基于Backstepping的双臂空间机器人惯性空间轨迹的鲁棒神经网络跟踪控制

Backstepping based robust ANN tracking in inertial space for space robot system with dual-arms

讨论了基于BackStepping方法,载体位置与姿态均不受控制的双臂空间机器人跟踪惯性空间期望轨迹的控制问题.首先基于拉格朗日第二类方法,结合系统动量(动量矩)守恒关系,推导得到了系统动力学方程,并转化为系统状态方程.基于Backstepping方法,针对具有不确定性的双臂空间机器人系统,设计了鲁棒自适应神经网络控制规律,保证了具有不确定性的双臂空间机器人系统末端手爪在惯性空间跟踪期望轨迹的控制.仿真实验证明了该方法的有效性.

Base on the backstepping method, the free-floating space robot with dual-arms tracking in inertial space is discussed. Considering the uncertainties of the system, the state function of the space robot system with dual-arms is established. Based on baekstepping method, a new adaptive robust ANN tracking control scheme is proposed. This scheme can guarantee the output tracking error＇s converging to zero while the states of the nonlinear system and the approximation errors of neural networks bounded. Simulation verified the proposed controller.