翻译文献原文.pdf

Innovative Applications of O.R. A case study of an integrated fuzzy methodology for green product development q Xiaojun Wang a, Hing Kai Chanb, Dong Lic aDepartment of Management, University of Bristol, Bristol BS8 1TN, UK bNorwich Business School, University of East Anglia, Norwich NR4 7TJ, UK cLiverpool Management School, University of Liverpool, UK a r t i c l ei n f o Article history: Received 20 March 2013 Accepted 4 August 2014 Available online 12 August 2014 Keywords: Multi criteria decision analysis Green product development Life cycle assessment Fuzzy Extent Analysis TOPSIS a b s t r a c t Green product development has become a key strategic consideration for many companies due to regu- latory requirements and the public awareness of environmental protection. Life cycle assessment (LCA) is a popular tool to measure the environmental impact of new product development. Nevertheless, it is often diffi cult to conduct a traditional LCA at the design phase due to uncertain and/or unknown data. This research adopts the concept of LCA and introduces a comprehensive method that integrates Fuzzy Extent Analysis and Fuzzy TOPSIS for the assessment of environmental performance with respect to dif- ferent product designs. Methodologically, it exhibits the superiority of the hierarchical structure and the easiness of TOPSIS implementation whilst capturing the vagueness of uncertainty. A case study concern- ing a consumer electronic product was presented, and data collected through a questionnaire survey were used for the design evaluation. The approach presented in this research is expected to help compa- nies decrease development lead time by screening out poor design options. ? 2014 Elsevier B.V. All rights reserved. 1. Introduction Due to increasing awareness of environmental issues, green product design (e.g., carbon reduction) has been a challenging new area of inquiry. There are many tools for green product design, some as simple as a checklist or the Materials, Energy, and Toxicity (MET) matrix. Among them, Life-Cycle Assessment (LCA) has gained noteworthy attention. LCA is a scientifi c model used to ana- lyse the environmental impacts of a product by taking its whole product life cycle, including material selection and production, manufacturing, usage, delivery, end-of-life treatment, and so on, into consideration (Hawkins, Hendrickson, Higgins, Matthews, Yung et al., 2012). Conducting an LCA can help design- ers understand the environmental impacts of a design by quantify- ing the secondary (i.e., undesired) outputs of the whole life cycle and then converting them into measureable impact items for analysis (Cerdan, Gazulla, Raugei, Martinez, Vallet et al., 2013). In summary, conducting an LCA is not an easy task. Therefore, there is a need to develop innovative approaches for, or to supplement, LCA. In this paper, a screening approach that can alleviate the shortcomings of LCA is proposed. To be precise, a hierarchical structure is employed to represent the life cycle of a /10.1016/j.ejor.2014.08.007 0377-2217/? 2014 Elsevier B.V. All rights reserved. q This extended paper is supported by the Euro Working Group on Decision Support Systems (EWG-DSS) / and has been originally presented in the Liverpool 2012 EWG-DSS Workshop at the University of Liverpool Management School. This event was organised by J.E. Hernndez, P. Zarat, F. Dargam, S. Liu, R. Ribeiro and B. Delibasic. Corresponding author. Address: Department of Management, University of Bristol, 8 Woodland Road, Bristol BS8 1TN, UK. Tel.: +44 (0)117 928 8608. E-mail address: xiaojun.wangbristol.ac.uk (X. Wang). European Journal of Operational Research 241 (2015) 212223 Contents lists available at ScienceDirect European Journal of Operational Research journal homepage: /locate/ejor product design in order to break down the complex problem into such a hierarchy. Then, fuzzy logic is used to take uncertainty into consideration as a screening tool. A hybrid, two-step approach is adopted (details to be discussed in Section 3). The proposed approach can be used as a screening tool to reduce the number of eco-design options and to identify key improvement areas. It is particularly useful in the early stages of design when different options can be evaluated and screened out. The rest of the paper is organised as follows. Section 2 briefl y reviews the two methods employed in this study, followed by the descriptions of the model in Section 3. Then, Section 4 presents how the method can be applied in a real-life case study in selecting eco-design options. Numerical examples are provided in this section. The fi ndings are then discussed in Section 5, and Section 6 concludes this paper. 