An intelligent PID controller for DC motor using RBF neural network and Kalman filter

作者：Liu Ning^*; Su Zhong; Peng Shu Hua

来源：Journal of Harbin Institute of Technology, 2010, 17(SUPPL. 2): 33-38.

• Adaptive PID
• Adaptive PID controller
• Control performance
• Control strategies
• DC motor system
• Measurement Noise
• MiroSot
• Model reference
• Model reference adaptive
• Noise influence
• PID controllers
• PID parameters
• Radial basis function neural networks
• RBF Neural Network
• Simulation result
• Single neural
• System response time
• Tracking performance

In this paper, to improve the control performance of FIRA MiroSot robot';s DC motor, we develop a model reference single neural adaptive PID controller based on radial basis RBF) neural network and Kalman filter. To compensate for the FIRA MiroSot robot';s DC motor actuator and to reduction of output and measurement noise, we adopt model reference adaptive control strategy and Kalman filter. Moreover, a RBF neural network is used to identify the robots'; DC motor system on-line and then regulate the PID parameters on-line, which makes the system more adaptive and reliability, makes the system response time more shorter, makes the MCU code more shorter and makes the noise influence less. Simulation results show the controller has good tracking performance and good robustness. This controller used on the robot successfully. So the control strategy presented in this paper is effective.

• 出版日期2010