A NeuroFuzzy Approach to Load Forecasting in Electric Power Systems

Labib M. Arafeh, Roger S. Powell


Clustering and NeuroFuzzy techniques are proposed for short-term-load forecasting. Assessments of the proposed models for different training periods along with and without the proposed weights are carried out. Preliminary results are encouraging and suggest that these models are capable of achieving a high degree of accuracy. Introduction of weights is shown to improve prediction performance, making it possible to shorten the necessary training data from that corresponding to10 weeks down to that corresponding to 3 weeks only.


Short-Term Load Forecasting, Fuzzy Logic, NeuroFuzzy, Sub-Clustering, ANFIS, Modeling Techniques.

Full Text:



  • There are currently no refbacks.