A Hybrid Genetic Algorithm for Multi-Objective Job Shop Scheduling in Manufacturing Systems

Abstract

Moreover, productivity and operational efficiency in a business drastically improves with specific JSS (Job Shop
Scheduling) system tailored for its unique functionalities. Traditional approaches to JSS are impractical because they
rarely solve multi-objective, intricate problems that require reasoning about many different objectives at the same
time. We introduce a Multi-Objective Job Shop Scheduling (MOJSS) problem using a tailored Hybrid Genetic
Algorithm (HGA) which, combine empirical work with locally optimal searches that accept worse solutions called
heuristics for augmenting GA (Genetic Algorithm) to optimize makespan problem, machine and job tardiness
utilization. Proposed HGA through better selection, crossover, mutation strategies and along with a local refinement
step increases speed and quality of results. Evaluations stem from benchmark instances with proven evidence of HGA
outperforming standard GA, other evolutionary methods, and enhancing convergence, diversity, and solution
optimality. Also, from a comparative perspective, makespan reduces while machine utilization increases proving
better performance HGA against GE. The effectiveness of hybridization in evolutionary algorithms shines through
complex scheduling with these findings proving further aid to the domain pedagogy that seeks robust, flexible, and
scalable solutions to the real-world challenges of scheduling problems with advanced efficiencies.