Research And Application On ANPID Control Algorithm For Electro-Hydraulic Servo System

The electro-hydraulic servo system has advantages such as high control accuracy, quick response, high power-weigh ratio, make itself to be widely used in military, aerospace and industry fields. Unfortunately, the electro-hydraulic servo system also has disadvantages like model uncertainty, load time-varying, system nonlinearity, so, the conventional linear control method can hardly get ideal control quality. Research on new type intelligent control algorithm for electro-hydraulic servo system is of great value in theoretical research and engineering application.The purpose of this paper is to improve the dynamic and static performance of electro-hydraulic servo system, research on the PID artificial neural network (ANPID) control algorithm, and apply it to practical engineering. Firstly, the paper introduces the classical PID control law, study on the combination of neural network and PID control. The existing ANPID control algorithm’s neurons transmit function is discontinuous differentiable, and the system error evaluation function adopted by the algorithm doesn’t make consideration on the actual characteristic of the servo system, the paper amend this two defects, and does a lot of research and design on the structure of algorithm, the control principle, initial weights setting, training algorithm and learning step which can guarantee convergence of the network. Finally, according to the three major working modes of testing machine control system, the paper accurately establishes the mathematical models of electro-hydraulic servo control system, and analysis the system’s dynamic and static performance in different working modes. This paper designs the PID and ANPID controller and applys them to the control system, conducts the computer simulation experiment and the prototype machine real-time control experiment.The experiments show that ANPID can intelligently adjust its parameters according to system error, has the feature of self-adaptive, when compared to the normal PID control law, the ANPID get the better control performance for the load time-varying, nonlinear servo system, this control algorithm has the value of popularizing in engineering application.