摘要
为了提高短期电力负荷预测的精度,提出了一种改进的粒子群算法和BP神经网络相结合的预测模型。综合考虑天气、温度等因素的影响建立了短期电力负荷预测模型,利用改进的粒子群算法对模型的初始参数进行优化,之后采用LM学习算法对优化后的网络进行训练。仿真结果表明,该预测模型的预测精度优于BP神经网络和PSO-BP神经网络,克服了BP神经网络和粒子群优化方法的缺陷,改善了BP神经网络的泛化能力,为短期负荷预测提供了一种有效的方法。
In order to improve the accuracy of short-term power load forecasting, a forecasting model combined with modified particle swarm optimization and BP neural network was proposed. Considering the weather conditions, tem- perature and other factors, we establish the short-term power load forecasting model, optimize it by the modified parti- cle swarm optimization and then train the optimized network by LM learning algorithm. The simulation results show that the prediction accuracy of the model which overcome the defects of the BP neural network and PSO-BP neural net- work, and improve the generalization ability of BP neural network was better than BP neural network and PSO-BP neu- ral network. Therefore, the model can be used effectively to forecast the short term load.
出处
《电测与仪表》
北大核心
2013年第3期29-32,共4页
Electrical Measurement & Instrumentation
关键词
粒子群算法
神经网络
负荷预测
particle swarm optimization, neural network, load forecasting