Goaf-collapse sites stability evaluation based on principal component hierarchical clustering model

Stability evaluation of goaf-collapse sites is the primary problem to be solved in the subsequent engineering construction. In this paper, a principal component hierarchical clustering analysis method for a goaf-collapse site stability evaluation has been proposed to solve the problem caused by multiple influencing factors and complicated geological and mining conditions of a steeply pitching phosphate orebody. On the basis of determining the stability evaluation range of the goaf-collapse site, 8 major indicators representing stability of goaf-collapse sites were selected after principal component analysis (PCA) as learning samples for training. AGNES (AGglomerative NESting) hierarchical clustering analysis model for evaluating stability of goaf-collapse sites was established. After dimensionality reduction, first four principal components of cumulative contribution rate were 81.8%. The results show that the goaf-collapse site can adapt to different land bearing capacity of urban planning in the study area, the discriminant result is consistent with other methods, it indicates that the hierarchical cluster analysis model has a good discriminant ability. The proposed approach demonstrates the feasibility and effectiveness in the field of stability assessment of goaf-collapse sites.