Wiener recurrent neural network adaptive inverse controller of hydraulic flight motion simulator

The traditional control strategy of hydraulic flight motion simulator (HFMS) cannot meet the high tracking performance requirement, and the control precision is sensitive to disturbance. This paper presents an adaptive inverse controller based on the Wiener-type recurrent neural network (WRNN) to deal with the parametric uncertainties and uncertain nonlinearities in HFMS. The WRNN is a dynamic linear subsystem cascaded with a static nonlinear subsystem. The controller contains two WRNNs, one to identify the Jacobian information of the controlled plant and another to approximate the inverse model of the plant. Since the inverse transfer function behaves sensitive to the initial value, a feedback controller is designed. The input of the controlled plant includes the feedback controller output and the WRNN inverse controller output. Simulations have confirmed the effectiveness and superiority of the proposed WRNN adaptive inverse control strategy.