مدل انتخاب تامین کننده مواد برای توسعه دهندگان اموال با استفاده از آنالیز اجزای اساسی فازی

Nowadays, the participants in construction industry are facing pressure to meet the higher customer expectations under a tighter budget. On the other hand, an effective and efficient material supplier selection system has a significant effect in the business success of property developers. The aim of this study is to investigate a selection model based on Fuzzy Principal Component Analysis (PCA) for solving the material supplier selection problem from the perspective of property developers. First, the Triangular Fuzzy Numbers is used to quantify the decision makers' subjective judgments. Second, PCA is employed to compress the data of the selection criteria and eliminating the multicollinearity among them. Third, the linear combined score of PCA (SCOREPCA) is used to rank the suppliers. Four material purchases are used to validate the proposed selection model. The results show that the proposed model can be adopted in construction material supplier selection by the property developers.

Owing to the construction material market in Mainland China that is not yet well developed, the quality and the price of the materials are varied among suppliers. For this reason, most of the Chinese property developers will purchase the construction materials directly from the material suppliers they selected in order to ensure the building quality and cost control. On the other hand, under the environment of keen bidding and higher customer expectations, Chinese property developers devote more time and resources on business process improvement. The cost of the construction project could be broadly divided into three major groups, namely: materials, labor and overhead. The materials can typically account for around 40% to 45% of the total cost in construction industry [1]. In addition, the labor cost in construction industry is generally governed by the availability of workers within the proximity, only the construction materials can provide the greatest flexibility in seeking the lower cost for the construction companies [2]. Therefore, an effective and efficient material supplier selection model which can help the developers to select the “best” suppliers at the right cost, in the right quantity, with the right quality at the right time has a significant effect in the business success of property developers [3]. The research of the supplier selection has a historical record. It is basically a multiple criteria decision making (MCDM) problem, which consists of three major stages: evaluation, prioritization and selection of alternatives [4]. There are two major approaches for researching supplier selection. The first approach focuses on which criteria are more effective, e.g. Dickson [5] identified 23 criteria for supplier selection based on the survey of 273 purchasing managers. The second approach concerns the selection methods for supplier selection and many methods were proposed to select suppliers in accordance with different selection criteria. However, there are some deficiencies of those methods which are affecting the quality of the result. First, the published methods do not deal with the inter-correlations, which will cause the problem of multicollinearity and add complexity to the evaluation process, among the supplier selection criteria. As a result, trade-offs and repetitive errors, which affect the final selection outcomes, are generated. Second, most of the criteria weighting allocation depend on the subjective judgments, which is not precisely made by the decision makers (DM) and the subjective errors shall affect the final selection outcomes. Third, during the selection process, most of the methods involve a large amount of selection criteria, which are not easy to implement in reality as they require high computational requirements and are time consuming. This study aims to provide a user-friendly supplier selection model for the property developers to select their best material suppliers effectively and efficiently while mitigating the problems identified above.

In this study, a construction material supplier selection model base on Fuzzy PCA is presented. Its merits can be generated as follows. First, the model can efficiently eliminate the multicollinearity among the supplier's attributes and help to reduce the trade-offs and repeatability errors in the selection process. Second, the proposed selection model can also reduce the subjective errors on the sense that the weight assigned for each ζ is generated automatically. The proposed model does not require the complex and subjective pair-wise comparison to determine the weights of the selection criteria, which mitigates such kind of subjective errors. Third, the proposed model can largely reduce the dimensionality of the selection data with little loss of the information. In the case study, the total variance (up to 87.855%) has been retained with the extraction rates of all the attributes range from 76.6% to 96.9%. Fourth, the qualitative data can be quantified by using TFN, which translates the DMs' subjective judgments into quantitative data efficiently and effectively. Therefore, the proposed model can handle both the quantitative and qualitative data at the same time. Finally, the proposed model is a user-friendly tool for supplier selection because of the availability of the software (SPSS and Microsoft Excel). The model presented in this study is just a tailor-made selection model for the target company which aims to present the ideas about how to set up a Fuzzy PCA based selection model to solve specific supplier selection problem. When it is applied to other company or other industry, one should first study the company's characteristics and the industry's requirements. For instance, in order to determine the tailor-made supplier selection criteria, a survey i.e. questionnaire is strongly recommended to firstly undertake so as to understand the target company's requirements and strategies towards the suppliers and the specific requirements of the industry. In addition, since the DM's linguistic judgments in accordance with the qualitative criteria are subjective, the qualitative factors involved in the decision are the preferences of the DM in accordance with his/her knowledge, experience and priority. Different DM may have different perceptions towards the same performance and such difference sometimes maybe quite explicit. When using the proposed model, it is recommended to provide guidance or benchmark for the DMs so as to assist their linguistic judgment markings and ensure the results do not have too much deviation.