Fractional ARIMA with an improved cuckoo search optimization for the efficient Short-term power load forecasting

Abstract

Short-term power load forecasting plays a key role in power supply systems. Many methods have been used in short-term power load forecasting during the past years. A new short-term power load forecasting method is proposed in this study. First, the study represents a Fractional Auto-regressive Integrated Moving Average (FARIMA) model based on long-range dependence (LRD). The LRD model is governed by the Hurst exponent, which shows whether the model exhibits the LRD or not. Then, the study employs Cuckoo Search (CS) algorithm based on two parameters dynamic adjustment for parameter optimization of the forecasting model. As test problem, we use the real power consumption data, and test it for different forecasting models. Our results indicate that the FARIMA model and the improved optimization algorithm show relatively high accuracy and effectiveness in forecasting short-term power load.

Keywords

• Fractional Auto-regressive Integrated Moving Average
• Long-range dependence
• Hurst exponent
• Improved Cuckoo Search