基于密度的聚类算法因其抗噪声能力强和能发现任意形状的簇等优点，在聚类分析中被广泛采用，本文提出

的基于相对密度的聚类算法，在继承上述优点的基础上。有效地解决了基于密度的聚类结果对参数值过于敏感、参数

值难以设置以厦高密度簇完全被相连的低密度簇所包含等问题。-Density-based clustering algorithm because of its strong resistance to noise and can find clusters of arbitrary shape, etc., in the cluster analysis is widely used, the proposed clustering algorithm based on the relative density, in succession on the basis of the above-mentioned advantages. Effectively solve the density-based clustering results on the parameter value is too sensitive, difficult to set the parameter value to high-density clusters of buildings connected to the low-density clusters are completely contained and other issues.