智能化地质统计学矿产资源储量计算方法研究

项目名称： 智能化地质统计学矿产资源储量计算方法研究

项目编号： No.40802082

项目类型： 青年科学基金项目

立项/批准年度： 2009

项目学科： 金属学与金属工艺

项目作者： 张夏林

作者单位： 中国地质大学（武汉）

项目金额： 19万元

中文摘要： 课题针对目前地质统计学储量计算过程中由于样点数据的选择方式和计算精度不确定性因素造成该类方法在实际应用过程中存在的问题以及计算结果和计算过程过分依赖计算者经验的现象，提出智能化地质统计学储量计算方法。研究有效样点数据的搜索方式和计算结果的精度利用遗传算法智能地加以优化和调整，以达到储量计算结果为最优的效果。研究地质统计学储量计算方法过程中实验变差函数的计算及拟合、套合结构等需要凭经验反复进行的操作，提取出适当的参数并设计成目标函数或适应度函数，建立起相应的方法模型，然后利用人工神经网络、遗传算法等智能方法对地质统计学储量计算方法中难以一次性确定下来的步骤进行优化，降低地质统计学储量计算方法对人的主观经验的依赖程度，从而使该类方法对矿山复杂的地质数据环境有一定的自适应的能力，可以使计算过程朝进化和智能的方向发展，智能地提供更高质量的储量计算结果。

中文关键词： 地质统计学；智能化；储量计算；GA；ANN

英文摘要： This project proposes the Intelligentized Geostatistical Ore Reserve Estimation Methods in order to resolve the problems that the present processes of geo-statistical ore reserve estimation methods are so dependent to the personal practical experience that it can't be used in a more widely extent and usually it will bring out a result with some randomicity. So, considering the several key steps in the geo-statistical ore reserve estimation methods, we attempt to research some intelligent and valid methods to: 1) pick up the statistical valid sample sets using the genetic algorithm; 2) confirm the typical variability directions, fit the experimental variograms and construct the nested structures of the various theoretical variogram models; 3)smartly choose the relevant samples, surrounding the point to be evaluated , to participate in the calculating process; 4)make the interpolation result accurate both in the local and the whole at the same time. Intelligentizing the processes, which are needed trial and error experimenta-lly by self-experience, can decrease the subjective experience-dependence of the geostatistical ore reserve estimation method and make our new method have some adaptive capability in the complicated mineral geological datum circumstance. Under the condition of implementing the operations simpler, the effect can be reached to make sure that the evaluation process can auto-evolve towards the more fitted parameters as well as the more accurate and credible results. Additionally, based on the theoretic research, the software will be developed to ensure the practicability and popularity of the method.

英文关键词： Geostatistics;Intelligentize;Ore Reserve Estimation;GA;ANN