A neural network-based design method of the fractional order PID controller for a class of motion control systems

作者：Zheng, Weijia^*; Chen, YangQuan; Wang, Xiaohong; Lin, Meijin

来源：ASIAN JOURNAL OF CONTROL, 2022, 24(6): 3378-3393.

DOI：10.1002/asjc.2727

• fractional order control
• motion control
• neural network
• proportional-integral-derivative control

A neural network-based design method for the fractional order proportional-integral-derivative ( PI lambda D mu) controller is developed in this paper. The tuning process is divided into two steps: the integral and derivative orders of the controller are estimated by the neural networks, which are established according to the design specifications and the features of the plant; the remaining controller parameters are calculated according to the flat-phase specification. In this way, the optimal PI lambda D mu controller can be obtained without the optimization process. The effectiveness of the proposed method is studied via the application to a motion control system. The control performance of the controller tuned by the proposed method is compared with those tuned by some existing methods. Simulation and experimental results show that the controller obtained using the proposed method can offer good tracking performance and disturbance rejection performance for the control system.

• 单位
  佛山科学技术学院