Bayesian Inferring Inverse Model Design of Nonlinear System

In this paper, a bayesian inferring inverse model was proposed for nonlinear system. The model directly utilizes the nonlinear system running data and obtains the nonlinear inverse relationship by probability inferring formula. In training of the bayesian inferring inverse model, the evolutionary algorithms and sliding window method are adopted to realize the parameters estimation in the threshold matrix and the on-line prediction application of the bayesian inferring inverse model. Some nonlinear systems are taken to validate the modeling effectiveness of the bayesian inferring inverse model. And the simulation results show that the bayesian inferring inverse model provides a valid method for the inverse modeling problem of nonlinear system.