基于快速人工神经网络的短期负荷预测

Short-term load forecasting using artificial neural network and adaptive fuzzy system

根据负荷的不确定性和非线性的特点 ,采用了ANN和AFS理论进行STLF ,分两个步骤 :在ANN中引入了平滑因子和遗忘因子 ,来加快收敛速度并解决ANN的遗忘问题 ;在AFS中对基本负荷预测值进行修正 ,引进不平均的隶属函数来体现负荷变化对温度的敏感性。实践表明该模型具有速度快。

This paper proposes a hybrid model for short term load forecast using ANN and AFS,according to the uncertainty and nonlinearity of the load.The forecasted load is obtained through two steps.In the first procedure,this paper introduces the smoothing factor β and the forgetting factor γ.In the second procedure,the uneven membership function is used.The result shows the efficiency and accuracy of this hybrid model.