非完整移动机器人的神经网络鲁棒自适应控制

在非完整移动机器人轨迹跟踪问题中，针对机器人运动学与动力学模型的参数和非参数不确定性，提出了一种混合神经网络鲁棒自适应轨迹跟踪控制器，该控制器由运动学控制器和动力学控制器两部分组成；其中，采用了参数自适应的径向基神经网络对运动学模型的未知部分进行了建模，并采用权值在线调整的单层神经网络和自适应鲁棒控制项构成了动力学控制器；基于Lyapunov方法的设计过程保证了系统的稳定性和收敛性，仿真结果证明了算法的有效性。

Neural network-based robust adaptive control of mobile robot with nonholonomic constraints

In the trajectory tracking of nonholonomic mobile robot,for solving the unknown factors of mobile robot,such as parametric and nonparametric uncertainties of the kinematics and dynamic models, a robust adaptive controller based on neural network is proposed,which includes a kinematics controller and a dynamic controller.Radial basis function neural network with adaptive parameter is uesed for modeling unknown parts of the kinematics model, and the dynamic controller is based on single-layer neural network with on-line adjustment of weights and adaptive robust controller.The proposed controller can overcome the uncertainties and the disturbances.The system stability and the convergence of tracking system are proved by Lyapunov stability theory.Simulation results show the effectiveness of the proposed tracking control law.