2. Literature review In recent years, there has been an increasing awareness of envi- ronmentally conscious practices (Carter Rao Sarkis, 1998; Yung et al., 2011). These practices include envi- ronmentally friendly design (sometimes referred to as eco-design), green procurement, sustainable operations, and a number of end- of-life practices such as recycling and remanufacturing. Environ- mental awareness may be a consequence of regulatory pressures to protect the environment. For example, the European Councils directive (2009) on energy related products (ErPs) requires manu- facturers to comply with eco-design principles in order to sell their products to the European Union. Preventive rather than corrective actions should be taken as early as possible during the design phase of ErPs in order to identify and reduce environmental impact of the products whole life cycle. This practice is becoming an important element in new product development. Decisions regard- ing raw materials selection, electricity consumption during use phase, packaging design, end-of-life treatment, etc. can potentially have a profound environmental impact. Adding eco-design princi- ples to the design process may further burden organisations. On the other hand, however, it also helps to boost the progression of organisations to reduce adverse effects on the environment (Zhu thus, the ErP Directive was created partly to address this issue (Yung et al., 2011). This is the motiva- tion for this studys proposal of an LCA-based fuzzy methodology for green product development. LCA is a systematic and scientifi c tool that can help designers analyse the environmental impact of a product and has been applied in various applications over the last three decades (Guine et al., 2011). In an LCA, a products whole life cycle is taken into consideration (Junnila, 2008). This means that LCA can provide the designers a complete view of the environmental output and, hence, the impacts of the product. Because of this unique feature, LCA has attracted increasing attention from both researchers and practitioners, and numerous studies can be found in the literature (e.g., Bovea Kobayashi, 2005; Thoming Yung et al., 2012). In essence, LCA involves multiple life cycle phases and requires the assessment of different environmental aspects (European Council, 2009). It is not uncommon that decision-making problems involve multiple criteria. The problems are even more diffi cult to address if some of the criteria are qualitative in nature. Saaty (1978)developedawell-knownAnalyticHierarchyProcess (AHP), which can handle such Multi-Criteria Decision-Making (MCDM) problems. The basic idea is to represent such problems by a hierarchical structure with different criteria and sub-criteria. Then, pairwise comparisons among those criteria are performed so that the weightings of the criteria (or priority in some applica- tions) with respect to the problem can then be estimated. AHP is one of the widely used approaches to prioritise multiple factors that can affect decisions involving multiple judging criteria, and tradeoffs can always be found between different factors (Tan, 2005). Applications of AHP are numerous (Ho, 2008). Although the discrete scale of AHP has the advantage of simplic- ity and ease of use for pair-wise comparison of alternatives, it has often been criticised from several perspectives in the literature (Bana e Costa Belton Smith Lenzen, 2006). Another stream of research uses fuzzy logic, which can handle uncertain information, to mitigate this weakness (Zadeh, 1965). In this paper, the two approaches employed are Fuzzy Extent Analysis and Fuzzy Technique for Order Performance by Similarity to Ideal Solution (TOPSIS). The former was developed by Chang (1996) and the non-fuzzy version of the latter was intro- duced by Hwang and Yoon (1981). This section briefl y summarises