3 resultados para iterative learning

em Chinese Academy of Sciences Institutional Repositories Grid Portal


Relevância:

100.00% 100.00%

Publicador:

Resumo:

为实现对模型不确定的有约束非线性系统在特定时间域上输出轨迹的有效跟踪,将改进的克隆选择算法用于求解迭代学习控制中的优化问题。提出基于克隆选择算法的非线性优化迭代学习控制。在每次迭代运算后,一个克隆选择算法用于求解下次迭代运算中的最优输入,另一个克隆选择算法用于修正系统参考模型。仿真结果表明,该方法比GA-ILC具有更快的收敛速度,能够有效处理输入上的约束以及模型不确定问题,通过少数几次迭代学习就能取得满意的跟踪效果。

Relevância:

70.00% 70.00%

Publicador:

Resumo:

In this paper, a disturbance controller is designed for making robotic system behave as a decoupled linear system according to the concept of internal model. Based on the linear system, the paper presents an iterative learning control algorithm to robotic manipulators. A sufficient condition for convergence is provided. The selection of parameter values of the algorithm is simple and easy to meet the convergence condition. The simulation results demonstrate the effectiveness of the algorithm..

Relevância:

60.00% 60.00%

Publicador:

Resumo:

本文为机器人机械手提出了一种基于离散时间的重复学习控制法,这种学习控制法利用机器人动力学模型的部分知识,从它的特性和实用观点看,这种控制法比现有的其它学习控制法更有吸引力.本文还给出了学习控制法的收敛性证明和计算机仿真结果。