Utilizing the Modified Self-Adaptive Differential Evolution Algorithm in Dynamic Cellular Manufacturing System

Utilizing the Modified Self-Adaptive Differential Evolution Algorithm in Dynamic Cellular Manufacturing System

Mohammad Hassannezhad (Polytechnic University of Turin, Italy) and Nikbakhsh Javadian (Mazandaran University of Science and Technology, Iran)
Copyright: © 2012 |Pages: 17
DOI: 10.4018/jamc.2012040101
OnDemand PDF Download:
$30.00
List Price: $37.50

Abstract

Today, Cellular Manufacturing Systems (CMS) have been introduced as a mixture of work-shop manufacturing and line-production systems for keeping efficiency and flexibility synchronously. One of the difficult steps of designing CMS is the Cell Formation (CF) problem in which parts with similar processes are made in one cell. Solving a dynamic integer model of CF with three sub-objective functions is considered using evolutionary algorithms. Due to the fact that CF is a NP-hard problem, solving the model using classical optimization methods needs long computational time. In this paper, a nonlinear integer model of CF is presented and then solved by proposed Modified Self-adaptive Differential Evolution (MSDE) and Modified Genetic Algorithm (MGA) using a set of 25 test problems. The results are compared with the optimal solution, and the efficiency of MSDE algorithm is discussed.
Article Preview

1. Introduction

With increasing market pressure for shorter product life cycle and need of responding to environmental issues, the dynamic aspects of manufacturing play a major role in designing and optimizing of a manufacturing system. CMS is a production approach aimed at increasing production efficiency and flexibility by utilizing the process similarities of the parts in order to minimize the intracellular cost of parts. The success of CMS is rooted in their ability to reduce set-up times, reduced work-in-process inventories, reduced material handling costs, simplified scheduling, shorter lead-times, faster response to internal and external changes, better production control, etc. (Askin & Strada, 1999). Most of the current cell formation design methods have been developed for a single-period planning horizon that called static. They assume that problem data such as product mix and demand is constant for the entire planning horizon. Sometimes we face dynamic condition which input data in each period is different from the other periods but it is deterministic.

1.1. CF Problem

CF is the first major step in designing a CMS, which involves identification of machine cells and part families. Most of the CF studies have focused on independence of cells and a few number of them point out the inter-cell movements (Vakharia & Wemmerlov, 1990). In the last three decades of research, several areas have been considered in CF problem. For example, Chen (2001) introduced a model for integrated production planning in CMS, while Liggett (2000) utilized CF problem for facilities layout by considering its features in both stable and stochastic conditions. Non-traditional methods such as neural networks, Tabu Search (TS), GA and Simulated Annealing (SA) have also been used to search for near-optimal solutions of various CF problems. Lei and Wu (2005) minimized the CF problem using TS approach based on similarity coefficient in generalized group technology. Furthermore, Wu, Chu, Wang, and Yan (2006) focused on designing of CMS based on GA approach, while Wu, Chang, and Chung (2008) utilized the SA algorithm for the CF problem objective. Recently, some researches focused on CMS with regard to variable demands. For instance, Geonwook and Herman (2006) formed part families and designed machine cells by using of GA under demand changes and Tavakkoli-Moghaddam, Javadian, Javadi, and Safaei (2007) designed a facility layout problem in CMS with stochastic demands. Also, Safaei, Saidi-Mehrabad, and Jabal-Ameli (2008) presented an extended model of dynamic CMS and proposed a hybrid SA to solve it.

Complete Article List

Search this Journal:
Reset
Open Access Articles: Forthcoming
Volume 9: 4 Issues (2018): 1 Released, 3 Forthcoming
Volume 8: 4 Issues (2017)
Volume 7: 4 Issues (2016)
Volume 6: 4 Issues (2015)
Volume 5: 4 Issues (2014)
Volume 4: 4 Issues (2013)
Volume 3: 4 Issues (2012)
Volume 2: 4 Issues (2011)
Volume 1: 4 Issues (2010)
View Complete Journal Contents Listing