Risk evaluation of construction safety hazard based on multi-source evidence fusion and LEC method

Abstract

Mega project represents a multifaceted and dynamic process, characterized by diverse and extensive safety data sources. The existing safety risk evaluation method fail to make full use of safety big data, resulting in the lack of comprehensiveness, objectivity and accuracy of the evaluation result. To accurately assess the risk of safety hazard sources in large-scale engineering construction, a risk evaluation method of construction hazard sources based on D-S evidence fusion and LEC method is proposed. Firstly, the source identification of construction safety hazards is achieved by using the keywords of safety hazards and strong association rules, and the hazard source risk estimation index and its quantification method are given. Then, the D-S evidence fusion theory is applied to process the multi-source information derived from experts' risk estimation result. This process yields corresponding values for human exposure and accident severity indices pertaining to the identified hazard sources. Finally, an improved LEC method is employed to establish a risk evaluation model. To demonstrate the effectiveness of this approach, a case study is conducted, focusing on the hazard source of construction safety hazards in an underground main powerhouse project. The result indicated that the method significantly enhances the accuracy and objectivity of hidden danger risk evaluation, providing scientific data support and a foundation for decision-making in the management of safety hazards during mega project construction.

Keywords

mega project / construction safety / hazards identification / risk evaluation / D-S evidence fusion theory / LEC method