Abstract
In order to solve the problems of model performance error and abnormal measurement value, the square-root Cubature Kalman filter (SRCKF) algorithm has problems such as filter performance degradation or filter effect divergence, this paper proposes a self-estimation based on M. Adaptive Square-root Cubature Kalman Filter based on M estimation (M-ASRCKF) algorithm, this filter algorithm uses an improved Huber weight function to improve the robustness of the SRCKF algorithm; and then adds the Sage-Husa order Excellent unbiased maximum a posteriori (MAP) estimator, real-time monitoring and estimation of unknown or inaccurate noise statistics, improving the tracking accuracy of the traditional SRCKF algorithm. MATLAB simulation results show that the proposed M-estimation-based adaptive SRCKF (M-ASRCKF) algorithm has better anti-non-linearity and anti-outlier interference capabilities, and this algorithm has strong robustness and high accuracy of results.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call