Zhang Muxiang, Ma Fulong, Xiao Guozhen. A New Neural Computing Method for Optimization[J]. Acta Electronica Sinica, 1993, (7): 1-7.DOI:
神经网络优化计算的新方法
摘要
本文在Hopfield神经网络优化方法的基础上
根据模拟退火算法逃离局部最优解的原理
提出了一种神经网络优化计算的新方法。通过调整神经网络的连接权
网络的演化不仅可以逃离目标函数的局部最优解
而且可以改善目标函数的局部最优解。实验结果表明
新方法求解最优解所需的计算时间比模拟退火算法少得多。
Abstract
According to the principle of escaping from local optimal solutions of simulated annealing algorithm
a new neural computing method for optimization is proposed based on Hopfield neural network method.By adjusting the connection weights of a neural network
the network can not only escape from local optimal solutions of objective functions
but also improve local optimal solutions of objective functions.Experimental results show that the new method requires much less computing time to obtain an optimal solution than simulated annealing algorithm does.