "particle smoothing" methods to compute required conditional expectations via a Monte Carlo approach. A simulation example demonstrates the efficacy of these techniques.

    Ladda ner fulltext (pdf)
    FULLTEXT01
  • 139.
    Schön, Thomas
    et al.
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Wills, Adrian
    University of Newcastle, Australia.
    Ninness, Brett
    University of Newcastle, Australia.
    Maximum Likelihood Nonlinear System Estimation2006Ingår i: Proceedings of the 14th IFAC Symposium on System Identification, Newcastle, Australia, 2006, s. 1003-1008Konferensbidrag (Refereegranskat)
    Abstract [en]

    This paper is concerned with the parameter estimation of a relatively general class of nonlinear dynamic systems. A Maximum Likelihood (ML) framework is employed in the interests of statistical efficiency, and it is illustrated how an Expectation Maximisation (EM) algorithm may be used to compute these ML estimates. An essential ingredient is the employment of so-called "particle smoothing" methods to compute required conditional expectations via a Monte Carlo approach. A simulation example demonstrates the efficacy of these techniques.

  • 140.
    Schön, Thomas
    et al.
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Wills, Adrian
    University of Newcastle, Australia.
    Ninness, Brett
    University of Newcastle, Australia.
    System Identification of Nonlinear State-Space Models2010Rapport (Övrigt vetenskapligt)
    Abstract [en]

    This paper is concerned with the parameter estimation of a general class of nonlinear dynamic systems in state-space form. More specifically, a Maximum Likelihood (ML) framework is employed and an Expectation Maximisation (EM) algorithm is derived to compute these ML estimates. The Expectation (E) step involves solving a nonlinear state estimation problem, where the smoothed estimates of the states are required. This problem lends itself perfectly to the particle smoother, which provides arbitrarily good estimates. The maximisation (M) step is solved using standard techniques from numerical optimisation theory. Simulation examples demonstrate the efficacy of our proposed solution.

    Ladda ner fulltext (pdf)
    FULLTEXT01
  • 141.
    Schön, Thomas
    et al.
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Wills, Adrian
    University of Newcastle, Australia.
    Ninness, Brett
    University of Newcastle, Australia.
    System Identification of Nonlinear State-Space Models2011Ingår i: Automatica, ISSN 0005-1098, E-ISSN 1873-2836, Vol. 47, nr 1, s. 39-49Artikel i tidskrift (Refereegranskat)
    Abstract [en]

    This paper is concerned with the parameter estimation of a general class of nonlinear dynamic systems in state-space form. More specifically, a Maximum Likelihood (ML) framework is employed and an Expectation Maximisation (EM) algorithm is derived to compute these ML estimates. The Expectation (E) step involves solving a nonlinear state estimation problem, where the smoothed estimates of the states are required. This problem lends itself perfectly to the particle smoother, which provides arbitrarily good estimates. The maximisation (M) step is solved using standard techniques from numerical optimisation theory. Simulation examples demonstrate the efficacy of our proposed solution.

    Ladda ner fulltext (pdf)
    FULLTEXT01
  • 142.
    Sjanic, Zoran
    et al.
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Skoglund, Martin A.
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Schön, Thomas B.
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Gustafsson, Fredrik
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    A Nonlinear Least-Squares Approach to the SLAM Problem2011Ingår i: Proceedings of the 18th IFAC World Congress, 2011: World Congress, Volume # 18, Part 1 / [ed] Sergio Bittanti, Angelo Cenedese and Sandro Zampieri, IFAC Papers Online, 2011, s. 4759-4764Konferensbidrag (Refereegranskat)
    Abstract [en]

    In this paper we present a solution to the simultaneous localisation and mapping (SLAM) problem using a camera and inertial sensors. Our approach is based on the maximum a posteriori (MAP) estimate of the complete SLAM problem. The resulting problem is posed in a nonlinear least-squares framework which we solve with the Gauss-Newton method. The proposed algorithm is evaluated on experimental data using a sensor platform mounted on an industrial robot. In this way, accurate ground truth is available, and the results are encouraging.

  • 143.
    Sjanic, Zoran
    et al.
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Skoglund, Martin
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Schön, Thomas
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Gustafsson, Fredrik
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Solving The SLAM Problem for Unmanned Aerial Vehicles Using Smoothed Estimates2010Rapport (Övrigt vetenskapligt)
    Abstract [en]

    In this paper we present a solution to the simultaneous localization and mapping (SLAM) problem for unmanned aerial vehicles (UAV) using a camera and inertial sensors. A good SLAM solution is an important enabler for autonomous robots. Our approach is based on an optimization based formulation of the problem, which results in a smoother, rather than a filter. The proposed algorithm is evaluated on experimental data and the resultsare compared with accurate ground truth data. The results from this comparisons are encouraging.

    Ladda ner fulltext (pdf)
    FULLTEXT01
  • 144.
    Taghavi, Ehsan
    et al.
    School of Computational Science and Engineering, McMaster University.
    Lindsten, Fredrik
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Svensson, Lennart
    Division of Signals and Systems, Chalmers University.
    Schön, Thomas B.
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Adaptive stopping for fast particle smoothing2013Ingår i: Proceedings of the 38th International Conference on Acoustics, Speech, and Signal Processing (ICASSP), IEEE , 2013, s. 6293-6297Konferensbidrag (Refereegranskat)
    Abstract [en]

    Particle smoothing is useful for offline state inference and parameter learning in nonlinear/non-Gaussian state-space models. However, many particle smoothers, such as the popular forward filter/backward simulator (FFBS), are plagued by a quadratic computational complexity in the number of particles. One approach to tackle this issue is to use rejection-sampling-based FFBS (RS-FFBS), which asymptotically reaches linear complexity. In practice, however, the constants can be quite large and the actual gain in computational time limited. In this contribution, we develop a hybrid method, governed by an adaptive stopping rule, in order to exploit the benefits, but avoid the drawbacks, of RS-FFBS. The resulting particle smoother is shown in a simulation study to be considerably more computationally efficient than both FFBS and RS-FFBS.

