Optimization of Safety Evaluation Indicator System of Prefabricated Building Construction Based on Grey Dynamic Clustering and Rough Set
The establishment of a rational safety evaluation indicators system for prefabricated construction is of paramount importance for the standardization of the safe construction of modular buildings.A novel method combining grey dynamic clustering and rough set attribute reduction is introduced to analyze and simplify the indicators.The grey relational analysis is utilized to establish a grey relational matrix between samples,and an F-statistic is used to determine the optimal threshold for the best categorization of samples.Subsequently,each indicator is removed one by one,and the remaining indicators are clustered using the grey dynamic clustering method to obtain their optimal clustering results.Finally,the theory of rough set reduction is applied to compare the optimal clustering result of a specific indicator with the results of removing that indicator,retaining only those indicators that differ from the original optimal clustering result and thus affect the classification of samples.This measure has proven effective in screening safety evaluation indicators,reducing the workload of subsequent construction safety evaluations,ensuring the accuracy of safety evaluations,and providing targeted guidance for the management of prefabricated building construction sites.