神经网络迭代学习控制快速算法的研究

A Modified Algorithm for Iterative Learning Controllers Using Neural Networks

  • 摘要: 研究了非线性系统的神经网络迭代学习控制问题,为了提高神经网络学习控制的收敛速度,改进了权值的迭代学习算法,提出了一种新的训练权值方法,它以递推的方法来计算网络权值,不仅能减少计算量且收敛速度快,最后,用仿真结果对该算法作了进一步的验证。

     

    Abstract: An iterative leaming controller using neural networks is presnted for nonlinear systems. ln order to speed up the convergenoe of iterative learning controller algorthm, the improved algorithm is employed in neural networks, which develops an iterative me-thod to compute the weights ,and this leads to reduce the amount of computation and speed up the convergence of learning Finally the simulation results ard are given

     

/

返回文章
返回
Baidu
map