    Ladda ner fulltext (pdf)
    fulltext
  • 145.
    Tidefelt, Henrik
    et al.
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Schön, Thomas
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Robust Point-Mass Filters on Manifolds2010Rapport (Övrigt vetenskapligt)
    Abstract [en]

    Robust state estimation for states evolving on compact manifolds is achieved by employing a point-mass filter. The proposed implementation emphasizes a sane treatment of the geometry of the problem, and advocates separation of the filtering algorithms from the implementation of particular manifolds.

    Ladda ner fulltext (pdf)
    FULLTEXT01
  • 146.
    Tidefelt, Henrik
    et al.
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Schön, Thomas
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Robust Point-Mass Filters on Manifolds2009Ingår i: Proceedings of the 15th IFAC Symposium on System Identification, 2009, s. 540-545Konferensbidrag (Refereegranskat)
    Abstract [en]

    Robust state estimation for states evolving on compact manifolds is achieved by employing a point-mass filter. The proposed implementation emphasizes a sane treatment of the geometry of the problem, and advocates separation of the filtering algorithms from the implementation of particular manifolds.

  • 147.
    Törnqvist, David
    et al.
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Schön, Thomas
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Gustafsson, Fredrik
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Detecting Spurious Features using Parity Space2008Rapport (Övrigt vetenskapligt)
    Abstract [en]

    Detection of spurious features is instrumental in many computer vision applications. The standard approach is feature based, where extracted features are matched between the image frames. This approach requires only vision, but is computer intensive and not yet suitable for real-time applications. We propose an alternative based on algorithms from the statistical fault detection literature. It is based on image data and an inertial measurement unit (IMU). The principle of analytical redundancy is applied to batches of measurements from a sliding time window. The resulting algorithm is fast and scalable, and requires only feature positions as inputs from the computer vision system. It is also pointed out that the algorithm can be extended to also detect nonstationary features (moving targets for instance). The algorithm is applied to real data from an unmanned aerial vehicle in a navigation application.

    Ladda ner fulltext (pdf)
    FULLTEXT01
  • 148.
    Törnqvist, David
    et al.
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Schön, Thomas
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Gustafsson, Fredrik
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Detecting Spurious Features using Parity Space2008Ingår i: Proceedings of the 10th International Conference on Control, Automation, Robotics and Vision, 2008, s. 353-358Konferensbidrag (Refereegranskat)
    Abstract [en]

    Detection of spurious features is instrumental in many computer vision applications. The standard approach is feature based, where extracted features are matched between the image frames. This approach requires only vision, but is computer intensive and not yet suitable for real-time applications. We propose an alternative based on algorithms from the statistical fault detection literature. It is based on image data and an inertial measurement unit (IMU). The principle of analytical redundancy is applied to batches of measurements from a sliding time window. The resulting algorithm is fast and scalable, and requires only feature positions as inputs from the computer vision system. It is also pointed out that the algorithm can be extended to also detect nonstationary features (moving targets for instance). The algorithm is applied to real data from an unmanned aerial vehicle in a navigation application.

  • 149.
    Törnqvist, David
    et al.
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Schön, Thomas
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Karlsson, Rickard
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Gustafsson, Fredrik
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Particle Filter SLAM with High Dimensional Vehicle Model2008Rapport (Övrigt vetenskapligt)
    Abstract [en]

    This work presents a particle filter (PF) method closely related to FastSLAM for solving the simultaneous localization and mapping (SLAM) problem. Using the standard FastSLAM algorithm, only low-dimensional vehicle models can be handled due to computational constraints. In this work an extra factorization of the problem is introduced that makes high-dimensional vehicle models computationally feasible. Results using experimental data from a UAV (helicopter) are presented. The algorithm fuses measurements from on-board inertial sensors (accelerometer and gyro), barometer, and vision in order to solve the SLAM problem.

    Ladda ner fulltext (pdf)
    FULLTEXT01
  • 150.
    Törnqvist, David
    et al.
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Schön, Thomas
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Karlsson, Rickard
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Gustafsson, Fredrik
    Linköpings universitet, Institutionen för systemteknik, Reglerteknik. Linköpings universitet, Tekniska högskolan.
    Particle Filter SLAM with High Dimensional Vehicle Model2009Ingår i: Journal of Intelligent and Robotic Systems, ISSN 0921-0296, E-ISSN 1573-0409, Vol. 55, nr 4, s. 249-266Artikel i tidskrift (Refereegranskat)
    Abstract [en]

    This work presents a particle filter (PF) method closely related to FastSLAM for solving the simultaneous localization and mapping (SLAM) problem. Using the standard FastSLAM algorithm, only low-dimensional vehicle models can be handled due to computational constraints. In this work an extra factorization of the problem is introduced that makes high-dimensional vehicle models computationally feasible. Results using experimental data from a UAV (helicopter) are presented. The algorithm fuses measurements from on-board inertial sensors (accelerometer and gyro), barometer, and vision in order to solve the SLAM problem.

1234 101 - 150 av 159
RefereraExporteraLänk till träfflistan
Permanent länk
Referera
Referensformat
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • oxford
  • Annat format
Fler format
Språk
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Annat språk
Fler språk
Utmatningsformat
  • html
  • text
  • asciidoc
  • rtf