Data-driven distributionally robust optimization for PVC Plant Planning under uncertainty of energy consumption

This paper studies the planning optimization problem of PVC production process with calcium carbide energy consumption uncertainty. A data-driven Wasserstein distributionally robust optimization approach is proposed to deal with the energy consumption uncertainty in PVC production planning. Firstly, the kernel density estimation technique is used to extract the distribution information of uncertain data of carbide furnace single consumption. Then, a support set was established using flexible uncertainty sets, and an ambiguity set was established based on Wasserstein distance to establish a distributionally robust model for the PVC production process. Finally, a PVC industry case shows that the proposed method can effectively reduce the cost of energy consumption and improve the robustness of PVC production planning under the uncertain factors of energy consumption.