摘要
在传统混沌神经网络模型的基础上,提出了一种具有衰减混沌噪声的混沌模拟退火神经网络模型(CSA-DCN)。该模型结合了 Hopfield 神经网络(HNN)与模拟退火算法(SA)的优点,并引入通过 Logistic 映射迭代函数产生的衰减混沌噪声,从而使该模型可以有效地解决高维、离散、非凸的非线性约束优化问题。例如电力系统经济负荷分配(ELD)问题,在考虑网损、阀点效应的情况下,将该模型应用于解决 ELD问题。通过多个算例仿真计算表明,该模型的算法是可行和有效的。CSA-DCN 模型是一种适用性很强的优化模型,可以应用于电力系统或其它行业系统的优化问题中。
Based on deeply discussing the principle of chaotic neural network model, the chaotic simulated annealing model with decaying chaotic noise (CSA-DCN) is presented. This model combines some advantages of Hopfield neural network (HNN) and simulated annealing (SA) algorithm, and decaying chaotic noise produced by iterated functions of Logistic map is inducted into this model, which makes it be used to solve many multidimensioned, discrete, non-convex, nonlinear constrained optimization problems, such as economic load dispatch (ELD) of power systems. Involved the transmission loss and valve point effect (VPE), the CSA-DCN model is applied to solve the ELD problem, simulation results of three examples show that the CSA-DCN model for the ELD problem is versatile, robust and efficient.
出处
《中国电机工程学报》
EI
CSCD
北大核心
2005年第3期65-70,共6页
Proceedings of the CSEE
