Fusion Estimation Of ICI And SICI States For Networked Systems With Multiplicative Noise

With the development of multi-sensor information fusion,the system with multiplicative noise attracts more and more attention of many researchers,and has made great breakthroughs in many research fields.This paper aimmed at the networked system with multiplicative noise,obtained the local estimator by Kalman filtering algorithm,and then applied the Inverse Covariance Intersection(ICI),the main research contents are as follows:1.The sequential idea of Sequential Covariance Intersection(SCI)fusion algorithm was integrated into the Inverse Covariance Intersectionfusion algorithm to obtain the Sequential Inverse Covariance Intersection(SICI)fusion algorithm.Based on the local Kalman filtering algorithm for the multi-sensor system with multiplicative noise,ICI and SICI fusion algorithm was applied to fuse the local information,and the corresponding fusion results are obtained.By comparing the fusion results of ICI and CI,SICI and SCI,it was concluded that the fusion accuracy of ICI and SICI was higher than the accuracy of CI and SCI respectively.2.The state fusion estimation of networked systems with multiplicative noise and two-step stochastic delay and networked systems with multiplicative noise and missing observations were studied respectively.Firstly,based on the fictitious noise technology and generalized Lyapunov equation method,the original system was transformed into a conventional system.Then the local estimation was obtained based on the local Kalman filtering algorithm.Finally,the fusion results were obtained by using ICI fusion algorithm and SICI fusion algorithm respectively.The results showed that ICI fusion algorithm and SICI fusion algorithm had better fusion accuracy than others.3.Matlab was used to simulate the above problems.The ICI fusion algorithm and SICI fusion algorithm were effective in dealing with the problem of networked system with multiplicative noise,which not only improved the accuracy,but also improved the calculation efficiency due to reducing the calculation of covariance.