1、 12 外文翻译 原文 Supplier evaluation and selection via an AHP Material Source: Int J Adv Manuf Technol (2006) Author: Wei-Ning Pi Chinyao Low Abstract The purchasing function directly affects the competitive ability of a firm. Purchasing managers need to periodically evaluate supplier performance in orde
2、r to retain those suppliers who meet their requirements. The importance of incorporating multiple attributes such as quality, on-time delivery, price and service, into vendor evaluation are well established in the literature. This report provides a more accurate and easier method for using an analyt
3、ical hierarchy process (AHP). An example for supplier evaluation and selection is also presented to demonstrate the functional application of the model. Keywords Analytical hierarchy process Supplier evaluation 1 Introduction Usually, quality is a critical concern for most manufacturers while purcha
4、sing. The need for high-quality suppliers has always been an important issue for many manufacturing organizations. The purchase price is also a highlighted consideration for the purchasing organization due to its impact on the product cost, but the purchase price is not all of the cost associated wi
5、th the material receipt. Additional costs are required by the purchasing organization to correct the deficiencies when a supplier fails to meet quality and delivery requirements. Hence, thepurchasing department must consider the full-part cost instead of a unit-price-oriented cost. Monczka and Trech
6、a provided a cost-based supplier performance evaluation system to evaluate key supplier performance. Service quality from the supplier is also very important to the manufacturer. Improving service quality is considered an essential strategy for success and survival in todays competitive situation. I
7、n order to meet the actual needs of customers, it is important to quantify service quality. Li proposed two methods to measure service quality. On the basis of the above discussion, it is not enough for the purchasing department to consider only one factor when purchasing a part. The purchaser may 1
8、2 try to find the optimal supplier not necessarily the supplier offering the best technical service, the lowest price or the shortest delivery. At this time, one of the important activities in the purchasing department is the evaluation and selection of a supplier from different multi-faceted inform
9、ation. Once an acceptable supplier is identified, the buyer has an opportunity to establish a long-term relationship with that supplier, which may provide a strategic advantage. In supplier evaluation processes, strategic importance has been addressed by several researchers; however, these studies h
10、ave mainly focused on the impact of supplier selection decisions on various functional areas. Financial measurement is also considered in the decision-making process for the traditional supplier evaluation, wherein manufacturing strategies such as just-in-time (JIT) delivery have placed increasing i
11、mportance on the incorporation of multiple supplier criteria. In particular, when manufacturers reduce their materials inventory, they increase their reliance on receiving the “right parts at the right time in the right condition” from their suppliers. In this study, a simple method for supplier eva
12、luation and selection based on quality, on-time delivery, price and service is developed. The model quantifies these four multiple criteria and then uses an analytical hierarchy process (AHP) to combine them into one global variable for decision-making. A numerical example is also presented to illus
