NEURAL NETWORK CONTROL OF SLIDE POSITION OF MECHANICAL PRESS

A new computer control system which can control the slide of mechanical press so as to stop on a accurate position is proposed. The control system consists of pneumatic proportional valve, pneumatic friction clutch and brake, displacement sensor etc. It is obvious that neural network control strategy is appropriate to compensate for the nonlinearness of the system. The neural network controller is designed in details. A lot of experiments on mechanical press model JH23-63 demonstrate that the new computer control system works well and neural network control strategy is effective.