INTERNATIONAL JOURNAL OF INDUSTRIAL ENGINEERING ...

1 INTERNATIONAL JOURNAL of INDUSTRIAL ENGINEERING Research and Development (IJIERD), ISSN 0976 6979(Print), ISSN 0976 6987(Online), Volume 5, Issue 3, May- June (2014), pp. 13-23 IAEME 13 AN INTELLIGENT HYBRID MULTI CRITERIA DECISION MAKING TECHNIQUE TO SOLVE A PLANT LAYOUT PROBLEM Indranil Ghosh Calcutta Business School, West Bengal, India ABSTRACT Multi criteria decision making (MCDM) techniques in today s organizations, as a key to performance measurement comes more to the foreground with the advancement in the high technology. During recent years, many studies have been conducted to obtain a ranking among many alternatives via measuring performance of each of them against many criteria. Managerial decision making problems like supplier selection, weapon selection, project selection, site selection etc are dealt with many multi criteria decision making methods like TOPSIS, AHP-TOPSIS (Technique for Order Preference by Similarity to Ideal Solution), PROMETHEE (Preference Ranking Organization Method for Enrichment Evaluation), ELECTRE, VIKOR etc in crisp throughout the literature.

2 In this work, we first compare several MCDM methodologies to validate the consistency of them on a standard dataset of plant layout problem. We proposed M-TOPSIS, A-TOPSIS procedure to select a suitable layout for the comparative study. Results of M-TOPSIS and A-TOPSIS have been employed to build an unsupervised artificial neural network (ANN) to obtain a new ranking of alternatives. This study proposes an approach of deriving the rank value, in order to get optimal configuration, from the average of more than one set of rank results obtained through the deployment of MCDM methodologies. Keywords: TOPSIS, M-TOPIS, VIKOR, Crisp, ANN. 1. INTRODUCTION Due to ever increasing complexity of performance measurements which is one of the most important processes in management literature and as its measurement is critical for judging the success or failure of a firm, multi criteria decision making (MCDM) techniques INTERNATIONAL JOURNAL OF INDUSTRIAL ENGINEERING RESEARCH AND DEVELOPMENT (IJIERD) ISSN 0976 6979 (Print) ISSN 0976 6987 (Online) Volume 5, Issue 3, May - June (2014), pp.

3 13-23 IAEME: JOURNAL Impact Factor (2014): (Calculated by GISI) IJIERD I A E M E INTERNATIONAL JOURNAL of INDUSTRIAL ENGINEERING Research and Development (IJIERD), ISSN 0976 6979(Print), ISSN 0976 6987(Online), Volume 5, Issue 3, May- June (2014), pp. 13-23 IAEME 14 have recently been in the limelight of research. MCDM techniques are tailor made to cater a systematic and deterministic approach to tackle complex real world decision making problems composed of several intertwining and incommensurate criteria. Roy (1990)[1] argues that solving MCDM problems and searching for an optimal solution are clearly two distinct measures, prime focus of MCDM is to assist Decision makers (DMs) evaluate the complex judgments and to carefully analyze data involved in their problems and advance towards an acceptable solution.

4 The entire process is subdivided in three parts, a set of alternatives, A, is evaluated to produce a final decision result: Choice- Choose the best alternative from A. Sorting- Sort the alternatives of A into relatively homogeneous groups in a preference order. Ranking- Rank the alternatives of A from best to worst. Unlike many off-the-shelf recipes that can be applied to every problem regardless of their constraints MCDM techniques have often beendictated by the essence of real-life MCDM techniques like TOPSIS, AHP, combined AHP-TOPSIS [2], VIKOR [3], PROMETHEE [4], ELECTRE [5] etc. have been successfully applied by many researchers addressing many MCDM problems. Artificial neural network (ANN), an evolutionary optimization based algorithm had been developed in [6, 7], and [8].

