Research And Application Of Fuzzy Cubature Kalman Filter Based On Truncated Singular Value

In this paper,a TSVD-FCKF(Fuzzy Cubature Kalman Filter based on Truncated Singular Value Decomposition)based on truncated singular value is designed,and the algorithm is validated with combining the SOC(State of Charge)estimation of lithium-ion battery under complex operating conditions and temperature variations.The specific work of this paper is as follows:1.The robust filtering algorithm for uncertain model parameters is studied.In this paper,Further study of the SVD-CKF(Cubature Kalman Filter based on Singular Value Decomposition)algorithm shows that the smaller non-zero singular values in the SVD decomposition process are sensitive to noise,which easily leads to poor robustness of the filter.In this paper,a TSVD-CKF(Cubature Kalman Filter based on Truncated Singular Value Decomposition)is designed to enhance the robustness of the algorithm by truncating smaller singular values,so that the estimation results are more stable and the robust estimation of the target state is realized.2.A TS fuzzy controller is designed to adjust the influence of noise on the convergence speed of the filter.In this method,the residual error and its change rate of the measurement information are used as the input,and the fuzzy control based on the TS model is used.Through the correction of the measurement noise covariance,the influence of the measurement information on the Kalman gain is reduced when the measurement noise is large,so as to achieve the rapid adaptation of the filter to the unknown noise,improve its convergence speed,and then achieve the purpose of improving the accuracy of state estimation.The TS fuzzy controller is combined with TSVD-CKF to form the TSVD-FCKF algorithm,which is designed in this paper.Finally,the IMM(Interacting Multiple Model)estimation method based on the TSVD-FCKF algorithm is designed for the application in this paper and the specific calculation process is given.3.The IMM estimation method based on the TSVD-FCKF algorithm is applied to estimate the state of charge of the lithium-ion battery under complex working conditions and temperature changes.Firstly,the mathematical model of the experimental object and the identification method of model parameters are given,and then the basic performance of the lithium ion battery is tested.Finally,using two sets of battery data with different aging degree and temperature,the effectiveness of the algorithm is verified by comparing the estimated value with the measured value,and the advantages of the algorithm are verified by comparative experiments.The results show that the algorithm based on TSVD-FCKF can achieve robust and fast estimation of the target state under the conditions of uncertain model parameters and unknown noise statistical characteristics.