Research On Robust Fusion Estimation Of Nonlinear Systems

Along with the rapid development of information era,people put forward higher and higher requirements for the accuracy of information processing and filtering system.However,in the practical application,the system still exists some uncertainties,which will reduce the performance of the filter and even lead to filter divergence.Therefore,in view of the uncertain system parameters,this paper has done the following research:(1)For the nonlinear systems with uncertain parameters,based on the recursive linear minimum variance estimation(RLMVE)framework,the paper analyzes the effect of the system parameters on the mean squared error(MSE).The result was that the greater the system parameters deviation(whether they were positive or negative deviation)is,the greater the MSE is.Based on the analyzed results and prior probability statistics(probability distributions or sampling distributions),an innovative robust estimation algorithm was proposed for nonlinear systems with uncertain system parameters.Three examples verified the effectiveness of the proposed algorithms.(2)For the uncertain nonlinear fusion systems(covariance intersection fusion,matrix weighted fusion and centralized fusion)with uncertain system parameters(noise variance or model parameters),it also proved that,the estimation error obtained by fusion always is larger than the estimation error obtained by the system without deviation,and the greater the deviation is,the greater the MSE is.Based on the conclusion,a nonlinear robust fusion estimation method was proposed.Finally,the superiority of the algorithm was verified by simulation.