Development of a Neural Network Algorithm for Estimating the Makespan in Jobshop Production Scheduling


Yıldız İ., Saygın A., Çolak S., Abut F.

TEHNICKI GLASNIK-TECHNICAL JOURNAL, vol.40, no.4, pp.1257-1264, 2023 (SCI-Expanded)

  • Publication Type: Article / Article
  • Volume: 40 Issue: 4
  • Publication Date: 2023
  • Doi Number: 10.17559/tv-20220818161430
  • Journal Name: TEHNICKI GLASNIK-TECHNICAL JOURNAL
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Emerging Sources Citation Index (ESCI)
  • Page Numbers: pp.1257-1264
  • Çukurova University Affiliated: Yes

Abstract

Since production scheduling is considered a short-term plan for future production planning, the advantages of effective scheduling and control and their contribution to the production process are numerous. Efficient use of resources improves productivity and ensures that customer orders are met on time. Even the simplest scheduling system has a complex solution structure. Long lead times also make it difficult to estimate the demand accurately. Therefore, it is important to solve scheduling problems effectively for such difficult-to-manage production processes. Job shop scheduling (JSS) problems are among the combinatorial problems in the NP-hard problems class. As constraints increase in such problems, the solution space starts to go to infinity, making it increasingly difficult to find the exact optimum solution. For this reason, metaheuristic algorithms have been used to solve such problems in recent years. This study aims to develop an artificial neural network (ANN)-based application to produce an optimal or near-optimal solution for JSS. Using the job shop type production data of Taillard comparison problems, the total processing time (i.e., makespan) has been calculated with the proposed ANN application. The results have been compared with the results of related studies in the literature, and the algorithm's efficiency has been evaluated in detail.