A Hierarchical Fuzzy Expert System for Contractor Prequalification

Abstract

The selection of qualified contractors for construction projects plays a pivotal role in ensuring the quality of project delivery, and as such, represents a critical decision for project owners. For an informed selection, multiple criteria, such as cost, quality, experience, safety records, and past performance, need to be considered. However, due to the multiple criteria for contractor prequalification, each with varying degrees of relative importance that are difficult to measure using quantitative data, the selection process can prove challenging for owners. The objective of this paper is to propose a model for contractor prequalification that combines the use of a fuzzy expert system and hierarchical diagramming. To achieve this aim, we first utilized various types of questionnaires to gather the main criteria, relevant subcriteria, and initial fuzzy numbers. Subsequently, we developed a fuzzy expert system for each subcriterion and the main system. Finally, by exploring different t-norms, s-norms, defuzzification types, and tuning membership functions, we selected the best type of fuzzy system for each criterion. MATLAB software was employed for coding purposes in this study. In order to validate the model’s efficacy, a comparative analysis was conducted between the scores assigned to individual contractors as generated by the fuzzy logic model, and the factual scores allocated to the respective contractors. The observed discrepancy in the model’s accuracy ranged between 10% and 15% which means an 85% to 90% similarity between the model’s evaluation and actual scoring. Selecting competent contractors for construction projects is crucial to ensuring project quality and success. The paper introduces a groundbreaking approach to simplify this process. Traditionally complex due to diverse criteria and their varying significance, contractor selection is now more accessible. In simple words, this research introduces a smart method for choosing the right contractors. By combining a fuzzy expert system and a clear hierarchy, the study makes it easier for project owners to pick the best team. Imagine having a tool that considers multiple factors, even those that are not easily measured, to help you choose contractors wisely. This approach, developed using various questionnaires and advanced algorithms, empowers project owners to make more informed decisions. In essence, this research streamlines contractor selection, ensuring projects have the best possible chance for success.