Towards an integrated, self-learning stochastic mining complex framework and new digital technologies for the sustainable development of mineral resources

为矿产资源的可持续发展建立一个集成的、自学习的随机采矿复杂框架和新的数字技术

• 批准号: RGPIN-2021-02777
• 负责人: Dimitrakopoulos, Roussos
• 金额: $ 6.41万
• 依托单位: McGill University
• 依托单位国家: 加拿大
• 项目类别: Discovery Grants Program - Individual
• 财政年份: 2022
• 资助国家: 加拿大
• 起止时间: 2022-01-01 至 2023-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=761083
• 关键词: Towards; integrated; self; learning; stochastic

Over the past decade, the concept of a mining complex or mineral value chain was introduced to reflect an integrated system that manages the extraction of materials from a group of mines, followed by the treatment of the extracted materials through different interconnected processing facilities. This system generates sellable products delivered to customers and/or the spot market. Mining complexes and related research contribute to the sustainable development of mineral resources and reserves that ensures the continued supply of raw materials and metals we rely upon, while managing environmental aspects. Technical aspects and components of a mining complex are substantially affected by uncertainties (stochasticity) stemming from multiple sources. These range from the valuation of assets to their operational performance and include their ability to adapt to endogenous and exogenous changes. Uncertainty-related effects are compounded with decision-making for a multitude of decisions from extraction to transportation. Critical sources of uncertainty of this integrated system include the quality and quantity of materials produced from the mines (supply uncertainty) and the metal's spot market price (demand uncertainty). Given new technological developments, it is important to address these uncertainties and assimilate new information collected as a mining complex operates, including from various sensors. This needs to be evaluated and used to update models, forecasts and further support complex decision-making. The applicant's ongoing research has generated substantial new developments to address these challenges. This has shifted the related paradigm and contributed to a new framework termed simultaneous stochastic optimization of mining complexes (SSOMC). It has also generated new stochastic optimization - simulation methods that integrate and manage uncertainty, including but not limited to geological uncertainty, while maximizing productivity, extending life of asset, and increasing return on investment. The new framework jointly optimizes components of a mining complex, including mine production schedules, blending, stockpiling, aspects of material processing streams and capital investments that define the critical bottlenecks in the mining system. These contributions support the development of a new generation of intelligent risk-management technologies. The proposed 5-year research program aims to build upon our previous work to develop the next level of state-of-the-art approaches. It intends to increase understanding of a new stochastic self-learning, modelling and optimization framework and related technologies by: developing an all-inclusive supply-meets-demand approach for SSOMC to respond to unveiling information; extending high-order quantification of spatial uncertainty to include statistical learning; and exploring intelligent approaches for jointly optimizing different time scales, while addressing multi-source data assimilation.

在过去的十年中，采矿综合体或矿产价值链的概念被引入，以反映一个集成系统，该系统管理从一组矿山中提取材料，然后通过不同的互连加工设施处理提取的材料。该系统生成交付给客户和/或现货市场的可销售产品。采矿综合体和相关研究有助于矿产资源和储量的可持续发展，确保我们所依赖的原材料和金属的持续供应，同时管理环境方面。采矿综合体的技术方面和组成部分很大程度上受到多种来源的不确定性（随机性）的影响。这些范围从资产估值到其运营绩效，包括其适应内生和外生变化的能力。与不确定性相关的影响与从开采到运输的众多决策的决策混合在一起。该集成系统的关键不确定性来源包括矿山生产的材料的质量和数量（供应不确定性）以及金属的现货市场价格（需求不确定性）。鉴于新技术的发展，解决这些不确定性并吸收采矿综合体运营时收集的新信息（包括来自各种传感器的信息）非常重要。这需要进行评估并用于更新模型、预测并进一步支持复杂的决策。申请人正在进行的研究已经产生了大量的新进展来应对这些挑战。这改变了相关范式，并促成了一个称为采矿综合体同时随机优化（SSOMC）的新框架。它还产生了新的随机优化——集成和管理不确定性（包括但不限于地质不确定性）的模拟方法，同时最大限度地提高生产率、延长资产寿命并提高投资回报。新框架联合优化了采矿综合体的各个组成部分，包括矿山生产计划、混合、库存、材料加工流程的各个方面以及定义采矿系统关键瓶颈的资本投资。这些贡献支持新一代智能风险管理技术的开发。拟议的 5 年研究计划旨在以我们之前的工作为基础，开发更高水平的最先进方法。它旨在通过以下方式增进对新的随机自学习、建模和优化框架及相关技术的理解： 为 SSOMC 开发一种全面的供需方法来响应揭晓的信息；扩展空间不确定性的高阶量化以包括统计学习；探索联合优化不同时间尺度的智能方法，同时解决多源数据同化问题。