A DWCVaR-Based Optimization Approach For Multi-Period Refinery Planning Incorporating Equipment Maintenance Under Uncertainty

Refinery production planning is the cornerstone of operational decision-making in refineries. However, current practices often optimize production planning independently, neglecting the influence of equipment maintenance planning on production planning and the coupling relationship between them. Furthermore, the pervasive issue of uncertainty poses significant challenges to decision-making processes. To address these issues, this paper proposes a collaborative optimization model that aims to optimize both production planning and equipment maintenance planning simultaneously, thereby maximizing economic returns under uncertainty. Specifically, this study utilizes the 1-norm and ∞-norm of uncertainty sets, applying the Data-driven Worst Conditional Value at Risk (DWCVaR) method to reformulate the uncertainty model and reduce decision-making risks. To accelerate the solution process, we propose a single-layer computational approach approximating the current two-layer iterative algorithm. Empirical validation demonstrates that the proposed methods not only achieve higher economic benefits but also accelerate the solution process without compromising solution accuracy.