论文部分内容阅读
本文以自由漂浮空间机器人为被控对象,针对机器人手臂在捕获大质量目标后,如何保证控制精度这一问题,提出基于神经网络的鲁棒控制策略.由于测量误差和外界干扰等因素的影响,空间机器人系统存在参数和非参数两种不确定性,利用径向基函数神经网络来学习系统的未知参数不确定性,设计隐层和输出层之间权值参数学习律来保证自适应在线调整,鲁棒控制器补偿非参数不确定性和神经网络的逼近误差,利用冗余机理来集成两种控制器,从而确保了捕获后的控制精度,基于Lyapunov理论证明了整个闭环系统全局渐进稳定.所提控制方法无需空间机械臂精确数学模型,仿真结果表明了这种控制器的有效性,且由于空间机械臂的低速工况,这为神经网络提供了充足的学习时间,保证了工程应用的实时性.该控制方法对于国防、航空航天及其安全领域具有重要的工程应用价值.
In this paper, a free-floating space robot is a controlled object, and a Robust control strategy based on neural network is proposed to solve the problem of how to ensure the control accuracy after the robot arm catches a large target.For the influence of measurement error and outside interference, The space robot system has two kinds of uncertainties, parametric and non-parametric. The radial basis function neural network is used to study the unknown parameter uncertainty of the system, and the parameter learning rule between hidden layer and output layer is designed to ensure adaptive online adjustment , The robust controller compensates for the non-parametric uncertainty and the approximation error of the neural network, and uses the redundant mechanism to integrate the two controllers so as to ensure the control accuracy after acquisition. The global asymptotic stability of the closed-loop system is proved based on Lyapunov theory. The proposed control method does not require an accurate mathematical model of the space manipulator. The simulation results show the effectiveness of this controller. Due to the low speed of the space manipulator, this provides enough time for the neural network to learn, Real-time. The control method for the defense, aerospace and safety areas have an important project application price value.