• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>International Congress on Automation in Mining >A Machine Learning and Data Analysis Approach to Improve Productivity in Mining Operations
【24h】

A Machine Learning and Data Analysis Approach to Improve Productivity in Mining Operations

机译：一种机器学习和数据分析方法，提高采矿业务的生产力

获取原文
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

With advances in communications, sensors, processing and other related technologies, companies are collecting more data than ever before. The challenge now becomes how to leverage this data to improve operations, productivity, reduce maintenance costs, and therefore, achieve operational excellence. In the mining industry, load and haul costs are significant but achieving accurate control of productivity, even with a fleet management system, is a difficult task. It is well known that each loading point exhibits different behaviours, but due to the inability to acquire more detailed information about what happened in the mine, it is not possible to detect which of them fails to achieve the desired productivity levels. Indeed, commonly the productivity plan is based on monthly results from all mine data. In order to improve the available information details, a simple and functional system was developed, which is able to analyse and consolidate mine productivity at the loading point level, using only data from haul trucks. Thus, the system helps to establish an improved mine plan in order to deploy an effective and informed course of action in each loading point that underperforms the operation, instead to apply actions at global level.The system is currently under validation in a Chilean mine where we are iterating the dashboard to corroborate the veracity of the information supplied and set what additional information the company is interested in seeing. One interesting discovered fact was the relationship between under and over load, while the overload represents 29% of the haul truck cycles with a total of 8000 Tonnes over the nominal operation, underload represent just 12% of the haul truck cycles but almost 12000 lost Tonnes, which may have big impact in maintenance plan and productivity. Furthermore, it may influence availability, which is an important KPI for mine companies.
机译：随着通信，传感器，加工和其他相关技术的进步，公司的收集量多于以往任何时候。挑战现在成为如何利用此数据来提高运营，生产率，降低维护成本，因此实现运营卓越。在采矿业，负载和运输成本显着但实现了对生产力的准确控制，即使是舰队管理系统也是一项艰巨的任务。众所周知，每个负载点呈现不同的行为，而是由于无法获取关于矿井发生的事情的更详细信息，因此无法检测到哪一个不能达到所需的生产率水平。实际上，通常，生产力计划是根据所有矿山数据的每月结果。为了改善可用的信息详细信息，开发了一种简单且功能的系统，能够仅使用来自Haul Trucks的数据来分析和巩固矿井生产力。因此，该系统有助于建立一个改进的矿山计划，以便在每个装载点中部署有效和明智的行动方案，而是在运行不足的每个装载点，而是在全局级别应用动作。当前在智利矿区验证系统我们正在迭代仪表板来证实所提供信息的准确性，并设置公司对看到的其他信息。一个有趣的发现事实是负载下和过负载之间的关系，而过载代表了总共8000吨的标称手术中的29％，欠载仅占运输卡车周期的12％，但近12000吨，这可能对维护计划和生产力产生了很大的影响。此外，它可能会影响可用性，这是矿山公司的重要KPI。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号