the two approaches and provides reasons for why an integrated approach is needed. The fi rst application (or evolution) of fuzzy AHP is to replace deterministic values in the pairwise comparisons process with lin- guistic parameters (e.g., more important, very important, and so on), which are characterised by fuzzy membership functions (Van Laarhoven M 2 gi;.;M m gi, i = 1, 2, . , n, where all of the M j gi (j = 1, 2, . , m) are TFNs. The value of fuzzy synthetic extent with respect to the ith object is defi ned as: Si X m j1 Mjgi? X n i1 X m j1 Mjgi “#?1 1 and Pn i1 Pm j1M j gi hi?1 can be calculated as: X n i1 X m j1 Mjgi “#?1 1 Pn i1m3i ; 1 Pn i1m2i ; 1 Pn i1m1i ? 2 The degree of possibility of M1P M2 is defi ned as: VM1P M2 sup xPy minuM1x;uM2y ? 3 When a pair (x, y) exists, such that x P y and uM1x uM2y 1, then V(M1P M2) = 1. Because M1and M2are convex fuzzy numbers, VM1? M2 1 if m12? m22; VM1? M2 hgtM1 M2 lM1d;4 where d is the ordinate of the highest intersection point D between lM1andlM2(see Fig. 2). When M1= (m11, m12, m13) and M2= (m21, m22, m23), then the ordinate of D is computed by VM2? M1 hgtM1 M2 m11? m23 m22? m23 ? m12? m11 5 Table 1 Linguistic classifi cation of triangular fuzzy numbers. Rating level Linguistic valuesTriangular fuzzy numbers (TFN) 1Equal(1, 1, 1) 3Moderately more important(2, 3, 4) 5Fairly more important(4, 5, 6) 7Much more important(6, 7, 8) 9Absolutely more important(9, 9, 9) 2, 4, 6, 8Mid-point preference values lying between above values (1, 2, 3), (3, 4, 5), (5, 6, 7), (7, 8, 9) X. Wang et al./European Journal of Operational Research 241 (2015) 212223215 To compare M1and M2, the values of both V(M1P M2) and V(M2P M1) are required. The degree of possibility for a convex fuzzy number to be greater than k convex fuzzy numbers Mi (i = 1, 2, . , k) can be defi ned by VM ? M1;M2;?;Mk VM ? M1 and M ? M2 and ? and M P Mk? minVM ? Mi; i 1;2;.;k:6 IfdXi minVSiP Sk;7 For k = 1, 2, . n; ki, then the rating vector is given by W0 dX1;dX2;.;dXnT8 where Xi(i = 1, 2, . , n) are n different criteria. The normalized rat- ing vectors are: W RX1;RX2;.;RXnT9 where W is a non-fuzzy number that provides priority weights of an uncertainty criterion or sub-criterion over others. To verify the accuracy of the method, the consistency measure is performed to screen out inconsistency between responses. Because Mi is a triangular number, it has to be defuzzifi ed into a crisp number to compute the consistency ratio (CR). The Centre of Area (COA) approach is used here for defuzzifying Mi. TFN Mi= (mi1, mi2, mi3 ) can be defuzzifi ed into a crisp value by PMi mi3? mi1 mi2? mi1?=3 mi110 Therefore, the CR of each judgment can be calculated and checked to ensure that it is lower than or equal to 0.1. 3.3. Evaluating alternative designs with Fuzzy hierarchical TOPSIS To evaluate alternative product designs, fi ve fuzzy decision matrices, e De , are constructed with respect to fi ve environmental assessments. Assuming that there are l alternatives Ak(k = 1, 2, . , l) and n life cycle phases, each main life cycle phase has cicri- teria, where the total number of criteria is equal to Pn i1Ci: xkij. This represents the value of the jth criterion within ith life cycle phase of the kth alternative, which can be crisp data or appropriate lin- guistic variables and can be further represented by fuzzy numbers, e.g.,xkij akij;mkij;bkij. A hierarchical MCDM problem can be con- cisely expressed in a fuzzy decision matrix as: k 1;2;.;l;i 1;2;.;n;j 1;2;.;Ci11 wherexkijis the fuzzy evaluation score of alternative Akwith respect to criterion Cij. Ciis the number of criteria within the life cycle phase Li. e is the number of environmental assessments. In general, the criteria can be classifi ed into two categories: benefi t and cost. For the benefi t criterion, a higher value is better, while the opposite is true for the cost criterion. The data of the decision matrix e Decome from different sources. Therefore, it is necessary to normalize it in order to transform it into a dimension- less matrix, which allows for the comparison of various criteria. In this research, the normalized fuzzy decision matrix is denoted by e R, shown as: e R rkij ? l?m and k 1;2;.;l;i 1;2;.;n;j 1;2;.;Ci;m X n i1 Ci12 The normalization process can then be performed by the