Towards sustainable scheduling of unrelated parallel batch processors: A multiobjective approach with triple bottom line, classical and data-driven robust optimization

Abstract

Sustainable scheduling has become a critical aspect of modern industrial practices, requiring the integration of economic, environmental, and social dimensions. This research introduces an integrated problem for scheduling unrelated batch processors, aiming to optimize total cost, energy consumption, and social benefit through a multiobjective mixed-integer linear mathematical programming model. The study addresses the uncertainty in job processing time using robust optimization to enhance the model’s reliability. A three-stage solution methodology is proposed to solve the problem. First, robust optimization approaches are suggested to handle job processing time uncertainty. To this end, classical and kernel learning data-driven robust approaches are employed for uncertainty handling in cases of interval-bounded and distributional asymmetry uncertainties. Then, the global criterion multiobjective method is presented to resolve goal conflicts. To tackle the NP-hard complexity, three efficient multiobjective metaheuristic algorithms, non-dominated sorting genetic algorithm-II, multiobjective particle swarm optimization, and multiobjective grey wolf optimizer, are designed. The developed model and methodologies are extensively evaluated through numerical experiments. Results demonstrate the efficiency of the current framework against the literature model in solving the studied problem. Also, the robust models can properly handle the problem’s uncertainty regarding the assumptions of studied cases. The global criterion method’s performance is acceptable for small instances, while metaheuristics excel in solving larger problems. Based on the assumptions of the studied robust cases, the developed framework is investigated for two case studies from poultry production and glass ceramization real-world industries to illustrate its applicability further.