基于模糊C-均值的空间机构轴系健康评估方法
Fuzzy C-means Based Health Assessment Method for Space Mechanism Shafting

针对现有轴承健康评估方法在空间机构轴系上应用难度大的问题，提出了一种基于模糊C-均值聚类的空间机构轴系健康评估方法。首先，定义了新特征参数——特征谱线明显度；其次，提取了优良、故障两类典型轴系及待评估轴系振动实测信号峭度和包络频谱特征谱线明显度；然后，以峭度和特征谱线明显度两种特征参数为输入，通过模糊C-均值聚类模型获得了各待评估轴系对优良轴系类的隶属度；最后，根据隶属度计算轴系健康指数。应用实例表明,该方法既可对轴系健康状态正确排序，又可在高转速下给出可靠的健康指数，验证了该方法的可行性和有效性。
A health assessment method for space mechanism shafting based on fuzzy C-means clustering is proposed in the light of applying the existing health assessment methods to space mechanism shafting. First, a new parameter named characteristic frequencies’ distinctness is defined. Then, the fuzzy C-means clustering model is trained by the kurtosis and characteristic frequencies’ distinctness of fine and faulted shafts. Finally, the health index is constructed according to the membership related to the fine shafting cluster. With the proposed method, the shafts can be sorted correctly according to the health state, and the reliable health index can be obtained at high rotation speeds. The practicability and effectiveness of the approach proposed have been demonstrated.