Nonlinear stochastic time-varying system identification based on multi-dimensional Taylor network with optimal structure

A novel method for nonlinear stochastic time-varying systems identification based on multi-dimensional Taylor network with optimal structure is proposed. In this paper, the connection weight coefficients of multi-dimensional Taylor network are regarded as the time-varying parameters, which are trained by the variable forgetting factor recursive least squares algorithm, to reflect the input-output change of nonlinear stochastic time-varying systems. Moreover, to avoid the dimension explosion, the weight-elimination algorithm is introduced to choose effective regression items of multi-dimensional Taylor network, thereby the simplest structure of network which has the best generalization ability is obtained. Simulation results show that the method proposed in this paper is valid.