中文摘要：

传统的模糊聚类分析方法分两个步骤,首先使用目标函数进行模糊生成,然后应用聚类有效性函数决定聚类的最佳数目。针对模糊形成和有效性验证函数的内在不同所导致的聚类不准确性,应用一种新的基于双目标模糊聚类分析（BOFCM）的聚类方法。同时将基于三角形隶属函数的T-S模糊系统应用于非线性系统辨识中,该方法可以很方便地确定输入空间的划分及隶属函数的形状,减少了计算量。将以上方法应用于一个二阶系统辨识分析,证实了该方法的有效性。

英文摘要：

The traditional fuzzy clustering analysis has two steps,the objective function was used for fuzzy generation and cluster effective function was used for validating numbers of cluster.However,the intrinsic differences of the formation and validation functions may cause Inaccuracy A fuzzy clustering based on bi-objective（BOFCM） analysis is proposed.Moreover,Triangular membership function based on TS Fuzzy Systems is applied to nonlinear system identification.This method can easily determine the input space location and the shape of membership functions,reducing the computational complexity.The above method is confirmed by applying it to analyse a second-order system.