Time-series load modelling and load forecasting using neuro-fuzzy techniques

This paper provides an intelligence method for medium and long-term energy demand forecasting of a complicated electrical system. A neurofuzzy methodology for load prediction is introduced that exploits the power of artificial neural networks (ANN) and fuzzy logic. In this method, energy data of several past years is used to train an Adaptive Network based on Fuzzy Inference System (ANFIS). ANN structure of ANFIS can capture the power consumption patterns, while the fuzzy logic structure of ANFIS performs signal trend identification. We tried to develop an intelligence approach for load forecasting, which is more accurate and it avoids the use of unavailable information. It is based on neurofuzzy statistical regression. We use this intelligence network to estimate the time series behaviour of energy consumption for several future years. Several simulations have been arranged to examine the efficiency of method.