Two-stage Cubature Kalman Filtering Fusion Algorithms for Nonlinear Systems

The traditional two-stage Kalman filtering fusion method can not deal with the case of non-linear systems. which limits its application scope and level in practical engineering. Aiming at the above problems, this paper studies the design of two-stage cubature Kalman filter fusion algorithm for non-linear systems, and establishes a two-stage cubature Kalman filter fusion algorithm for multi-sensor systems. The main filter uses the output information of each local filter, fuses the global values, and feeds back to each local filter reasonably. The simulation results show that the proposed two-stage cubature Kalman filter fusion estimator (MTSCKF) has better performance than the two-stage cubature Kalman filter estimator (TSCKF) in filtering estimation accuracy.