A parallel search genetic algorithm for the flexible job shop scheduling problem with regular machine halt time

EasyChair Preprint 1631

Abstract

This paper presents a more realistic flexible job shop scheduling problem, the flexible job shop scheduling problem with regular machine halt time (RMHT-FJSP). Different from machine breakdowns and maintenances, the regular machine halt time is deterministic, regular and frequent. The objective is to minimize the number of tardiness jobs (JTN), total tardiness time (TTT) and average machine idle time (AMIT). The objective of minimizing AMIT is to ensure operations on a machine close to each other, and it is only used as a non-important selection criterion in this paper. To enhance the optimization ability of genetic algorithm, several kinds of crossover operators and mutation operators are adopted simultaneously, and the buffer population to integrate old population and new individuals generated by these operators during evolution process is proposed, we called this algorithm as parallel search genetic algorithm (PSGA). Further, a fitness function designed by pre-experiment is studied. A computational experiment is made. Comparisons between FJSP and RMHT-FJSP are also represented.

Keyphrases: Flexible job shop, Genetic Algorithm, Tardiness, parallel search, regular machine halt time

@booklet{EasyChair:1631,
  author    = {Guangcan Yang and Hegen Xiong},
  title     = {A parallel search genetic algorithm for the flexible job shop scheduling problem with regular machine halt time},
  howpublished = {EasyChair Preprint 1631},
  year      = {EasyChair, 2019}}