Stochastic gradient algorithm for dual-rate sampled data nonlinear systems based on the missing outputs identification model

This paper combines a missing outputs identification model with a stochastic gradient algorithm to estimate the parameters of dual-rate sampled data systems with a preload nonlinear input. A switching function is provided to turn the model of the nonlinear systems into an identification model. Based on the identification model, a missing outputs identification model is introduced to predict the unmeasurable outputs and a stochastic gradient algorithm is derived to identify the parameters of the identification model by all the inputs and all the outputs. The simulation results indicate that the proposed algorithm is effective.