论文：受限机器人的神经鲁棒力位置控制.pdf

受限机器人的神经鲁棒力/位置控制 王洪瑞 郭继丽 温淑焕 （燕山大学电气工程学院，河北，秦皇岛066004） 摘 要： 考虑了受限机器人的力/位置鲁棒控制问题。首先使用微分几何方法，将受限机器人力/位置控 制的动力学模型，转化为非线性系统。然后应用输入输出线性化方法以及 Lyapunov 函数法，推导出鲁棒 跟踪控制器。所获得的控制器可确保系统按指数规律跟踪期望轨迹和期望的约束力，同时相应闭环系统的 状态一致最终有界。文中还利用一个RBF 神经网络学习系统不确定性的未知上界，提高控制精度。仿真算 例验证了控制律的有效性。 关键词： 受限机器人；非线性系统；神经网络；鲁棒控制。 Neural Robust force /Position control for Constrained Robot WANG Hong-rui GUO Ji-li WEN Shu-huan （Institute of Electrical Engineering, Yanshan University, Qinhuangdao 066004, China ） Abstract ： The robust force/position control of constrained robot manipulators was considered. First, using differential geometry theory, uncertain nonlinear system was derived from the force/position nonlinear dynamic model of the constrained robot. Second, robust tracking controller is obtained by employing input/output feedback linearization approach and Lyapunov method. The proposed controller guarantees that the system can track the desired output with exponential law and all states of the obtained closed-loop system are uniformly ultimately bounded. Still, an RBF neural network used to learn the unknown upper bound of system uncertainties effectively eliminate the effects of system uncertainties. Simulation result of an illustrative example justifies the effectiveness of the control law 。 Key words: constrained robot;nonlinear systems;neural network;robust control.