Acceptable cost-driven multivariate load forecasting for integrated coal mine energy systems

Abstract

Forecasting errors are inevitable in integrated energy systems with multivariate loads. To minimize costs, specialized forecasting models are essential. In this study, we propose a method for acceptable cost-driven multivariate load forecasting for the integrated coal mine energy system. By analyzing the relationship between dispatch costs and forecasting errors, the forecasting accuracy requirements for different types of loads are determined, based on which appropriate models for forecasting loads are selected. Firstly, the impact degrees of forecasting errors on dispatch costs for different kinds of loads are determined. Following that, the forecasting accuracy requirements for different types of loads within the acceptable costs are calculated by solving an optimization problem. Finally, the models for forecasting different types of loads are selected based on the forecasting accuracy requirements and the Bayesian information criterion. The proposed method is applied to an integrated coal mine energy system, and the experimental results show that the proposed method is capable of forecasting multivariate loads of the system within acceptable cost ranges.