基于H_∞跟踪的不确定机器人模糊神经网络变结构控制

Tracking-Based Fuzzy Neural-Network Variable Structure Control for Uncertain Robot Manipulators

针对不确定机器人的跟踪控制,提出了一种基于H_∞跟踪的模糊神经网络变结构控制方案。该方案利用模糊神经网络的学习和函数逼近能力,补偿机器人系统的内部不确定性和外干扰,H_∞跟踪设计与变结构控制的结合削弱了控制输入的抖振,提高了机器人的鲁棒性,保证了整个系统的快速跟踪性能,仿真实例证实了该控制方案的有效性。

This paper presents a Fuzzy neural-network variable structure control based on H∞ tracking for uncertain robot manipulators. The scheme can use the learning ability and function approximation capability of the fuzzy neural-network to compensate the robotic internal uncertainties and external disturbances.The H∞ tracking design technique and the variable structure control improve the robustness for the robot, and weaken the chattering of the control input. Simulation results prove the control law can guarantee fast trajectory tracking performance for the overall system.