抄録
It is generally not easy to acquire high positioning accuracy in hydraulic servocontrol systems, since they have an intense non-linearity, such as dead zone, which is originally caused by the leakage and friction. It is considered that the Neural Network compensator is one of the available options to get over the inherent problem. This paper deals with the eect of dither/PWM signals on the steady position error in the hydraulic motor drive system with a state feedback Neural Network. From the computer simulation, it is clarified that the application of the dither/PWM signals to the Neural Network control system improves the positioning performance by accompanying the assistance of the Neural Network capability. In addition, the effectiveness of the dither/PWM signals is also verified by experimental results.
