Evaluation of mineral resources carrying capacity based on the particle swarm optimization clustering algorithm

Abstract

SYNOPSIS As minerals are a non-renewable resource, sustainability must be considered in their development and utilization. Evaluation of the mineral resources carrying capacity is necessary for the sustainable development of mineral resource-based regions. Following the construction of a comprehensive evaluation index system from four aspects, namely resource endowment, socio-economic status, environmental pollution, and ecological restoration, a method combining particle swarm optimization (PSO) and the K-means algorithm (PSO-Kmeans) was used to evaluate the mineral resources carrying capacity of the Panxi region southwest Sichuan Province, China. The evaluation method is data-driven and does not consider the classification standards of different carrying capacity levels. At the same time, it avoids the problems of local optimization and sensitivity to initial points of the K-means algorithm, thereby providing more objective evaluation results and solving the problem of subjective division of each grade volume capacity in carrying capacity evaluation. The algorithm was verified through UCI data-sets and virtual samples. By superimposing a single index on the carrying capacity map for analysis, the rationality of the evaluation results was validated. Keywords: particle swarm optimization, K-means algorithm , mineral resources, carrying capacity, sustainability.