基于遗传算法的WNN在电力负荷短期预测中的应用

提出将小波神经网络和遗传算法相结合,用于电力系统短期负荷预测的新方法。具体是充分利用遗传算法的优越性,对小波神经网络的权值进行优化,然后利用优化得到的权值,对原始数据进行W N N训练。通过仿真,该种方法比传统利用神经网络进行负荷预测具有更高的精度。

Application of power load forecasting using Wavelet Neural Network Based on Genetic Algorithm

It is proposed a new method for using Short-term load forecasting in power system, it combined Wavelet Neural Networks and Genetic algorithm. Specifically take advantage of the superiority of genetic algorithm, optimize the weights of Wavelet Neural Network, and then use the optimize weights; train the original data using Wavelet Neural Network. Through simulation, this method has higher accuracy than traditional load forecasting using neural networks. This work is supported by Department of Education of Gansu Province