Cubature Kalman Filter Fusion Method Of Complex Nonlinear Systems

With the rapid development of science and technology,the actual engineering application environment is becoming more and more complex,and higher requirements are placed on the accuracy of target tracking and the reliability of the algorithm.Because the sensor is affected by various uncertain factors in a complex environment,the established model is seriously mismatched with the actual system,resulting in a decrease in estimated performance.For example,when the data obtained by the sensor contains outliers,the problem of detecting and processing the outliers,when the channel attenuation is not ideal during communication,the problem of multiplicative noise occurs in the measurement data.These issues will be a barrier to the application of Kalman filter technology in actual engineering environments.Therefore,it is of great significance to study the nonlinear Kalman filter algorithm in complex engineering environments.Aiming at the above problems,this paper studies the Cubature Kalman filter fusion method of complex nonlinear systems based on the analysis of aircraft flight test data in complex environment.The specific research contents are as follows:(1)A Fault-tolerant Cubature Kalman filter(CKF)algorithm for nonlinear systems is studied for the problem of outliers in measurement data.Firstly,the problem that the outliers occurrence causes the Cubature Kalman filter estimation performance mismatch is analyzed.Then,based on the error caused by the sampling approximation method and the introduction position of the adjustment factor,a new Fault-tolerant CKF filtering algorithm is proposed.Finally,by comparing the relative closeness of filter MSE and true MSE of the CKF and Fault-tolerant CKF,it provides theoretical support for the superiority of the Fault-tolerant CKF filtering algorithm.(2)Aiming at the problem that the nonlinear multiplicative noise is correlated and the correlation of noise is uncertain,the nonlinear adaptive CKF algorithm for multiplicative noise related systems is studied.Firstly,the problem of estimation performance mismatch of CKF algorithm under multiplicative noise is analyzed.The necessity of improving the filtering algorithm for multiplicative noise is proved.The CKF filtering algorithm based on multiplicative noise is proposed.Aiming at the inaccurate model parameters in actual engineering,the adaptive estimation filteringalgorithm based on correlated multiplicative noise is studied.Finally,the influence of multiplicative noise correlation on the performance of filter estimation is analyzed in depth,and the effectiveness of the proposed algorithm is proved theoretically.(3)The filtering estimation for single sensor can not satisfy the filtering accuracy problem in complex engineering environment.The fault-tolerant CKF filtering fusion algorithm is studied.Aiming at the shortcomings of no-reset mode federal filter fault isolation and the related multiplicative noise in the measurement data,a fault-tolerant CKF filtering algorithm with correlated multiplicative noise is proposed for the sub-filter.The fault detection method is improved for the filter fusion,which effectively improves the performance of the fusion estimation.Finally,the rationality of the fault-tolerant CKF filter fusion algorithm is theoretically analyzed.