In this thesis we study a Bayesian estimation formulation of the target tracking problem. Traditionally, linear or linearized models are used, where the uncertainty in the sensor and motion models is typically modeled by Gaussian densities. Hence, classical sub-optimal Bayesian methods based on linearized Kalman filters can be used. The sequential Monte Carlo method, or particle filter, provides an approximative solution to the non-linear and non-Gaussian estimation problem. The particle filter approximates the optimal solution, hence it can outperform the Kalman filter in many cases, given sufficient computational resources. A survey over relevant tracking literature is presented including aspects as estimation, data association, sensor fusion and target modeling. In various target tracking related estimation and data association applications, we extend or modify particle filtering algorithms.
The passive ranging application when only angle information is available is discussed for several problems. In an air-to-sea application it is shown how to incorporate terrain induced constraints using a terrain database. The algorithm is also successfully evaluated on experimental sonar data acquired from a torpedo system.
In a multi-target data association application a simulation based approach for data association is proposed and compared to classical algorithms for an air-to-air tracking application. Moreover, the number of particles needed in the particle filter is adapted using a control structure to reduce the computational complexity.
Linköping: Linköpings universitet , 2002. , 124 p.