Adaptive and Nonlinear Kalman Filtering for GPS Navigation Processing
Dah‐Jing Jwo, Mu‐Yen Chen, Chien-Hao Tseng, Ta-Shun Cho · InTech eBooks · 2009
The divergence problem due to modelling errors is critical in Kalman filter applications. The conventional extended Kalman filter does not present the capability to monitor the change of parameters due to changes in vehicle dynamics. In this chapter, three feasible ways to avoid the divergence problem and to further improve the GPS navigation accuracy are discussed: (1) adaptive approaches assisted by heuristic search techniques to fit the dynamic model process of interest as precisely as possible; (2) utilization of an appropriate nonlinear estimation approach after deriving a better nonlinear dynamic process model; and (3) interactive multiple model approach accounting for different manoeuvring conditions. In Example 1, the FLAS is incorporated into the traditional strong tracking Kalman filter (STKF) approach for determining the softening factors, resulting in the adaptive fuzzy strong tracking Kalman filter (AFSTKF). Through the use of fuzzy logic, the FLAS has been employed as a mechanism for timely detecting the dynamical changes and implementing the on-line tuning of filter parameters by monitoring the innovation information so as to maintain good tracking capability. By use of the FLAS, lower order of filter model can be utilized and, therefore, less computational effort will be sufficient without compromising estimation accuracy significantly.