Occupational Health and Safety Risk Assessment Using Social‐Trust‐Driven Consensus Method and Deep Contrastive Clustering

ABSTRACT

As a crucial process in occupational health and safety management, the occupational health and safety risk assessment (OHSRA) aims to evaluate, prioritise, and reduce the risk of occupational hazards in workplaces to ensure employee health and safety. In this paper, a new integrated OHSRA model combining the social‐trust‐driven consensus (STDC) method and the deep contrastive clustering is suggested for the risk assessment and classification of occupational hazards under a polytopic fuzzy linguistic environment. Specifically, the STDC method is applied to achieve expert consensus on the risk assessments of occupational hazards, and the deep contrastive clustering is utilised to obtain the risk classifications of occupational hazards. In addition, the polytopic fuzzy linguistic theory is employed to deal with experts' hesitant risk assessment information. Finally, a practical case in coal mining is provided to illustrate the applicability and effectiveness of the proposed OHSRA model. The results show that equipment‐related risks and injuries caused by falling objects are the highest‐risk occupational hazards and should be controlled by appropriate precautions. Moreover, a comparative analysis confirms that the proposed model can effectively cope with inconsistent linguistic risk assessments given by experts and produce reliable risk classifications of occupational hazards in OHSRA.