Metaheuristics for warehouse storage location assignment problems
© 2018 Inderscience Enterprises Ltd. This study addressed warehouse storage location assignment problems (SLAP) to minimize total travelling distances in an order-picking process. The problem was formulated and presented as a mixed integer programming model. The LINGO optimization solver was then us...
Saved in:
Main Authors: | , |
---|---|
Format: | Journal |
Published: |
2018
|
Subjects: | |
Online Access: | https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85047991018&origin=inward http://cmuir.cmu.ac.th/jspui/handle/6653943832/54911 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Institution: | Chiang Mai University |
Summary: | © 2018 Inderscience Enterprises Ltd. This study addressed warehouse storage location assignment problems (SLAP) to minimize total travelling distances in an order-picking process. The problem was formulated and presented as a mixed integer programming model. The LINGO optimization solver was then used to find solutions for a set of generated problems. The results showed that the LINGO optimization solver easily attained optimal solutions for small-sized problems; however, as the problem size increased, computational time increased rapidly. Eventually, when the problem size became very large, LINGO was unable to find solutions. Due to the competence limitations of the exact solution method, this research presented two effective metaheuristic approaches - Differential Evolution (DE) and Global Local and Near-Neighbor Particle Swarm Optimization (GLNPSO) - to solve SLAP. To illustrate the performance of the algorithms, the numerical results were evaluated and compared with a set of generated problems. The experimental results showed that, for small-sized problems, DE and GLNPSO found optimal solutions equal to those obtained from LINGO, with fast computing times. For medium-sized problems, DE and GLNPSO were not significantly different in terms of solution quality, but DE found solutions approximately twice as fast as GLNPSO. For large-sized problems, DE was significantly superior to GLNPSO in terms of both solution quality and computational times. The average DE solutions, obtained from five independent runs, were equal to or better than those obtained from GLNPSO in all large-sized instances. In addition, DE showed faster convergence behavior than GLNPSO, since it yielded better solutions while using a fewer number of function evaluations. |
---|