Comparison of Artificial Intelligence based Techniques for Short Term Load Forecasting

Abstract

Electrical load forecasting plays an important role in planning, operation and control of power system. The accuracy of this forecasted value is necessary for economically efficient operation and also for effective control. Accurate load and price forecasting are very essential in power system planning. This will increase the efficiency of electricity generation and distribution while maintaining sufficient security of operation. Due to deregulation in an energy sector and the energy market, there is a pressing need of accurate STLF method. Short-term load forecasting is an important basis of the secure and economic operation in power systems. Accurate load forecasting is helpful to improve the security and economic effect of power systems and can reduce the cost of generation. Therefore, finding an appropriate load forecasting method to improve accuracy of forecasting has important application value. Accuracy of STLF depends on forecasting techniques and input variable. There are many techniques for short term load forecasting such as mathematical and soft computing. Mathematical technique includes time series, expert system, data mining and regression method etc. Soft computing includes artificial neural network, fuzzy logic, fuzzy neural network and genetic algorithm. This dissertation work presents an investigation for the short term (up 24 hours) load forecasting of the load demand of Power utility, by using artificial neural network, fuzzy logic and fuzzy neural network. A comparative study of these methods has been carried out with regression method. These methods give better and accurate results than mathematical model for short term load forecasting. The past data of UK based power utility is used, in which independent variable such as dry bulb, wet bulb temperature, previous load, energy PR and ten minute spinning reserve (TMSR) have been used to explore the applicability of the proposed method.