Short-Term Power Load Interval Forecasting Based on Nonparametric Bootstrap Errors Sampling

Abstract

Short-term power load forecasting plays a vital role in the planning of distribution network and the development of social economy, and it is a very important task to forecast the power load accurately and reliably. Many researchers have devoted their attention to construct point forecasting models. However, due to errors in load forecasting are simply unavoidable and always are significant, point forecast methods are difficult to capture the characteristics of large fluctuation for power load. Therefore, the construction of interval forecasting is particularly important to improve the accuracy of power load forecasting. In this paper, the nonparametric Bootstrap sampling method is used to estimate the confidence interval of different interval errors and construct the interval forecasting to reduce the error of load forecasting. Sliding windows method is firstly used to pre-process original data, Extreme Learning Machine AdaBoost model is adopted to simulate points forecasting. Then, intervals are divided according to the actual values, and the errors in each interval are sampled by nonparametric Bootstrap method. Finally, the ultimate forecasting error is reduced by the confidence interval of the error. The proposed model is used to forecast power load for Queensland, Australia. The experimental results show that the proposed method performs better than other comparative models.