fol- lowing fuzzy operations: rkij akij u ij ; mkij u ij ; bkij u ij ? ;8ij;xijis a benefit criterion u? ij akij; u? ij mkij; u? ij bkij ? ;8ij;xijis a cost criterion 8 : 13 whereu ij andu? ij present the largest and the lowest value of each criterion, respectively. The weighted fuzzy normalized decision matrix is shown as: e Ve vkij?l?m;k 1;2;.;l;i 1;2;.;n;j 1;2;?;Ci;m X n i1 Ci14 wherevkijrkij? Wij. Here, Wij is the fi nal weight score for each criterion, which is the product of the main category weight score and the criterion weight score with respect to the corresponding main category as follows: Wij wLi? wCij wi? wi1 wi2 . . . wici 2 6 6 6 6 4 3 7 7 7 7 5 ;i 1;2;.;n15 where wLiand wCijdenote the weight score of the ith main life cycle and the criterion Cijrespectively. Both wLiand wCijare obtained through the Fuzzy Extent Analysis method discussed in Section 3.2. The results of Eq. (14) can be summarized as: 16 Subsequently, the fuzzy addition principle is used to aggregate the values within each main criterion as follows: v0ki X Cj j1 vkij; k 1;2;.;l;i 1;2;.;n17 The matrix e V is thus converted into the fi nal weighted normal- ized fuzzy decision matrix e V0, 318 Fig. 2. Membership functions of the set of importance ratings. 216X. Wang et al./European Journal of Operational Research 241 (2015) 212223 Again, the fuzzy addition principle is used to aggregate the val- ues of fi ve environmental assessments as follows: e Y X 5 e1 e V0e;e 1;2;.;519 420 The addition operation is important because the hierarchical structure can be refl ected only when aggregation of the weighted values within each main life cycle phase and fi ve environmental assessments are conducted. Now, let A+and A?denote the fuzzy positive ideal solution (FPIS) and fuzzy negative ideal solution (FNIS), respectively. According to the aggregated fuzzy-decision matrix, we have: Ay 1;?; y i ;?;y n ? A?y? 1;?; y? i;?; y? n ? 21 wherey i andy? i are the fuzzy numbers with the largest and the smallest generalized means, respectively. For each column i, the greatest generalized mean ofy i and the lowest generalized mean ofy? i can be obtained. Consequently, the FPIS (A+) and the FNIS (A?) are derived. Then, the distances (d+and d?) of each alternative from A+and A?can be calculated by the area compensation method as: d k X n i1 dvki;v i ?; k 1;2;.;l; i 1;2;.;n22 d? k X n i1 dvki;v? i ?; k 1;2;.;l; i 1;2;.;n23 deA;eB ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi 1 3 a1? b12 a2? b22 a3? b32 hi r 24 The relative closeness index is calculated as follows: /k d? k d k d? k 25 According to the index value, the alternative design options can be ranked from the most preferred to the least preferred feasible solutions. 4. Case study In this section, a case study on a manufacturing company is pre- sented to illustrate how the proposed fuzzy methodology can be applied to support decision making for environmentally friendly product design evaluation. The company produces wireless per- sonal electronic products and would like to incorporate environ- mental issues into their product design because any decisions made in this stage could have a profound environmental impact throughout its entire product life cycle. An LCA was initially con- ducted for one of their electronic products. Details of the LCA results were reported in Yung et al. (2011, 2012), and Chan et al. (2013). In these studies, LCA was also proven to be a time-consum- ing and tedious process. In this study, the authors make reference to this case to demonstrate how the proposed model can facilitate and simplify new product development from an eco-design per- spective, especially when an LCA has already been conducted. With reference to the case (Yung et al., 2011, 2012; Chan et al., 2013), the main life cycle phases are defi ned, and the key criteria under each phase are identifi ed, as illustrated in Table 2. Relevant data such as bill of material and manufacturing processes have to be collected to construct a hierarchical structure. After construct- ing the hierarchical model, it is essential to know how important one life cycle phase (or its associated criterion) is over another for eco-design purposes. In other words, decision makers have to determine the weights between main life cycle phases and the associated c