Hybrid Micro Genetic Multi-Population Algorithm With Collective Communication for the Job Shop Scheduling Problem

Abstract

This paper presents a hybrid genetic algorithm with collective communication (HGACC) using distributed processing for the job shop scheduling problem. The genetic algorithm starts with a set of elite micro-populations created randomly, where the fitness of these individuals does not exceed a tuned upper bound in the makespan value. The computational processes distribute the micro-populations collectively. In the micro-populations, each individual's search for good solutions is directed toward the solution space of the fittest individual, identified by an approximation of genetic traits. In each generation of the genetic algorithm, the best individual from each micro-population migrates to another micro-population to maintain diversity in populations. Changes in the genetic sequence are applied to each individual by the simulated annealing algorithm (iterative mutation). In this paper, the results obtained show that the genetic algorithm achieves excellent results, as compared to other genetic algorithms. It is also better than other non-genetic meta heuristics or competes with them.

Keywords

• Genetic approximation
• hamming distance
• processes
• barrier
• landscape
• micro-population