Wave Reproduction Simulation for Road Simulator with Iterative Learning Control Applied to Nonlinear Plant Model

摘要

This article applies iterative learning control (ILC) to road simulation test system and simulates the control system to reproduce a stochastic pavement profile. With uniform white noise as input, using actual measured input-output data and dynamic neural network, system nonlinear autoregressive moving average model (NARMA) was established. Regarding road simulator control mission as a perfect tracking problem in finite time interval, considering the pure time delay of system model, a P-type open loop iterative learning law was devised using output error and control parameters of learning law were determined with relative root mean square error (RMSE) of output error as iterative stop condition. By designed learning law and system NARMA model formulating road simulation ILC system, simulation went on. Simulation results prove that with different initial inputs, utilizing designed learning law, model output can approach target signal with expected precision and that designed iterative control system in this article is feasible.