基于多层Hopfield神经网络的配电网网络规划

New distribution network planning algorithm based on multilayer Hopfield neural network

分析了配电网网络结构规划模型,在此基础上提出用Hopfield神经网络进行网络结构规划.针对城市电网辐射状运行的特点,提出多层Hopfield神经网络模型、对应的能量函数以及参数选择规律.多层Hopfield神经网络的每一层对应于一个负荷点的供电线路,能量函数的建立同时考虑到各层的状态.提出一种新的基于多层Hopfield神经网络的配电网网络结构的规划算法,该算法通过使能量函数降到最低值,可以求得配电网网络规划问题的最优或近似最优解.新算法无需对线路编码、无需对数据进行归一化处理,更加易于编程实现.实例计算表明该方法可行、有效.

This paper analysis the distribution network planning model, and introduces Hopfield neural network to plan distribution network. We present muhiplayer Hopfield neural network model for planning a radical distribution network. The energy function and the regular of choosing parameters are also given in this paper. Each layer of neural network decide a power line for one node, the energy function can optimize all power lines for all nodes simultaneously. A new algorithm based on the muhilayer Hopfield neural network is put forward. By minimizing the value of the energy function, the algorithm can obtain the optimal or nearly optimal solution for the topology of radical distribution network. The new algorithm doesn＇ t need coding of city streets and normalizing of data, so the program is easier to be realized. An example shows that the algorithm is practical.