不确定机器人系统无模型自适应滑模控制方法

Model-free adaptive sliding mode control method for uncertain robot system

为了弥补不确定机器人系统模型不准确、参数时变的问题,该文提出了一种数据驱动的无模型自适应滑模控制方法。利用一种新的动态线性化方法转换不确定机器人动力学模型。采用数据驱动无模型自适应控制方法设计控制器。引入离散滑动模态指数趋近律保证其收敛性。以五自由度外骨骼上肢康复机器人为仿真对象,通过Sim Mechanics进行仿真实验。结果证明即使在无法建立准确模型的情况下,该文所提出的无模型自适应滑模控制方法也可使不确定时变的机器人系统沿着给定的轨迹运动且系统稳定。仿真结果证明了该方法的可行性。

To compensate for the problems of inaccurate and having time-varying parameters of a uncertain robot system model,a model-free adaptive sliding mode control method of data driven control is proposed here. A new dynamic linear method is used to transfer a uncertain robot dynamics model.A controller is designed using a model-free adaptive sliding mode control method of data driven control. The discrete sliding mode index reaching law is introduced to ensure the convergence. An exoskeleton upper-limb rehabilitation robot with five degrees of freedom is simulated by Sim Mechanics.The simulation results prove that even in the case of being unable to establish an accurate model,the model-free adaptive sliding mode control method proposed here can make the uncertain time-varying robot system move along a given path and the system is stable. Simulation results show the feasibility of this method.