遗传神经网络在电力系统短期负荷预测中的应用

为了克服传统BP神经网络中存在的一些缺陷，实现准确、快速预测电力系统负荷的目的，作者通过将遗传算法与神经网络结合，构造了一种遗传神经网络来进行电力系统短期负荷预测，方法的思路是：首先，利用遗传算法有指导地计算神经隐层节点数，从而确定一个较合理的神经网络结构；其次，由遗传算法从初始权值的解群中选取出一个优秀的初始权值，克服初始权值选取的盲目性；最后，将得到的神经网络结构和优秀的初始权值结合起来，利用改进的BP算法进行电力系统短期负荷预测，仿真计算表明该方法达到了提高预测精度和改善网络性能的要求。

APPLICATION OF GENETIC ALGORITHM NEURAL NETWORK FOR SHORT TERM LOAD FORECASTING OF POWER SYSTEM

To overcome the defects of BP neural network and to make short term load forecasting more accurate and fast, a kind of genetic algorithm neural network is established to forecast short term load of power system by combining genetic algorithm and neural network. For establishing this genetic algorithm neural network, a method consisting of three steps is applied. In the first step, the number of hiddennodes of this network is calculated by use of genetic algorithm. Thus, a more rational neural network structure is determined. In the second step, by use of genetic algorithm a fittest initial weight value is selected from the solution group of initial weight values to avoid the blindness in the selection of initial weight value. In the third step, combining the structure of the obtained neural network and the fittest initial weight value, the short term load forecasting of power system can be performed by use of improved BP algorithm. Simulation results indicate that this method can meet the need of improving forecast accuracy and enhancing the performance of the network.