【作者】 程勖；

【导师】 杨毅恒；

【作者基本信息】 吉林大学 ， 地球探测与信息技术， 2009， 博士

【副题名】新疆东天山化探数据插值方法研究

【摘要】 本文针对在靶区采集信息点的不足而导致在资源评价中储量估算失真问题,深入研究多种矿业生产中广泛使用的克立格方法,提出了多种克立格法的适用范围及优缺点。使用VC.net、Sql Server开发储量估算可视化操作系统。本文解决了储量估算过程中一些参数的设置问题并提出了变差函数拟合方法的优化;阐述地学空间数据可视化储量估算系统的需求分析;基于传统地质矿产储量估算的工作流程,提出系统的储量估算流程;说明地学储量估算分析系统的总体设计(系统的开发环境、总体结构和功能分析),对系统界面及系统各功能模块的集成与实现进行了分析与总结;解决了实际生产中地质建模和储量估算的可视化、模块化、自动化;详细分析各种克立格方法的适用条件并对各种克立格法插值产生的结果进行对比。在地质统计学理论的发展状况和常用方法的基本原理的基础上,阐明了它优于其它统计方法的原因,并在储量计算模块中,设计了地质统计学的计算方案与实现方法。改进了变差函数的拟合过程,结合加权多项式法和线性规划法的优点,提出使用处理滞后距的方法得到权系数的线性规划法进行球状模型的参数估计。更多还原

【Abstract】 Reserves estimate of mineral resources research is an important element in the field of geology and mineral resources, and geostatistics methods for mineral resources estimating reserves of mineral resources is one of the most effective tools. Using computer technology to enhance the skills of evaluation of mineral resources, improve the credibility, reliability and efficiency of prospecting of the evaluation of results, geological workers are generally concerned about the importance of research topic. Geostatistics is based on regionalized variable theory, in order to function as the basic tool for variation to study the distribution of those in space and showing a certain degree of structural and random natural phenomenon of science.This paper, aiming to the issue of spatial interpolation, developed a geostatistical software system, and made the successful application of the Eastern Tianshan in Xinjiang geochemical data interpolation method of research. The system through analysis of geological prospecting data, established reserves estimates of interpolation model, and estimated target by use of different kriging interpolation methods. Combined with the software development ideas composed and reusable, it determined the components of the system design methodology. Model of mineral resources and resource assessment studies are mainly based on two aspects:1. Interpret the orebody model according to the known geological data;2. Make a correct choice of spatial interpolation methods in the process of estimates of reserves of mineral resources, and complete efficient and accurate estimation of reserves in a short period of time.To solve above the issues, in the geology and mineral resources exploration, the selection of interpolation methods is particularly important, but interpolation methods are diverse (such as Inverse Distance method, cross-section method, geostatistics, etc.), in which the one of better effect interpolation is geostatistics. In geostatistics there are ordinary kriging, kriging association, indicator kriging, universal kriging and restrictive kriging. Geostatistics do not apply directly into the geological and mining area by the current theory of probability and statistics, but from the geological and mining reality, which chooses the appropriate mathematical concepts, theories, models, methods based on the geological characteristics of the variable itself, after that reconstructs, innovates, and adapts to the needs of the special nature of geological movement. Its investigation maximizes the information provided . Such as ,estimating the average grade of deposit of a block by kriging ,not only taking into account the block falls on the sample data, but also the block adjacent to the sample data fall outside.This paper depicted kriging of research and development profiles in the estimates of mineral resources reserves, described detailed kriging principles of mathematical model, including the regionalized variable, variogram, the value above the grade determination and a priori probability select issues.Estimates of reserves of mineral resources described in the work steps and work flow, elaborate on the actual parameter settings and the need to pay attention to some problems. From the ore types and kriging method of time complexity and space complexity, it analyzed the scope of work of the kriging methods , providing a reference how to select the appropriate basis in practical work.In the system development process, this article focuses on the function of the weighted coefficient of variation with the fitting kriging methods to strike a problem. Variogram is the basic tool of geostatistics. For how to improve the theoretical variogram model fitting results has not been a good solution. In this paper, it makes a subtract of the maximum lag distance and lag from the valuation point, after that makes a quotient with the lag from the point, the result of K times as weights to be spherical variogram model fitting. In this way, it solved the case of a negative coefficient, and realized the automation process. Generally, The value of K changes from 1 to 10. When the continuity of mineral deposits is better and the search distance is long, value is in multiples of small K; on the contrary, values is in multiples of large K. In solving the ordinary kriging equations, this paper advanced four applications techniques: automatic fitting spherical model, no nested methods of dealing with anisotropy, the dynamic position of the search parameter estimation, the lower kriging equations dimension. In the indicator kriging, the point of non-sampling the uncertainty of estimates was described in detail, focusing on the median indicator kriging.In this paper, the results of the work achieved and innovation are as follows:1. Summed up the development of the domestic and international reserves estimation methods, discussed the fact the exploration of data read how to create a data model, data structure and modeling methods.2. Explained the requirement analysis of interpolation to estimate reserves, based on traditional reserve calculation of Geology and Mineral Resources of the work flow, advanced the system interpolation method to estimate the reserves process, described the analysis of geo-reserve system design, including system development environment, the overall structure and function analysis of the system interface and system integration of functional modules and implementation of an analysis and summary.3. In the development of geostatistical theory and the basic principles of common methods, discussed that the statistical method is superior to the others, and in the reserve calculation module, designed the methods of interpolation to estimate reserves and implementation of programs.4. Improved the process of variogram fitting, weighted polynomial and the linear programming theory and method as a tool to deal with lag from the proposed use of the method, the spherical model to estimate model parameters.更多还原