论文部分内容阅读
研究了混沌优化方法中混沌变量的初值设定和载波过程中放大倍数等参数调整的实用方法 .在此基础上 ,为了克服BP网络收敛速度慢和易陷入局部极小点的不足 ,提出了基于混沌优化的BP网络学习算法 ,该方法主要利用混沌运动的遍历性为梯度算法创造一个良好的搜索界面 .仿真结果表明 ,把混沌优化方法用于神经网络权值优化 ,方法简单可行 ,搜索速度快 ,是一种有效的新途径 .
In order to overcome the shortcomings of slow convergence rate of BP network and easy fall into local minimum point, this paper proposes a new method to adjust the initial value of chaotic variables and the magnification of carrier in the process of chaos optimization. This method mainly uses the ergodicity of chaotic motion to create a good search interface for gradient algorithm.The simulation results show that the chaos optimization method is used to optimize the weight of neural network, the method is simple and feasible, the search speed Fast, is an effective new way.