Parallel Genetic Algorithm for solving Job-Shop Scheduling Problem Using Topological sort

Author

Somani, Archit ; Singh, Dhirendra Pratap

Author_Institution

Dept. of Comput. Sci. & Eng., Maulana Azad Nat. Inst. of Technol., Bhopal, India

fYear

2014

fDate

1-2 Aug. 2014

Firstpage

1

Lastpage

8

Abstract

Job Shop Scheduling Problem (JSSP) is one of the hard combinatorial optimization problems for machines having multiple jobs and multiple processing elements. This paper presents Parallel Genetic Algorithm (PGA) by Using Topological sorting, which is able to improve the solution of JSSP. Our proposed algorithm minimizes the execution time for Make span calculation by using PGA. Proposed PGA applies parallel topological sort on initial populations to generate linear sequences. After that whatever acyclic schedules come it applies crossover mutation on that and finally calculates the minimum Make span parallely for each linear sequence in optimal time. We implemented the PGA on GPU based machine using CUDA 5.0. Results are evaluated for different combinations of JSSP. Experimental results are compared with sequential genetic algorithm for JSSP.

Keywords

genetic algorithms; graphics processing units; job shop scheduling; parallel algorithms; parallel architectures; CUDA 5.0; GPU; JSSP; PGA; combinatorial optimization problems; job-shop scheduling problem; linear sequences; multiple processing elements; parallel genetic algorithm; parallel topological sorting; Algorithm design and analysis; Biological cells; Electronics packaging; Genetics; Schedules; Sociology; Statistics; Crossover and Mutation; JSSP; Make span; PGA; Topological sort;

fLanguage

English

Publisher

ieee

Conference_Titel

Advances in Engineering and Technology Research (ICAETR), 2014 International Conference on

Conference_Location

Unnao

ISSN

2347-9337

Type

conf

DOI

10.1109/ICAETR.2014.7012818

Filename

7012818