Environment
Ali Rasouli; Akbar Esmaeilzadeh; Reza Mikaeil; Solat Atalou
Abstract
Identifying joint sets is essential in engineering geology for rock mass classification and slope stability analysis in mining. Accurate clustering of joint sets based on dip and dip direction enhances the understanding of rock behavior and ensures stability in mine walls. This study presents a novel ...
Read More
Identifying joint sets is essential in engineering geology for rock mass classification and slope stability analysis in mining. Accurate clustering of joint sets based on dip and dip direction enhances the understanding of rock behavior and ensures stability in mine walls. This study presents a novel clustering approach integrating the Harmony Search (HS) and Particle Swarm Optimization (PSO) algorithms to classify joint sets in the Sungun copper mine. Initially, joint characteristics were classified using the Fuzzy C-Means (FCM) method, with the elbow method selecting a four-class clustering solution. To optimize clustering, FCM was combined with HS and PSO, and joint data were assessed using Davies-Bouldin, Calinski–Harabasz, and Silhouette indices. The results demonstrated that the hybrid FCM-PSO method outperformed alternatives, achieving scores of 0.80, 347.48, and 0.57, respectively, indicating superior clustering performance and stability. In contrast, the FCM-HS method performed worse than FCM alone, ranking third overall. The findings confirm that FCM-PSO effectively classifies joint sets, providing reliable insights into rock mass behavior in the Sungun mine. Considering the features and advantages of the FCM-PSO method, it is concluded that the proposed approach has significant potential for effective joint classification in mining engineering. This improved clustering approach enhances geological analysis, supporting safer and more efficient mining operations.
