基于神经网络的变量泵模型跟踪控制

Model-following Controller Based on Neural Network for Variable Displacement Pump

本文用神经网络逼近对象的非线性逆模型 ,通过神经网络模型参考控制实现电液比例斜盘倾角和流量的控制 ,使系统的不确定性和非线性等得到补偿。仿真和实验表明 ,所开发的神经网络控制器能够较好地实现VDAPP控制 ,有较好的自适应性和鲁棒性 ,跟踪性能有较大改善。

A neural network model-following controller was proposed to approach the nonlinear anti-model of the object,and to realize the control of electro-hydraulic swashplate angle and flow,so that the uncertainty and nonlinear of the system can be compensated.The results of simulation and experiment show that the proposed neural network controller can conduct nonlinear control in VDAPP,enhance adaptability and robustness,and improve the performance of the control system.