Two-stage distributionally robust integrated scheduling of oxygen distribution and steelmaking-continuous casting in steel enterprises
Liu Zhang,Kaitian Zhang,Zhong Zheng,Yi Chai,Xiaoyuan Lian,Kai Zhang,Zhaojun Xu,Sujun Chen
DOI: https://doi.org/10.1016/j.apenergy.2023.121788
IF: 11.2
2023-09-14
Applied Energy
Abstract:In the steel industry, the imbalance between fluctuating oxygen demand and stable supply generally results in excessive oxygen emissions and power waste. Independent optimal scheduling of oxygen distribution (OD) or steelmaking-continuous casting (SCC) has limited ability to handle this issue. To address this challenge, we propose an integrated scheduling model that allows for flexible interaction between OD and SCC. We first develop an SCC scheduling model that applies to machine sharing of duplex and conventional smelting, and derive an OD scheduling model. A critical demand interface is further designed to connect the scheduling models of SCC and OD, resulting in a mixed integer linear programming-based integrated scheduling model. Existing research overlooks the probability information of demand uncertainty posed by unpredictable interferences, which limits the ability to handle uncertain demands. To address this limitation, we embed demand probability distribution via a data-driven ambiguity set and upgrade the deterministic integrated scheduling model via two-stage distributionally robust optimization . Our model was tested with real data and resulted in a 15% reduction in total costs compared to independent scheduling of OD and SCC. Additionally, it effectively reduced uncertainty risk to 8.3% with low conservativeness, demonstrating superior performance compared to recent methods.
energy & fuels,engineering, chemical