Multi-Objective Flexible Job Shop Scheduling Using Genetic Algorithms

Flexible Job Shop Scheduling is an important problem in the fields of combinatorial optimization and production management. This research addresses multiobjective flexible job shop scheduling problem with the objective of simultaneous minimization of: (1) makespan, (2) workload of the most loaded machine, and (3) total workload. A general-purpose, domain independent genetic algorithm implemented in a spreadsheet environment is proposed for the flexible job shop. Spreadsheet functions are used to develop the shop model. Performance of the proposed algorithm is compared with heuristic algorithms already reported in the literature. Simulation experiments demonstrated that the proposed methodology can achieve solutions that are comparable to previous approaches in terms of solution quality and computational time. Flexible job shop models presented herein are easily customizable to cater for different objective functions without changing the basic genetic algorithm routine or the spreadsheet model. Experimental analysis demonstrates the robustness, simplicity, and general-purpose nature of the proposed approach.