论文部分内容阅读
采矿工程师需要的辅助工具应该具有可靠、灵活、很快形成矿体模型的特征。本文的独创之处就在于把以下三种类型的资料结合在一起:地质描述的平面图;地质统计估值的结果;从矿体“专家系统”获得的矿化带的一般几何形态的知识。其主导思想是先对克里格方块和地质资料进行分类整理,然后通过一个参考椭球的修正导出一个“矿体”模型,并确定其空间位置。这种方法已被用于索杰雷姆公司的萤石矿。该方法用了人工智能的技术,从而可建立一个能通过掌握知识库和模拟采矿方法来辅助矿山计划和生产管理系统。
Auxiliary tools that mining engineers need should be reliable, flexible and quickly form the characteristics of the ore body model. The original idea of this article is to combine the following three types of data: a plan of the geological description; the result of a geostatistical valuation; and knowledge of the general geometry of the mineralized zone obtained from the “expert system” of the ore body. The dominant idea is to classify the kriging cube and geological data first, and then derive an “ore body” model by a reference ellipsoid correction and determine its spatial location. This method has been applied to Sogerem’s fluorite mine. This approach uses artificial intelligence techniques to create a system that assists mine planning and production management through mastering knowledge bases and simulating mining methods.