Intelligent control of alternative current permanent magnet servomotor using neural network

A momentum back-propagation neural network is applied to control of alternative current permanent magnet servo-motors operating in a high performance drives environment to inhibit the network from converging at the local minimum point. The emerging backpropagation neural networks have the capability to learn arbitrary nonlinearity and show great potential for intelligent control applications. A scheme for combing momentum back-propagation neural networks with traditional PID control techniques is proposed. The control action emulated with the aid of MATLAB shows that the new self-turning scheme can deal with a large unknown non-linearity.