模糊C-均值聚类算法和可能性C-均值聚类算法是经典的聚类算法。为克服模糊C-均值聚类算法运行慢的不足，申请者采用竞争学习机制，曾提出抑制式模糊C-均值聚类算法，获得国际同行的认可。鉴于可能性C-均值聚类算法存在聚类中心趋同现象，本项目借鉴抑制式模糊C-均值聚类算法的研究思路，在可能性聚类中引入竞争学习机制，提出一系列抑制式可能性聚类算法，并研究抑制率Alpha和相关参数的选择问题。. 目前，国际上提出的克服可能性聚类算法的聚类中心趋同现象的改进算法均是从目标函数入手。与现有改进方式不同的是，本项目从改变数据的可能性分布入手来解决问题。相比而言，本项目研究具有原理简单、表述直观、易于使用的优点。本项目的研究思路在国际上属首次提出，因此本项目的研究成果具有重要的理论和应用价值。

Fuzzy c-means clustering algorithm (FCM) and possibilistic c-means clustering algorithm (PCM) are two classical fuzzy clustering algorithms. In order to overcome the low running speed of FCM, the applicant adopted the mechanism of competitive learning and proposed suppressed fuzzy c-means clustering algorithm (S-FCM). This work has been recognized by international counterparts. Considering the existence of the phenomenon of the clustering centers convergence in PCM, this project learns from the research method of S-FCM and then introduce the mechanism of competitive learning into the PCM. A series of suppressed possibilistic c-means clustering algorithms are proposed and further the selection of the suppressed rate parameter Alpha and other parameters are studied.. At present, all the modified algorithms which are proposed to overcome the convergence of the clustering centers in PCM are based on the objective function. Being different from these modified algorithms of PCM, this project adopts to change the possibilistic distribution of data to solve the convergence of the clustering centers in PCM. Comparatively speaking, this project behaves well in simple principle, intuitive formulation and being easy to use. The research method of this project is first introduced in the world. Therefore, the research results of this project have important theoretical and applied value.