摘要
在传统Hopfield神经网络的基础上,提出了自适应坡度调整法和自适应偏移量调整法的Hopfield神经网络模型(AHNN)。结合传统Hopfeild网络优点,将Sigmoid函数(S型函数)中的恒定参数变为随动,自动调整随动参数运动方向,减小了能量函数的振荡和迭代步数,从而可有效的解决高维、非凸、非线性约束条件的优化问题。将该方法应用于电力系统经济负荷分配(ELD),通过多个算例仿真表明,AHNN有效可行。
Based on principle of the traditional Hopfield neural network, an adaptive Hopfield neural network model with adaptive slope - adjusting method and adaptive bias - adjusting method has been put forward. The said model, combining with some advantages of traditional Hopfield neural network, can convert the invariables in Sigmoid function into adjustable parameters. The movement direction of said adjustabele parameters can automatically be adjusted, thus, the model can decrease oscilation in energy function and iterative steps, thereby, many multi- dimensional, non convex, and non- linear constrained optimization problems can be effectively solved. This method can be used for economic load dispatch in electric power system. Simulation results of several examples show that the said model to be effective and feasible.
出处
《热力发电》
CAS
北大核心
2007年第8期35-39,43,共6页
Thermal Power Generation
作者简介
周明（1981-），男，就职于中国船舶重工集团第709研究所，主要从事计算机控制、系统辨识电力系统规划的研究。E-mail：elkahlil@gmail．com
