摘要
本文采用凝聚式层次聚类算法对块体数据进行预处理,显著降低了混合整数线性规划(MILP)数学模型在矿业应用中的复杂度,以解决矿业排产优化中的复杂问题。该方法的核心在于根据块体的地质和采矿相关属性的相似性,将大量小块体合理聚合成相对较少的大聚合单元,从而简化了模型的变量和约束条件,减少了求解优化问题所需的计算资源和时间。通过某大型露天金矿采矿联合体的应用案例研究,证明了凝聚式层次聚类算法在实践中的有效性。原始的MILP模型由于块体数量巨大和计算复杂度高无法直接求解。应用聚类算法后,将36183个块体合理减少到5810个聚合单元,显著降低了问题的规模。
This study aims to significantly reduce the complexity of Mixed Integer Linear Programming(MILP)mathematical models in mining applications by using a hierarchical clustering algorithm for pre-processing block data,thereby solving complex problems in mining scheduling optimization.The essence of this method lies in the aggregation of numerous small blocks into relatively fewer and larger consolidated units based on the similarity of geological and mining-related properties of the blocks.This approach simplifies the variables and constraints of the model,reducing the computational resources and time required to solve optimization problems.An application case study of a large open-pit gold mining complex demonstrates the effectiveness of the hierarchical clustering algorithm in practice.The original MILP model,due to the immense number of blocks and high computational complexity,was not directly solvable.After applying the clustering algorithm,the number of blocks was sensibly reduced from 36183 to 5810 consolidated units,significantly reducing the scale of the problem.
出处
《中国矿山工程》
2024年第2期12-16,共5页
China Mine Engineering