5 ANN based algorithms are claimed to be helpful for practical INDUSTRIAL applications especially for dynamic situations. ANN is categorized in two sections- Supervised ANN & Unsupervised ANN which we discuss in section 4. ANN has been successfully applied in many real life INDUSTRIAL problems including MCDM problems too [9, 10]. One famous work of Kumar & Roy [11] deploys an Unsupervised artificial neural network to evaluate rank of suppliers. This work avails the model of to rank the layouts based on the results of M-TOPSIS & A-TOPSIS. The remainder of the paper is organized is as follows section 2 outlines the plant layout problem, section 3 depicts the mathematical steps involved with TOPSIS, A-TOPSIS, M-TOPSIS respectively, section 4 presents the unsupervised ANN model and algorithm to generate composite ranking, section 5 presents the comparative analysis of results and proposed methods and results of Yang & Hung [12] an approach of deriving the rank value, in order to get optimal configuration.

6 2. PLANT LAYOUT Designing and implementation of plant or facility layout is the most critical phase of setting up new facility in existing unit both in manufacturing and service sectors. It directly affects the performance of an entire unit. Layout design can influence quality of manufactured products or service delivery as checking or testing locations needs to be incorporated in the integrated system in most befitting manner besides the fact that in certain situations material damages are obviated by reducing its handling requirement. So choosing an appropriate layout among several layout configurations that can be generated by software such as ARENA, CORLAP, CRAFT etc is indeed a typical MCDM problem which contains several conflicting criteria associated with possible alternatives (plant configurations).

7 A good layout design ensures increase in productivity reducing overheads. Some notable works on this domain include Karray et al[13] where he proposed an integrated methodology using INTERNATIONAL JOURNAL of INDUSTRIAL ENGINEERING Research and Development (IJIERD), ISSN 0976 6979(Print), ISSN 0976 6987(Online), Volume 5, Issue 3, May- June (2014), pp. 13-23 IAEME 15 the fuzzy set theory and genetic algorithms to investigate the layout of temporary facilities in relation to the planned buildings in a construction site, (TOPSIS) and fuzzy TOPSIS[12, 14](Yang and Hung ,2007, Grey Relational Analysis(Kuo, Yang, and Huang, 2008). Yang and Hung [12] mentioned six criteria out of which three are quantitative and rest are qualitative.)

8 Thequantitative criteria included material handling distance(in meters ), adjacency score and shape ratio, which are thedirect outputs of Spiral. The handling distance is calculated by the sum of the products of flow volume and rectilinear distance between the centroids of two departments. The adjacency score is the sum of all positive relationships between adjacent departments. Whereas, shape ratio is defined as the maximum of the depth-to width and width-to-depth ratio of the smallest rectangle that completely encloses the department. For a layout design problem, it is needed to minimize both the shape ratio and flow distance, while maximizing adjacency score. There are three qualitative attributes are flexibility, accessibilityand maintenance.

9 These are the six attributes chosen by Yang and Hung to evaluate their 18 alternatives. 3. MCDM METHODOLOGIES TOPSIS: The TOPSIS (technique for order performance by similarity to ideal solution) method [15](Hwang & Yoon, 1981) constitutes a usefultechnique in solving ranking problems. The basic idea of the TOPSISis simple and intuitive: measure alternatives distances to predefinedideal and anti-ideal points first and, then, aggregate theseparate distance information to reach overall evaluation features of TOPSIS, as summarized in [16] (Kim, Park, and Yoon(1997)) and [17] (Shih, Shyur, and Lee (2007), include clear and easilyunderstandable geometric meaning, simultaneously considerationfrom both best and worst points of view, and convenient calculationand implementation.)

10 The procedural steps of TOPSIS are mentioned below: Construct a matrix based on the priority scoresassigned to each alternative simulator on each attributedenoted by X = (xij)nxm (1) Determine the importance weight (wj) of the attributes such that: = 1, j=1, 2, 3,..m. (2) Obtain the normalized decision matrix: = / ( ) j = 1, 2,..m; i = 1, 2, ..n. (3) Obtain the weighted normalized decision matrix, = ; j = 1, 2, .., m; i = 1, 2, .., n. (4) Determine the PIS and NIS: = ( , , .. , ) = {( { }| j B ), ( { | j C)} , (5) INTERNATIONAL JOURNAL of INDUSTRIAL ENGINEERING Research and Development (IJIERD), ISSN 0976 6979(Print), ISSN 0976 6987(Online), Volume 5, Issue 3, May- June (2014), pp.}