An integrated framework for effective safety management evaluation: Application of an improved grey clustering measurement

Abstract

Safety evaluation is an important and challenging issue in many industries and is also a key component of risk management. Various evaluation methods have been proposed to make safety evaluation more consistent and objective. However, a major concern is that many existing safety evaluation measurements are still subjective and are not easy to obtain in a uniform way. This paper aims to develop a framework suitable for evaluating the safety performance at organizational or project levels in a comprehensive way that may be expected to reflect all risk assessment aspects and make best use of professional talent and experiences from different evaluators. In this paper, a structural evaluation logic is proposed, based on an improved grey clustering method. First, a grey clustering-based indicator system is developed to avoid the arbitrary selection of indicators. Then, a novel interval-grey-number reciprocal-judgment-matrix based AHP (GRAHP) is proposed that extends the classical analytic hierarchy process to deal with the possible contradictory opinions from experts in different fields as well as the standardization problem of collected independent and uncertain data in evaluation. Additionally, the transformation and optimization method of the new proposed grey hierarchy analysis model are given. Finally, an improved grey variable weight clustering evaluation model is built based on the above described methods. We illustrate the practical implementation of the proposed methods using actual aviation data from China. The results show that the proposed framework and methods have good ability in safety evaluation for large and complex engineering projects.