Sequential Fusion Estimation Of Complex Networked Systems

With the development of networked systems in many fields,the estimation for the mixed uncertainties networked systems has become an important direction of modern control theory,such as military,navigation,aviation and so on.For complex networked systems,this paper uses the Kalman Filter algorithm to obtain the local filter,and then applies the Sequential Fast Covariance Intersection(SFCI)fusion algorithm to obtain the fusion filter result.Compared with Sequential Covariance Intersection(SCI)fusion and Sequential Inverse Covariance Intersection(SICI)fusion,the algorithm can not only significantly reduce the computational burden and improve the fusion accuracy,but also the fusion results are not affected by the fusion order.The main research contents of this paper are as follows:1.On the basis of Fast Covariance Intersection(FCI)fusion algorithm,the sequential idea of fusion algorithm is integrated.For networked systems with random one-step lag and missing observations,based on local Kalman filtering algorithm,the FCI and SFCI fusion algorithms are respectively applied to fuse the local information,and the corresponding fusion filter is obtained.2.For mixed uncertain networked systems with multiplicative and correlated noise,Firstly,based on the augmented state method and fictitious noise technology,the mixed uncertain networked system model is converted into standard Kalman filter model,the Kalman filter algorithm is used to obtain the local filter estimation,then apply the SFCI fusion algorithm to obtain the fusion result,and compare with the SCI,SICI and weighted fusion by matrix algorithms.3.For networked system with fading measurement,multiplicative noise and correlated noise,Firstly,the system model is converted into a standard Kalman filter model based on model transformation and fictitious noise technology,the Kalman filter algorithm is applied to obtain the local estimation,then the SFCI fusion algorithm is used to obtain the fusion filter,and compare with the SCI,SICI and weighted fusion by matrix algorithms.