Multi-sensor Information Fusion Incremental Kalman Filter

Multi-sensor information fusion filtering theory has been widely used in many fields such as aviation,aerospace,navigation,industrial process control,target tracking and so on.Information fusion can make full use of the observation information from different sensors,so as to obtain an optimal description of the system state and ensure the reliability of the system.In complex environment,if all kinds of errors or error information for the multi-sensor system can be effectively identified and eliminated,the accuracy of system state estimation can be further improved.Kalman filtering algorithm is a very common state estimation method.Its recursive form and small data storage make it better than the other general filtering algorithms.However,in the actual application process,due to the influence of the surrounding environment,errors caused by the measuring equipment itself,or improper selection of models and parameters and other reasons,the systematic errors in measurement data often drift with time.Such system observation errors are often difficult to be verified or calibrated,and the direct use of traditional Kalman filtering algorithm will also cause large filtering errors.To solve this problem,the Kalman filtering algorithm and fusion algorithm for the multi-sensor systems under poor observation condition are studied based on the incremental equation in this paper.The main contents include the following aspects:Firstly,a new incremental equation is proposed for linear discrete systems under poor observation condition.Moreover,the incremental Kalman estimators are proposed based on two incremental equations.They can effectively solve the state estimation problem for the systems under poor observation condition,which can not be solved by the traditional Kalman filter algorithm.Secondly,under the linear minimum variance optimal fusion criterion,the multi-sensor weighted state and weighted measurement fusion incremental Kalmanestimators are presented for the systems under poor observation condition.They improve the state estimation accuracy for the multi-sensor systems under poor observation condition.Finally,considering the incremental observation noise as colored noise,the local and weighted measurement fusion incremental Kalman estimators with colored measurement noises are proposed.Compared with the incremental Kalman estimators with white measurement noises,the estimation accuracy is further improved.Applying above algorithms,the specific simulation application examples are given,and the simulation results show the effectiveness and practicability of the proposed algorithm in this paper.