Energy Saving Control System of Long Stroke Pumping Unit Based on RBF Neural Network

Based on the highly nonlinear electromagnetism characteristics switched reluctance motor (SRM) of long stroke pumping, traditional PID controller can´t achieve good performance index and meet energy-saving requirements. This paper presents a novel approach of RBF neural network PID adaptive control for SRM based on RBF neural network on-line identification and learning algorithm of variable learning rate. The experimental results show that a high control performance is achieved. The control method has fast response, small overshoot, strong robustness and adaptivity, and the system has better energy-saving effect.