13、trate the model and to demonstrate its utility. 2 Review of literature on supplier evaluation Methodologies for supplier evaluation have included conceptual, empirical and modeling approaches. Cost, quality and delivery performance are the three most important criteria that need to be considered for
14、 supplier evaluation in the initial work. The conceptual research primarily emphasizes the strategic importance of supplier evaluation and the trade-offs among cost, quality and delivery performance. The empirical research mainly focuses on studying the relative importance of various supplier attrib
15、utes such as price, quality and delivery performance. Although the conceptual and empirical research both stress the strategic importance of supplier evaluation and the consideration of multiple measures, they do not specifically propose any evaluation models. A lot of literature has accumulated on
16、evaluation models. Most of these models finalize the supplier selection decision making process on the basis of a set of supplier performance criteria. Some important models are summarized below. (1)Categorical model in the categorical model, suppliers are evaluated by criteria 12 such as cost, qual
17、ity, speed of delivery, etc. With regard to each criterion, suppliers are classified as good, fair, poor and assigned a (+), (0) or () for each level, respectively. The supplier receiving the most (+) ratings is considered the best. Being simplistic and easy to use, this model can include both quali
18、tative and quantitative criteria. However, all of the criteria are weighted equally, intuitively and subjectively by this model. In practice, some of the attributes are more important than others. Alternatively, this method can be useful if weights are assigned to each criterion and the (+), (0) and
19、 () are replaced by (+1), (0) and (1), respectively. On the basis of the total score, suppliers can then be ranked and the one with the highest score can be selected (2)Cost-ratio method Timmerman, believing that part-cost is not the total price of purchasing parts, proposed the cost-ratio method. T
20、his model collects all costs related to quality, delivery and service and evaluates them as a percentage of the total price. Then, the supplier who can provide the lowest cost is the best choice. However, it is difficult to develop cost-accounting systems for this purpose. (3)Cost-based models Moncz
21、ka and Trecha recognized that material price is only a fraction of the cost of the purchased material and that the measurement and evaluation of the suppliers performance should accurately reflect the total cost of doing business with that individual supplier. Hence, they provided a cost-based suppl
22、ier performance evaluation system to reflect the actual total cost of doing business. In this model, two indices, namely a supplier performance index (SPI), and a service factor rating (SFR) were used. The SPI recognizes costs attributed to non-performance by suppliers for delivery, material quality
23、 and price. These costs are identified and collected after which the total cost of the suppliers performance is used to develop an index number for each supplier for each major item. The SPI can be expressed as SPI = Extended purchase cost +Non -performance cost/Extended purchase cost The SFR shows
24、the performance factors that are difficult to quantify from the cost viewpoint, but which are important to the suppliers success. This evaluation system can indicate when a suppliers relative performance is exhibiting variability; also, it can provide insight into the suppliers ability to maintain i
