Neural networks based PID adaptive control of redundant servo system

A mathematical model of redundant direct drive electro-hydraulic servo system is built up. However it is not easy to control such a system precisely for it is time-variant and non-linear, besides that there is an unsteady transitional state during switching between different redundancies and characteristic of the servo system deteriorates with the degrading of the redundancies. Considering these factors, an adaptive control algorithm based on neural networks (NN) is introduced to improve the traditional algorithm which has such disadvantages as poor self-adaptability and poor robustness. According to characteristics of the servo system and the development of the neural network nowadays, a RBF neural network based PID controller is adopted in the controlling of the redundant actuator to solve the above problems. Results of the simulation show that the controller adjusts controlling parameters according to the instructions and structure of the controlled object. It is of high precision and strongly adaptive and it solves the problems brought by redundancy technology.