基于Hopfield神经网络的电力系统扩展规划新方法

A New Method of Power System Expansion Planning Based on Hopfield Network

提出了一种基于Ｈｏｐｆｉｅｌｄ神经网络模型的电力系统扩展规划新方法，方法考虑了规划问题的特点，提出以二值状态实现数字编码的方法来构成Ｈｏｐｆｉｅｌｄ网络的模型结构，可用来求解输出为整数的优化问题，能量函数中以架线投资和运行费用为目标函数，并考虑了等式和不等式约束，对ＩＥＥＥ６节：点系统的计算表明这一方法是可行且有效的。

A new method of power system expansion planning(PSEP)based on Hopfield neural net-work model is presented in this paper.According to the special features of PSEP problem,the methodis taken out to form a new structure of Hopfield model. The success of such model in finding good solu-tions to complex problems of integer output is found to be dependent on the number representationschemes used.The equality constraints and inequality constraints are considered in Lyapunov energyfunction,It has been tested on IEEE-6 nodes system and achieved a satisfactory result.The resultdemonstrate that the new method is possible, capable and effective。