应用最大熵方法改进克里金负权系数  

Improvement of Kringing negative-weight based on method of maximum entropy

作　　者：陈天伟[1] 卢献健[1] 江海东[2,3] 

机构地区：[1]桂林理工大学测绘地理信息学院,广西桂林541004 [2]中南大学地球科学与信息物理学院,长沙410083 [3]中南大学有色金属找矿预测教育部重点实验室,长沙410083

出　　处：《桂林理工大学学报》2015年第2期340-342,共3页Journal of Guilin University of Technology

基　　金：国家自然科学基金项目(41161072);广西自然科学基金项目(2011GXNSFA018001);广西空间信息与测绘重点实验室项目(桂科能1207115-08)

摘　　要：针对基于地统计学的克里金数学模型,利用最大熵原理分析模型系统对空间权重系数的依赖度增设非负约束条件,研究消除构建格网DEM时的负权问题。通过MATLAB编程验证算法的正确性、准确性,并与拟牛顿法、内点算法解算二次规划的结果进行了比较。对比结果显示,最大熵法得到的权系数大小比例与点位关系相适应,且其估值精度优于后两者。In allusion to Kriging method model based on geologic statistics,principles of maximum entropy are utilized to analyze the model system which depends on modulus of space weight.The nonnegative constraint was set up in the research to eliminate the problem of negative-weight while grid DEMis built.The algorithmic ac-curacy and precision are validated by MATLAB,and compared with the result of quadratic program.Compari-son shows that the volume and scaling of maximum entropys weight is corresponding to the relation of space, with superior accuracy than other two methods.

关 键 词：克里金法 负权 最大熵 

分 类 号：P208[天文地球—地图制图学与地理信息工程] P209[天文地球—测绘科学与技术]