基于Elman神经网络模型的短期电力负荷预测模型

Short term power load forecasting model based on the Elman neural network model

在国家经济发展的过程中,电力行业是不可或缺的支柱,并且与国家的安全稳定、人民的安居乐业息息相关。为了提高电力负荷预测系统的预测精度,这次研究结合电力系统的动态特性,将动态型Elman神经网络算法应用于短期电力负荷预测模型的研究,从激励函数、学习规则两个角度进行了算法优化,并与目前研究较多的BP神经网络算法进行了MATLAB仿真实验对比。研究结果显示,在短期电力负荷的预测上,BP神经网络和优化后的Elman神经网络平均预测精度分别是97.75%和98.74%,优化后的Elman神经网络模型预测效果明显更优。Elman神经网络算法的优化思路,可以为电力负荷预测系统的研究提供一些参考和数据支撑,并能在实际应用中为提高电力负荷预测精度提供一些帮助。

In the process of the country’s economic development,power industry is indispensable.And it is closely bound to country’s safety and stability,and people’s living and working in peace and contentment.In order to improve the prediction accuracy of the power load forecasting system,this study combined the dynamic property of power system,applied dynamic Elman neural network algorithm to short term power load forecasting model and optimized the algorithm from the excitation function and learning rules’point.MATLAB simulation experiments were carried to compare with current well-studied BP neural network algorithm.The result revealed that the average prediction accuracy of BP neural network algorithm and optimized Elman neural network algorithm were 97.75%and 98.74%respectively in short term power load forecasting.It is obviously that optimized Elman neural network algorithm is better.The thinking of optimizing Elman neural network algorithm can provide some reference and data supporting for power load forecasting system,and can provide some help to improve the prediction accuracy of power load in real life.