多关节机器人的自学习模糊全局滑模控制

针对模型不确定性多关节机器人的轨迹跟踪控制问题,研究多关节机器人全局滑模控制,为了削弱系统在滑动模态上的抖振,将模糊控制和全局滑模控制相结合,提出一种自学习模糊全局滑模控制方法.该方法利用模糊系统的输出代替全局滑模控制中的非连续开关切换量,并根据滑模变结构原理,设计自学习算法,动态调整模糊隶属函数的参数.通过对2关节机器人的仿真,结果表明在存在模型误差和外部扰动的情况下,该方法既能达到快速跟踪,又能很好地消除控制器的抖振.

Global Fuzzy Self-learning Sliding Mode Control for Multi-link Robot Manipulators

In this paper global sliding control for multi-link robot manipulators is studied and a global fuzzy self-learning sliding mode controller is proposed,which combines fuzzy control and global sliding control to restrain the chattering around the sliding plane.The output of a fuzzy control system is substituted for the non-continuous switching Control volume and a self-learning algorithm is devised to regulate the fuzzy membership functions parameters according to the sliding mode variable structure theory.Some simulation results of a two-link robotic manipulator show that the control scheme can achieve tracking effect with high precision and speediness,as well as eliminate chattering of control under the condition of existing model error and external disturbances.