Advanced CRITIC–GRA–GMM model with multiple restart simulation for assuaging decision uncertainty: An application to transport safety engineering for OECD members
Ziyu Zhou,Yuhao Zhang,Yi Zhang,Bodong Hou,Yunhao Mei,Peijun Wu,Yichu Chen,Weijie Zhou,Haoyan Wu,Faan Chen
DOI: https://doi.org/10.1016/j.aei.2024.102373
IF: 8.8
2024-01-30
Advanced Engineering Informatics
Abstract:When dealing with multi-criteria decision-making (MCDM) activities in an uncertain environment, one of the fundamental requirements of the methodology used is that it provides creditable and defensible decisions. This is particularly true for transport safety engineering. In the current study, a hybrid decision-making model was developed by integrating several advanced mathematical models and an embedded simulation. Specifically, this study integrates criteria importance through intercriteria correlation (CRITIC), grey relational analysis (GRA), and the Gaussian mixture model (GMM), namely CRITIC–GRA–GMM with multiple restart simulation; the ultimate aim is to offer a decision tool with substantial stability and reliability. In particular, this study addresses the issue faced by conventional GMM – namely uncertain initialization and trapping in local optima – by embedding a multiple restart simulation to obtain a more robust model. The multilevel contrasts of the results in the application to transport safety engineering in Organisation for Economic Co-operation and Development (OECD) countries verified the efficiency, quality, and reliability of the proposed model, indicating its practicability in real-life decision-making activities. This novel framework contributes to existing decision approach databases by enhancing the decision-making process. Overall, the current study provides decision-makers and policymakers in OECD countries with a valuable tool for identifying the strengths and weaknesses of transport safety, enabling the formulation of effective measures and action plans to maximize safety performance.
engineering, multidisciplinary,computer science, artificial intelligence