ABSTRACT
This chapter presents some of the widely used adaptive Kalman filtering algorithms. The adaptation algorithms are applicable to linear Kalman filter (KF) so the resulting algorithms are all for linear systems. The KF approach to the state estimation is quite sensitive to any uncertainties and malfunctions. If the condition of the real system does not correspond to the models, used in the synthesis of the filter, then these differences, which may be due to some possible failures in the sensors and actuators or any other uncertainties in the models, significantly decrease the performance of the estimation system. The KF can be made adaptive and hence insensitive to the uncertainties and faults by using various different techniques. The essence of this approach is to investigate the behavior of the innovation sequence and determine whether the real characteristics of the noise match with its a priori characteristics.
