A Progressive Hedging-Based Solution Approach for Integrated Planning and Scheduling Problems under Demand Uncertainty

• Published in: Industrial & engineering chemistry research Vol. 58; no. 32; pp. 14880 - 14896
• Main Authors: Peng, Zedong, Zhang, Yi, Feng, Yiping, Rong, Gang, Su, Hongye
• Format: Journal Article
• Language: English
• Published: American Chemical Society 14.08.2019
• Subjects: algorithms; computers; deterministic models; ethylene; planning; process design; stochastic processes; uncertainty
• ISSN: 0888-5885; 1520-5045; 1520-5045
• DOI: 10.1021/acs.iecr.9b02620

Progressive hedging (PH) is a classical decomposition algorithm for solving multistage stochastic problems. However, due to the exponentially growing model size of real-world enterprise-wide optimization problems, critical issues arise when implementing PH in practice. In this work, we propose a novel PH-based algorithm to address integrated planning and scheduling problems under demand uncertainty in a general mathematical formulation. Strategies are proposed to accelerate and guarantee the convergence of the algorithm. Through application of the enhanced PH to solve variants of a typical state–task network example and a real-world ethylene plant case, computational results demonstrate that the proposed algorithm outperforms directly invoking commercial solvers and gets a better solution within nearly two-thirds of the direct solution time on a serial computer. The advantage of the multistage stochastic programming method is also demonstrated by comparing the model solution with the counterparts of an expected value-based deterministic model and a two-stage stochastic model.