Document Type : Research Paper
Authors
1 Department of Economics, Ar.C., Islamic Azad University, Arak, Iran.
2 Department of Computer, Ar.C., Islamic Azad University, Arak, Iran.
Abstract
Load forecasting is crucial for planning power plants, meeting electricity demand, and building reliable and efficient energy infrastructure. It is divided into three types: long-term, medium-term, and short-term. Short-term load forecasting (STLF) is particularly important due to the use of distributed energy resources, integration of renewable energy sources, and demand management in smart grids. Various models, such as artificial intelligence, traditional, or hybrid models, are used for this purpose. Accuracy in short-term forecasting is vital for efficient management of electrical systems. Factors like weather, holidays, and temperature affect forecasts, but using large amounts of data can be time-consuming and costly. Therefore, this study uses a simpler model that meets practical needs. Two new models for short-term load forecasting are proposed: one is a hybrid model based on the Mackey-Glass time series with fuzzy logic, and the other is a predictive fuzzy logic model. Both are trained using the Wild Horse Optimization (WHO) algorithm. These models were tested with multi-year data from Iran’s Markazi Province. The models’ performance was evaluated using metrics like Mean Absolute Percentage Error (MAPE) and Mean Square Error (MSE) and compared with three other methods (LSTM, ARIMA, and ANFIS). Results show that the proposed model significantly reduces MAPE and MSE compared to other methods, offering better performance.
Keywords
- Short-term load forecasting (STLF)
- Demand response
- Smart grids
- Mackey-Glass time series
- wild horse optimization (WHO) algorithm
Main Subjects
)