A novel improved data-driven subspace algorithm for power load forecasting in iron and steel enterprise

Abstract

Electricity is one of the main energy in iron and steel enterprise, it is very important to forecast power load accuracy. Accurate power load demands estimation is an important way to reduce production cost, thus data-driven subspace (DDS) method is proposed to forecast power load. Considering the needs in the load forecast period of enterprises in the different sectors, the load forecasting systems are classified into daily load forecasting and ultra-short term load forecasting. The subspace method is improved by introducing the feedback factor and the forgetting factor. The values of these factors are optimized by particle swarm optimization (PSO) algorithm to improve the prediction accuracy. The performance of the improved method is verified by Bao steel's practical data. Forecasting results of the improved method can provide beneficial advice in power load management.