基于人工神经网络与主分量分析的短期电力负荷预测方法

Short-term Load Forecasting Based on Neural Network and Principle Component Analysis

电力系统短期负荷预测是保证电力系统安全经济运行和实现电网科学管理及调度的重要依据,目前的电力系统短期负荷预测方法存在着一些不足。提出了基于人工神经网络与主分量分析的短期负荷预测方法,在试验中分别采用该方法和单一的人工神经网络对辽宁省某电网的短期负荷进行了预测,试验结果表明本文提出的方法与单一的人工神经网络预测法相比,不但减少了预测的时间,而且避免了过拟合现象,提高了预测精度。

Short-term load forecasting is power system in an important basis for guarantee a safe and economical operation and implementation of a scientific management and dispatch of the system.Against existing shortcomings,a short-term load forecasting method based on neural network and principle component analysis is proposed.In the test process,short-term forecasting of given Liaoning power system is done both by this method and by neutral network,the result shows that compared with neutral network forecasting,time period is shortened,over-fit is avoided and forecasting accuracy is enhanced with this method.