基于神经网络的不确定性空间机器人自适应控制方法研究

提出了一种针对自由漂浮状态的空间机器人模型不确定性的神经网络自适应控制方法。通过RBF神经网络逼近模型的非线性函数和不确定性上界,无需预先估计系统的不确定性程度和外部干扰,提出的自适应控制律保证了权值的有界性,解决了神经网络权值的UUB(Unknown Upper Bound)问题,即未知上界有界问题,完成了笛卡尔空间内空间机器人轨迹规划任务。证明了所提出的控制方法的稳定性,仿真结果表明控制方法避免了对空间机器人动力学模型的参数线性化要求降低了计算量,能够满足实际任务中的实时性要求。

Study on Neural Network Adaptive Control Method for Uncertain Space Manipulator

An adaptive control method based on neural network for the model uncertainty of free float space manipulator is proposed. A RBF neural network is used to approximate the nonlinear model of the space manipulator and learn the upper bound of the dynamic uncertainty. The neural network UUB problem is solved by the adaptive law, and path planning of free float space manipulator in Cartesian space is completed. The adaptive law for the weights of RBF neural network is presented which can ensure the stability of the space manipulator system. Simulations illustrate that the method avoids parameter linearization of the space manipulator dynamic model and decreases the computation, satisfy the actual mission of real-time as well.