A systematic design method for reducing bias in observers is developed. The method utilizes an observable default model of the system together with measurement data from the real system and estimates a model augmentation. The augmented model is then used to design an observer which reduces the estimation bias compared to an observer based on the default model. Three main results are a characterization of possible augmentations from observability perspectives, a parameterization of the augmentations from the method, and a robustness analysis of the proposed augmentation estimation method. The method is applied to a truck engine where the resulting augmented observer reduces the estimation bias by 50% in a European transient cycle.

A method for bias compensation and online map adaptation using extended Kalman filters isdeveloped. Key properties of the approach include the methods of handling component aging, varyingmeasurement quality including operating-point-dependent reliability and occasional outliers, andoperating-point-dependent model quality. Theoretical results about local and global observability,specifically adapted to the map adaptation problem, are proven. In addition, a method is presented tohandle covariance growth of locally unobservable modes, which is inherent in the map adaptationproblem. The approach is also applicable to the offline calibration of maps, in which case the onlyrequirement of the data is that the entire operating region of the system is covered, i.e., no specialcalibration cycles are required. The approach is applied to a truck engine in which an air mass-flowsensor adaptation map is estimated during a European transient cycle. It is demonstrated that themethod manages to find a map describing the sensor error in the presence of model errors on ameasurement sequence not specifically designed for adaptation. It is also demonstrated that themethod integrates well with traditional engineering tools, allowing prior knowledge about specificmodel errors to be incorporated and handled.

Maps or look-up tables are frequently used in engine control systems, and can be of dimension one or higher. Their use is often to describe stationary phenomena such as sensor characteristics or engine performance parameters like volumetric efficiency. Aging can slowly change the behavior, which can be manifested as a bias, and it can be necessary to adapt the maps. Methods for bias compensation and on-line map adaptation using extended Kalman filters are investigated and discussed. Key properties of the approach are ways of handling component aging, varying measurement quality, as well as operating point dependent model quality. Handling covariance growth on locally unobservable modes, which is an inherent property of the map adaptation problem, is also important and this is solved for the Kalman filter. The method is applicable to off-line calibration ofmaps where the only requirement of the data is that the entire operating region of the system is covered, i.e. no special calibration cycles are required. Two truck engine applications are evaluated, one where a 1-D air mass-ffow sensoradaptation map is estimated, and one where a 2-D volumetric efficiency map is adapted, both during a European transient cycle. An evaluation on experimental data shows that the method estimates a map, describing the sensor error, on a measurement sequence not specially designed for adaptation.

Computation and sampling time requirements for real-time implementation of observers is studied. A common procedure for state estimation and observer design is to have a system model in continuous time that is converted to sampled time with Euler forward method and then the observer is designed and implemented in sampled time in the real time system. When considering state estimation in real time control systems for production there are often limited computational resources. This becomes especially apparent when designing observers for stiff systems since the discretized implementation requires small step lengths to ensure stability. One way to reduce the computational burden, is to reduce the model stiffness by approximating the fast dynamics with instantaneous relations, transforming an ordinary differential equations (ODE) model into a differential algebraic equation (DAE) model. Performance and sampling frequency limitations for extended Kalman filter (EKF)s based on both the original ODE model and the reduced DAE model are here analyzed and compared for an industrial system. Furthermore, the effect of using backward Euler instead of forward Euler when discretizing the continuous time model is also analyzed. The ideas are evaluated using measurement data from a diesel engine. The engine is equipped with throttle, exhaust gas recirculation (EGR), and variable geometry turbines (VGT) and the stiff model dynamics arise as a consequence of the throttle between two control volumes in the air intake system. The process of simplifying and modifying the stiff ODE model to a DAE model is also discussed. The analysis of the computational effort shows that even though the ODE, for each time-update, is less computationally demanding than the resulting DAE, an EKF based on the DAE model achieves better estimation performance than one based on the ODE with less computational effort. The main gain with the DAE based EKF is that it allows increased step lengths without degrading the estimation performance compared to the ODE based EKF.

A method for bias compensation in model based estimation utilizing model augmentation is developed. Based on a default model, that suffers from stationary errors, and measurements from the system a low order augmentation is estimated. The method handles models described by differential algebraic equations and the main contributions are necessary and sufficient conditions for the preservation of the observability properties of the default model during the augmentation.