电液位置伺服系统的再励学习控制研究

针对非线性电液位置伺服系统的不确定性控制问题，提出了一种带有小脑模型（ＣＭＡＣ）神经网络的再励学习控制方法。将ＣＭＡＣ神经网络融入再励学习控制结构中，并进行了撞化与改进杯仅使再励学习控制器具备了泛化能力，而且提高了其学习速度，因此竽电液位置伺 服系统的快速跟踪控制。仿真结果表明，控制器不仅具有良好的处理非线性能力而且对时变外扰支具有明显的抑制作用。

Reinforcement Learning Control of an Electrohydraulic Position Servo System

A reinforcement learning control method is presented for position control of an electrohydraulic servo system with uncertainties, where use was made of the cerebellar model articulation controller (CMAC). The method integrates the CMAC neural network into a reinforcement learning control scheme, which improves learning ability and speed. The control algorithm is also simplified for fast tracking control of the electrohydraulic position servo system. The simulation results demonstrate that the presented method is of good robustness and tracking property in the presence of nonlinearity and time varying external load disturbances.