Key points are not available for this paper at this time.
Many scenarios in the process manufacturing industry, such as production in steel plants, can be modeled through hybrid flow shop scheduling problems. Continuous processing is a crucial constraint in this type of production, playing a vital role in maintaining the continuity of operations. Meanwhile, limited processing resources create threshold constraints that affect the conditions for starting production and lead to new bottlenecks. Therefore, we introduce and study the hybrid flow shop scheduling problem with continuous processing and resource threshold constraints (HFSSP-CPRT). To solve this problem, a population-based iterative greedy algorithm with bidirectional decoding (PBIGA-BD) is proposed. To improve encoding accuracy and decoding efficiency, stages with continuous processing and resource threshold constraints are encoded separately, and bidirectional decoding is applied. Subsequently, six initialization solutions are designed to improve the quality of the initial population. To enhance the diversity of the superior population, an improved destruction-construction method is designed. Then, a hyper-heuristic local search is developed to reduce the blindness of local search. It consists of six neighborhood search actions by considering the characteristics of HFSSP-CPRT. We tested the effectiveness of PBIGA-BD by applying it to an HFSSP-CPRT case of steel plant in China. Comprehensive experimental results showed that the proposed PBIGA-BD is able to outperform several heuristic rules and state-of-the-art algorithms being compared.
Su et al. (Thu,) studied this question.