Enhancing robustness: Multi-stage adaptive robust scheduling of oxygen systems in steel enterprises under demand uncertainty

In steel industries, oxygen demands are generally uncertain due to unpredictable disturbances in steel production, which seriously impacts system safety. Conventional static oxygen scheduling exhibits excessive conservativeness and limited robustness. Thus, a Multi-Stage Adaptive Robust Scheduling method for oxygen systems is proposed to adaptively address uncertain demands over time. To achieve this, an oxygen system scheduling model that integrates multiple load-adjustment modes of the air separation unit is first established to improve responsiveness to oxygen demands. Then, the scheduling time horizon is divided into multiple stages. For each stage, a data-driven uncertainty set is constructed to capture the most critical uncertainty distribution interval with low conservativeness. In addition, computationally efficient recourse decisions are designed with a few additional variables. These designs upgrade the deterministic oxygen scheduling model to a computationally tractable Multi-Stage Adaptive Robust Scheduling model that can dynamically adjust evaporators and emissions to address observed uncertainty promptly. The superiority of the proposed model is demonstrated using real data of an oxygen system. Compared to conventional oxygen scheduling, the adjustable robust scheduling improves the ability to handle uncertainty at least three-fold and saves cost, thereby enhancing system robustness and while reducing conservativeness. The designed recourse decisions greatly improve solution efficiency by two orders of magnitude, and the constructed data-driven uncertainty set also benefits robustness and cost optimality.