IDENTIFICATION OF ATYPICAL HYDRAULIC ACTUATOR SYSTEM USING A NEURAL NETWORK BASED ARX MODEL

摘要

This paper investigates the identification of a typical hydraulic actuator system using a model based on two-layer multi-layer perception neural network. It is shown that the neural based ARX model clearly outperforms the possible linear models in simulation and correlation characteristics of the prediction error and thus, it is possible to obtain reasonable results for the hydraulic actuator system with neural network based ARX model. As correlation functions of the prediction errors tend to remain confined more closely to the specified confidence regions, it is likely that most of the information has been extracted from the training set. It is by no means claimed that the "optimal" solution is found, however, it is shown that the proposed neural network based ARX model has provided a simple means for solving a challenging nonlinear problem, such as the one considered in this paper.