Title: Programming tasks in business processes like a realistic hybrid flexible flow shop using genetic algorithms

Authors: Jaime Antero Arango-Marin

Addresses: Universidad Católica Luis Amigó, Centro Regional Manizales, Carrera 22 N° 67A, 49. Manizales, Colombia

Abstract: An adaptation of the job scheduling to the programming of business process tasks is made in a hybrid flexible flow shop environment. The problem is modelled considering realistic situations: sequence-dependent task change times, malleability of batch sizes, variable transfer batch, objective function of minimising average tardiness, unrelated parallel resources and more than two stages. To solve the problem, the proposed standard and modified genetic algorithms were presented. The results of the experimentation allow us to appreciate that both genetic algorithms achieve average tardiness values between 20% and 60% better than the dispatch rules with best performance of the modified genetic algorithm. The conclusions are that it is possible to schedule business process tasks as an industrial plant, that it is necessary to take account of the real environment requirements and that the best solution is reached when a smart technique adapted to the features of the problem is used.

Keywords: business process management; BPM; genetic algorithms; combinatorial optimisation; flow shop; scheduling.

DOI: 10.1504/IJPMB.2022.121590

International Journal of Process Management and Benchmarking, 2022 Vol.12 No.2, pp.131 - 146

Received: 13 Nov 2019
Accepted: 23 Dec 2019
Published online: 21 Mar 2022 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article