基于聚类和微元法的林草业智能规划系统  

Intelligent Planning System of Forestation and Grass Based on Clustering and Infinitesimal Dividing Method

作　　者：张勇进[1,2] 李淼[1,2] 张建[1,2] 孙涛[1,2] 

机构地区：[1]中国科学院合肥智能机械研究所,合肥230031 [2]中国科学院研究生院,北京100039

出　　处：《计算机工程与应用》2005年第25期180-182,共3页Computer Engineering and Applications

基　　金：国家863高技术研究发展计划:信息技术项目基金(编号:2001AA115180)

摘　　要：文章用改进的K-Means聚类方法对甘肃省林草业智能规划系统中三种种植作物(天然草、人工草和树)的各适宜地点分别进行了聚类,然后用微元划分法将聚类得到的各簇中的所有点包围在一个相应的多边形中,从而直观地表达了三种种植作物的适宜种植区域。This paper presents a modified K-Means clustering which is respectively employed to cluster the planting localities of artificial grass,inartificial grass and tree in intelligent planning system of forestation and grass,then the planting localities of each cluster are surrounded with a corresponding polygon by means of infinitesimal dividing method,as a result,the areas which are befitting to plant artificial grass,inartificial grass and tree are well sketched respectively.

关 键 词：K-均值 聚类 微元划分法 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]