Abstract

Job-shop scheduling is one of the most difficult production scheduling problems in industry. This paper proposes an adaptive neural network and local search hybrid approach for the job-shop scheduling problem. The adaptive neural network is constructed based on constraint satisfactions of job-shop scheduling and can adapt its structure and neuron connections during the solving process. The neural network is used to solve feasible schedules for the job-shop scheduling problem while the local search scheme aims to improve the performance by searching the neighbourhood of a given feasible schedule. The experimental study validates the proposed hybrid approach for job-shop scheduling regarding the quality of solutions and the computing speed.

Highlights

  • The job-shop scheduling problem (JSP) is one of the most difficult production scheduling problems

  • We propose a schedule relaxing technique to obtain a quite relaxed schedule from a schedule obtained by constraint satisfaction adaptive neural network (CSANN)-II

  • Given a feasible schedule obtained by CSANN-II, a relaxed schedule can be obtained in the follow steps: 1) Calculate the distance between the expected makespan and the makespan of the given schedule

Read more

Summary

INTRODUCTION

The job-shop scheduling problem (JSP) is one of the most difficult production scheduling problems It aims to allocate a number of machines over time to perform a set of jobs with certain constraint conditions in order to optimize certain criterion, e.g., minimizing the makespan. In CSANN-II, the resource constraint block is constructed adaptively from actual resource constraint satisfactions during the running, which is achieved by quick sorting the jobs on each machine according to their starting time and orderly pair two neighbouring jobs into resource constraint units. The experimental results show that CSANN-II with the local search scheme has good performance regarding the quality of solutions and the computing speed.

Description of the Job-Shop Scheduling Problem
Classification of Feasible Solutions for JSPs
Giffler and Thompson Heuristics for JSPs
Neurons of CSANN-II
Adaptive Connection Weights and Biases
Adaptive RC-Block Scheme
Combined Heuristic Algorithms for CSANN-II
Local Search Mechanism
Schedule Relaxing Technique
Framework of the Hybrid Approach for JSPs
Experimental Setting
Experimental Results and Analysis
CONCLUSIONS AND FUTURE WORK
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call