25、tem performance consistency. There are several advantages in this evaluation model. First, it allows for qualitative and quantitative evaluation criteria. Second, evaluation on qualitative criteria is done by those who have direct contact with the suppliers. Third, the two indices are complementary
26、to each other and, if integrated properly, can make this 12 model superior to other available models. However, with this and other models, the process of evaluation is still subjective. (4)Weighted point model The weighted point method quantifies the factors with relevant weights and then rates the
27、potential suppliers according to these weighted factors. Thomson stated that this decision begins with the identification and weighting of key dimensions (evaluative or choice criteria) required for evaluating alternative vendors. The decision-makers next rate the expected performance of the supplie
28、rs by each evaluation criterion under intuitive judgment. The supplier performance ratings are multiplied by their respective importance weights to yield a weighted value. Then, the vendor with the highest summated score is the superior choice. Thompson pointed out that the mathematics underlying we
29、ighted point models are simple and can be adapted to any type of purchase decision. However, weighted point models also have some disadvantages, two of which are the subjectivity of the decision-maker in the identification of weights and the assumption of an ordinal scale as a cardinal scale. (5)Ven
30、dor profile analysis Thomson provided a modified weighted point model, called vendor profile analysis, to reduce the uncertainty innate to the rating mechanism. This model incorporates decision-makers understanding of uncertainty surrounding vendor performance by using a Monte Carlo simulation techn
31、ique instead of a rating from human intuitive judgment. The simulation algorithm randomly samples values from within each estimated performance range and combines these values with importance weights, in accordance with linear compensatory rules, to produce a distribution of summated scores. This pr
32、ocess can be repeated up to several thousand times for each supplier. The use of the Monte Carlo simulation simplifies the decision makers input to the evaluation model and provides output that contains considerably more information upon which to base purchase decisions than do standard weighted-poi
33、nt decision models. (6)Dimensional analysis In this model, the evaluation process involves a series of one-on one comparison and can compare only two suppliers each time. The dimensional analysis ratio (DA) can be greater than 1, equal to 1 or less than 1. Youssef pointed out that this evaluation me
34、thod has two disadvantages. First, a value of DA = 1 will cause the decision-maker to be indifferent about which supplier is chosen. Second, 12 the process becomes very tedious and time-consuming if a large number of suppliers must be evaluated. 3 Conclusions A supplier evaluation and selection syst
35、em via an AHP has been proposed in this study. The decision criteria are quality, on-time delivery, price and service. The performance on each criterion for each supplier has been transferred to quality loss by using Taguchi loss functions, and an AHP has been used as a framework to formalize the ev
36、aluation of tradeoffs between the conflicting selections criteria associated with various suppliers offers. One numerical example has also been used to demonstrate the functional application of the model. This evaluation program can address buying needs by monitoring and evaluating suppliers on thei
