Symmetrical valve controlled asymmetrical cylinder based on wavelet neural network

Abstract

Purpose The symmetrical valve is usually used in the hydraulic servo control system to control the asymmetrical cylinder, but this system’s structure involves asymmetry, and so its dynamic characteristics are asymmetrical, which causes issues in the control system of symmetric response. The purpose of this paper is to achieve the aim of symmetric control. Design/methodology/approach In this paper, the authors proposed a method that combined wavelet neural network (WNN) and model reference adaptive control. The reference model determined the dynamic response that the system was expected to achieve, and the WNN adaptive control made the system follow the reference model to achieve the purpose of symmetric control. Findings The experimental results show that the method can achieve a more accurate symmetric control and position control compared with the solutions via the classical PID control. Originality/value The proposed combination of the WNN and the reference model can effectively compensate for the asymmetry of dynamic response of the asymmetric cylinder in forward and return directions, which can be extended to deal with other classes of applications.