An inventory-theory-based interval stochastic programming method and its application to Beijing’s electric-power system planning

In this study, an inventory-theory-based interval stochastic programming (IB-ISP) model is proposed through incorporating stochastic programming and interval parameters within an inventory model. IB-ISP can tackle uncertainties expressed as probability density functions (PDFs) and interval parameters in constraints and objective function. The developed IB-ISP is then applied to planning electric-power generation system of Beijing. Support vector regression (SVR) is used for forecasting the electricity demand, which is useful for coping with the uncertainty of customer demand. During the coal transportation processes, various factors may affect the time consumption of coal transportation, leading to uncertainties existing in energy generation and energy inventory. Under different delay times of coal transportation, different safety stocks and inventory patterns are generated to minimize the system cost and ensure the regular operation of the coal-fired power plants. The results obtained can not only help the managers to identify desired policies for safety stock in electricity-generation processes, but also be used for minimizing system cost and generating desired inventory pattern (with optimal transferring batch and period). Compared with the traditional economic order quantity (EOQ) model, the IB-ISP model can provide an effective measure for not-timely coal supplying pattern with a reduced system-failure risk under uncertainty.