Neural network internal feedback control for electrohydraulic servo loading

To solve the electrohydraulic servo loading extra force problem, a new control method is presented through feedback internal different force signals in loading system to suppress the extra force. To overcome the effects of the nonlinearity, parameter time variation and uncertainties, the neural network adaption based on internal reference model control scheme is further proposed in servo loading. The internal reference model is the load pressure dynamic model of the cylinder on condition of no disturbance caused by position system. In loading control, the measured result of load pressure is compared with the output of the ideal reference model. By controlling the neural network adaptive controller, the disturbances of velocity and acceleration of the position system are eliminated and the negative effects of hydraulic nonlinearity, parameter time variation and uncertainties are simultaneously reduced. The method is suitable for more applications than the conventional ones in servo loading.