Middle Anatolian Region Short-Term Load Forecasting Using Artificial Neural Networks

In recent years, several studies of short-term load forecasting using different of artificial neural network structures have been reported. In this paper, an application of short-term load forecasting is investigated by multilayer perceptron structure. Actual load and temperature data of the Middle Anatolian Region in the years 2002 and 2003 are used for this investigation. In this study, maximum temperature, minimum temperature, and day type factors are used to construct the forecasting model. Also, load forecasting for the same region is obtained by the regression method to compare the effectiveness of the artificial neural network method.

Keywords:

• short-term load forecasting
• artificial neural networks
• similarity based load forecasting

Acknowledgments

The authors are greatly indebted to TEDAS Research—Development Planning and External Relation Department Office for providing Turkey Middle Anatolian Region Electric Energy Distribution and Custom Statistics used in this paper.