广东石油化工学院学报
廣東石油化工學院學報
엄동석유화공학원학보
JOURNAL OF MAOMING COLLEGE
2011年
6期
47-50
,共4页
电力负荷短期预测%BP网络%L—M算法%粒子群算法
電力負荷短期預測%BP網絡%L—M算法%粒子群算法
전력부하단기예측%BP망락%L—M산법%입자군산법
short- term power load forecasting%BP network%L- M algorithm%Partial Swarm Optimization Algorithm
针对常规BP网络收敛速度慢,易陷入局部极小值等问题,采用L—M算法对网络进行训练,利用改进粒子群算法优化BP网络初始权值和阈值。将该方法应用在南方某市短期电网负荷预测中,预测结果表明,相较于常规BP网络、L—M算法改进预测模型,该预测算法在预测结果精度和速度上均有较大幅度提高。
針對常規BP網絡收斂速度慢,易陷入跼部極小值等問題,採用L—M算法對網絡進行訓練,利用改進粒子群算法優化BP網絡初始權值和閾值。將該方法應用在南方某市短期電網負荷預測中,預測結果錶明,相較于常規BP網絡、L—M算法改進預測模型,該預測算法在預測結果精度和速度上均有較大幅度提高。
침대상규BP망락수렴속도만,역함입국부겁소치등문제,채용L—M산법대망락진행훈련,이용개진입자군산법우화BP망락초시권치화역치。장해방법응용재남방모시단기전망부하예측중,예측결과표명,상교우상규BP망락、L—M산법개진예측모형,해예측산법재예측결과정도화속도상균유교대폭도제고。
As conventional BP network for the slow convergence and easy to fall into local minimum problem, the use of LM algorithm for network training, the improved particle swarm optimization BP network initial weights and threshold. Application of this method in the text grid in a city short - term load forecasting, showed that compared with conventional BP network, L- M algorithm to improve prediction models. This article describes the results of the prediction algorithm in the prediction accuracy and speed have increased greatly.