Research And Application Of Combined Adaptive Kalman Filtering Method Based On Weighted Correlation Method

In Kalman Filtering theory,the Weighted Correlation Method is an adaptive filtering method that uses Kalman filtering innovation to estimate unknown noise statistics.However,it is limited by the impact of dynamic model anomalies and gross error observations on filter state estimation.The Weighted Correlation Method can not obtain better quality parameter estimates.In order to overcome this limitation,Focusing on the adaptive robust Kalman filtering method and noise covariance estimation method,a Robust Weighted Correlation Method is proposed,and the correctness and effectiveness of the proposed method are verified by numerical simulation examples and GPS single-point positioning experiment examples.The main research work of the paper is summarized as follows:1.the research status of Kalman filtering method at home and abroad is summarizes,including noise covariance estimation method,adaptive factor-based filtering method and combined filtering method,pointing out the shortcomings of the above methods in dynamic data processing And the necessity of constructing reasonable filtering models and methods.2.In order to achieve the complementary advantages of weighted correlation method and adaptive robust filtering method,noise covariance estimation method based on Robust Weighted Correlation is proposed to resist the influence of unknown noise covariance and gross observation on parameter estimation and improve the accuracy of parameter estimation.At the same time,the correctness and effectiveness of the method are verified by numerical simulation examples.The numerical simulation results show that the method improves the estimation accuracy while ensuring the unbiased noise covariance estimation results.3.In order to verify the reliability of the Robust Weighted Correlation method in practical applications,the performance of the two noise covariance estimation methods,the Weighted Correlation Method and the Robust Weighted Correlation Method,is compared and analyzed in conjunction with the application of GPS single-point positioning.The experimental results show that: Robust WeightedCorrelation Method can effectively resist gross error and has strong resistance;Weighted Correlation Method can not resist gross error and has poor performance.