Design optimization of access roads and networks in hard-rock underground mines

矿业是加拿大经济的重要组成部分，为加拿大人的财富和繁荣做出了贡献。地下采矿作业面临着重大挑战。在过去的五年里，商品价格一直很低。由于能源成本高和钻井深度增加，作业成本持续增长。为了维持业务，地下采矿作业需要找到新的方法，以便提高作业的生产力和效率。新的优化工具和技术为管理“新的工作方式”提供了巨大的潜力。本研究计划旨在开发一种用于地下采矿物料搬运和通风的通道网络优化工具。研究将开始探索一种决策支持工具，用于选择中央通道作为竖井或坑道，以及它的尺寸。这种选择将减少问题的规模。然后，利用斯坦纳树和杜宾斯规划路径算法中的新兴方法，实现网络优化。该问题将被表述为在最小转弯半径、最大坡度、稳定性要求和生产能力约束下运输成本的最小化。在研究的第二阶段，重点将网络优化与之前解决的采场布置和排序优化相结合，实现地下矿山规划的全局优化。目标将从成本最小化转向净现值最大化。为此，将构建一个基于分而治之方法的框架，并通过并行模拟退火对框架进行求解。这种求解方法的主要思想是将地下矿山规划的三个子问题并行化，使其在合理的时间内达到最优。在本研究结束时，将向加拿大采矿作业提供一个优化工具。此外，研究成果将有助于矿山规划软件行业的产品组合多样化。

The mineral industry is a significant part of the Canadian economy and contributes to the wealth and prosperity of Canadians. Underground mining operations have important challenges. The commodity prices have been quite low for the last five year. Operating costs continuously grow due to high energy costs and increasing depth. To remain business, underground mining operations need to find new ways of doing such that productivity and efficiency of the operations will be enhanced. New optimization tools and technologies present great potential to manage "new ways of doing". This research proposal aims to develop an optimization tool for access road network used for material handling and ventilation in underground mining. The research will start to explore a decision support tool for selection of central access as shaft or adit, and its dimensioning. This selection will reduce the size of the problem. Then, using emerging approaches in the Steiner tree and Dubins planning path algorithms, the network optimization will be fulfilled. The problem will be formulated as the minimization of haulage costs under constraints of minimum turning radius, maximum gradient, stability requirements, and production capacities. In the second stage of the research, the focus will be the integration of network optimization with the previously solved stope layout and sequencing optimizations such that a global optimization for underground mine planning is attained. The objective will shift from cost minimization to maximization of net present value. For this reason, a framework based on a divide and conquer approach will be constructed, and the framework will be solved by parallel simulated annealing. The main idea behind this solution approach is to parallelize three sub-problems of underground mine planning in such a way as to reach the optimality in a reasonable time. At the end of this research, an optimization tool will be delivered to Canadian mining operations. Also, the research outcomes will assist in diversifying the product portfolio of the mine planning software industry.