Greedy-Based Non-Dominated Sorting Genetic Algorithm III for Optimizing Single-Machine Scheduling Problem with Interfering Jobs

Abstract

Given the importance of production planning and control in the design of flexible services and manufacturing systems, scheduling problems with interfering jobs are much-needed optimization tools to respond to heterogeneous and fluctuating market demands in a timely fashion. This study contributes to the scheduling literature developing an effective multi-objective (M-O) metaheuristic to solve the Single-machine Scheduling Problems with Interfering Jobs (SSP-IJs). Integrating a local search-based mechanism into the evolutionary search procedure, a Greedy-based non-dominated sorting genetic algorithm III (GNSGA-III) is proposed that effectively explores multi-objective solution environments. Various performance indicators within extensive numerical tests are used to compare the performance of the GNSGA-III with that of the best-performing benchmark algorithm in the literature developed to solve the SSP-IJs. Statistical tests verify that the developed multi-objective optimization algorithm is superior with respect to various performance indicators. Applications of the developed solution approach are worthwhile topics to help advance multi-objective optimization problems.