Abstract:

To improve the safety management level of multi-industry groups, a safety performance evaluation system based on the classification and evaluation of the safety management difficulty of subsidiary companies was constructed. First, six key indicators including industry attributes, scale, and employee count were comprehensively analyzed. Indicator weights were determined through the AHP and Entropy Weight Method(EWM). Subsidiary safety management difficulty categories were identified via K-means clustering. Then, such main elements as organizational systems, risk control, emergency management, and accident indicators were comprehensively analyzed. From three dimensions of control indicators, constraint indicators, and incentive indicators, a Plan-Do-Check-Act (PDCA) performance evaluation mechanism of "quarterly self-assessment by subsidiaries + annual comprehensive evaluation by the group" was established to achieve differentiated management for different subsidiaries. Finally, a SRI monitoring and early warning model was developed to quantitatively assess key indicators in the safety production process and enable graded early warnings and interventions. Research results show that the safety management difficulty evaluation model enables the quantitative classification of the safety management difficulty of subsidiaries. The safety performance evaluation system, together with its PDCA cycle mechanism, realizes differentiated assessment and incentives for subsidiaries of different categories, thereby promoting continuous improvement of safety performance. The SRI monitoring and early warning model achieves quantitative evaluation and graded early warning of key indicator failures.

Key words: multi-industry groups, safety performance evaluation, safety management difficulty, analytic hierarchy process (AHP), safety risk index (SRI)