37、r actual performances. It communicates the purchasing priorities to the supplier in a manner that is easy to understand. In actual application, managers must carefully select the factors that best represent their competitive priorities, goals and objectives, and also construct pairwise comparison ma
38、trices in order to obtain more objective weights. Such procedures will make the application of the model more realistic. 译文 基于 层次分析法的供应商评价和选择 资料来源： Int J Adv Manuf Technol (2006) 作者： Wei-Ning Pi Chinyao Low 摘要：采购能力直接影响到一个公司的竞争能力。为了留住供应商，满足自己的需求，采购经理需要定期评估供应商业绩。供应商评价的多属性体系重要性比如质量、准时交货，价格和服务，在文献中能很好的构
39、建。这份报告提出了一个更为准确、更加容易的方法，用层次分析法 (AHP)将质量损失转移成决策变量。举例给出供应商评价和选择的功能来证明应用该模型。 关键词层 ：层次分析法，供应商评价 1 介绍 通常，对于大多数制造商采购而言质量是最值得关注的。供应商高品质的需求对许多制造组织一直是一个非常重要的问题。采购组织也要高度考虑到采购价格对其产品成本的影响，进货价格并不是所有成本与原材料的收据。当供应商质量和交货期需求不符合，采购组织需要额外费用来纠正存在的不足。因此，采购部门必须考虑全部或部分的成本而不是每一项价格的成本。 Monczka和 Trecha 提供了一种基于价格的供应商评估体系来评价关键
40、供应商的绩效。 12 对制造商来说供应商的服务质量也很重要。为了成功生存在今天的竞 争局面，提高服务质量被认为是一种基本的策略。为了满足客户的实际需求，量化服务质量是很重要的。李提出两种方法来衡量服务质量。 在上述讨论的基础上，当采购一个部分时已经不足以采购部门仅考虑其中的一个因素。买方可能试图找到最佳的供应商 不一定是供应商最佳的技术服务，最低的价格或最短的交付。这时，在采购部门其中一个重要的活动是从很多不同方面的信息来评估和选择供应商。一旦确认一个可接受的供应商，买方有机会与供应商建立长期的合作关系，可以提供一个战略优势。在供应商评估过程中，战略重要性由几位研究人员处理；然而，这些研 究主
41、要着眼于影响决定供应商选择的各种功能区。对传统供应商评价金融测量在决策过程也是要考虑的，其中制造战略提供多供应商标准就像生产计划 (JIT)将增加公司的重视。特别是当中国制造商降低材料库存，会增加他们依赖收到他们的供应商“在正确的情况下在正确的时间正确的零部件”。 在这项研究中，供应商建立在保证质量的基础上的一个简单的评价和选择方法，准确的交货，设计与开发价格和服务。该模型定量这四个多目标，然后使用层次分析法 (AHP)结合成一个全局变量决策，给出了一个算例说明该模型，从而显示它的用途。 2 供应商评估相关的文献 供应商评估的方法，包括理论、实证和建模。成本，质量和交付业绩是三种最重要衡量标准
42、，需要考虑供应商在最初工作中的评估。概念研究主要强调供应商评估战略的重要性以及权衡它们之间的成本，质量和交付业绩。实证研究主要重点研究的相对重要性各种供应商属性，如价格、质量和交货期的性能。虽然概念和实证研究在考虑供应商评价多种措施时起到了战略重要性，但他们没有明确提出任何评价模型。很多文学积累了评估模型。大部分的这些模型上确定了供应商选择的决策过程。 （ 1）分类模型在分类模型，供应商要进行评估，比如成本、 质量、收发速度等。对于每个标准，供应商分为好、公平、差，并标记为 (+)， (0)或 ()，为每一个水平。供应商收到最多 (+)的评分被认为是最好的。被过分单纯化的、易于使用，本模型可以
43、包括定性和定量标准。然而，此模型所有的标准加权同样、直观、主观。在实践中，一些属性比别的更加重要。另外，该方法可以是有用的如果重量被分到每准则和 (+)， (0)、 ()都分别换成了 (+ 1)， (0)、 (1)。根据总分，供应商可以被排序，可以选择最好的成绩。 （ 2） Timmerman 成本比例方法提出了成本比率方法，认为部分不是购买部 分总价格。该模型收集有关的一切费用、质量、交货和服务并评估他们的百分比总价。然后，供应商可以提供最低的成本是最好的选择。但是，这种成本12 核算系统很难开发。 （ 3）基于成本的模型 Monczka 和 Trecha 意识到材料价格只是采购材料成本的一
44、小部分并且供应商绩效的测量和评估要精确的反映与个体供应商交易时的总成本。因此，他们提供了一种基于价格的供应商绩效评价指标体系来反映了实际交易总成本。在该模型中两项指标，即供应商性能指标 (SPI)，服务因素评级 (SFR)被应用。SPI 认为成本是供应商未履行交付、材料质量和价 格而造成的。这些成本是在供应商绩效的总成本被用来开发一个指数为每个供应商的对应每个主要的项目之后被确定和收集。 SPI 可以被表示为： SPI=扩展采购成本 +违约成本 /扩展采购成本 SFR 表明绩效因素从成本的角度看是难以定量的，但是对供应商的成功是极其重要的。当一个供应商的相对性能发生变异性展示时该评价体系可以指
45、出，也可提供深入了解供应商的内部能力来维持项目绩效的一致性。这个评价模型有几个优势：首先，它允许定性和定量评价准则。第二，定性指标是由那些直接接触的供应商来评定的。第三，这两个指标互补，如果综合 得当，可使该模型优于其他现有的模式。不过，有了这些模型，评估的过程仍然是主观的。 （ 4）加权点模型 加权点法与相关因素权重，然后根据这些加权的因素平衡潜在供应商。汤姆森表示，这个决定开始鉴定的关键维度（评价或选择标准）评价所需选择供应商。在每个评价标准直觉判断下决策的预期业绩的供应商。供应商表现评级都乘以各自的重要性权重，这样会产生一个加权值。然后，卖方以最高的求和得分作为更好的选择。汤普森指出数学
46、潜在的加权点模型简单、能适应任何类型的购买决定。然而，加权点模型也有一些缺点，其中两决策的一个普通的假设主观权 重的识别规模被当成一个至关重要的规模。 （ 5）供应商个人能力测试图分析 汤姆森提供了改性加权点模型，减少对该评价机制，称为供应商剖面分析先天的不确定性。该模型包含了决策者通过利用蒙特卡洛仿真技术等级理解周围供应商绩效，而不是从人类的直觉判断。仿真算法根据线性赔偿规则，从随机采样值在每个估计性能范围和组合这些价值观与重要性权重，来生产一种配送的求和分数。这个过程可以为供应商重复几千次。蒙特卡罗的使用使决策者简化了评价模型的输入输出，并提供含有更多信息的相关资料，购买决策的基础比重点的
47、决策模型标准。 （ 6）量纲 分析 在该模型中，每次评估过程涉及一系列一对一比较和可以比较的两家供应12 商。大的空间分析比大于 1 的，等于 1 或不到 1。 Youssef指出此评价方法有两个缺点。首先，价值 = 1 会使决策者选中冷门的供应商。第二，如果大量的供应商必须被评估，过程将变得很漫长和费时的。 3 结论 基于该研究的 AHP 评价和选择供应商系统中提出了这个决定标准质量、准确的交货，价格和服务。在每一个标准的性能，并已转到每个供应商质量损失，对不同供应商的报价 AHP 已经被用来作为评价固化框架之间的交易相关冲突的选择标准。一个数值例子用该模型的功能 也被用来示反应。该评估程序可以就实际的表现通过监控和评估供应商的地址购买需要。它沟通采购优先供应商的态度是很容易理解的。在实际应用中，管理者必须仔细选择因素，最能代表他们竞争的工作重点，目标，也构建两两比较矩阵为了获得更高的目标权重。这样的程序应用该模型将具有更强的真实性。
