【摘要】 针对一类非线性系统的跟踪控制问题 ,首先提出了一种遗忘因子迭代学习控制算法 ,给出了算法收敛的充分条件 ,然后 ,利用神经网络原理 ,对要求跟踪的新的期望轨迹 ,在系统的历史控制经验基础上 ,用神经网络估计系统的期望控制输入 ,然后将其作为迭代学习控制器的初始控制输入 ,再由迭代学习律逐步改善控制输入 ,使系统的实际输出只需较少的迭代次数就能达到跟踪的精度要求。机器人系统的仿真结果表明了该算法的有效性。更多还原

【Abstract】 An iterative learning controller based on neural network learning is presented for trajectory-tracked task of a class of nonlinear systems.In the first part of the paper,sufficient condition for learning algorithm with forgetting factor is derived to guaratee convergence of learning system.In the second part of the paper,desired control input of iterative learning controller is estimated by neural networks incorporated experience for a new desired trajectory-tracked task.If the selection of initial control input has considered previous experience of the controller for a new desired trajectory tracked task,then convergence of error can be improved and accuracy of tracking can be full only few iterative number.Simulation examples of robot show their effectiveness.更多还原