基于改进MF-DFA和SSM-FCM的液压泵退化状态识别方法

Degradation state identification method of hydraulic pump based on improved MF-DFA and SSM-FCM

针对液压泵振动信号通常具有非线性强和信噪比低的特点,提出了一种基于改进多重分形去趋势波动分析(MF-DFA)和半监督马氏距离模糊C均值(SSM-FCM)的液压泵退化状态识别方法。该方法首先引入滑动窗口技术改进传统MF-DFA方法在时间序列数据分割过程中存在的不足,提高MF-DFA方法的计算精度;然后利用改进MF-DFA方法计算液压泵多重分形谱参数,并分析了不同分形谱参数对液压泵退化状态的反映能力,选取奇异指数α_0和广义Hurst指数波动均值Δh(q)作为退化特征量;最后利用半监督马氏距离模糊C均值方法实现了液压泵退化状态识别,并以液压泵实测数据为例验证本文所提方法的有效性。

Aiming at the characteristic that the vibration signals of hydraulic pump usually have strong nonlinearity and low signal to noise ratio,this paper presents a hydraulic pump degradation state identification method based on improved multi-fractal detrended fluctuation analysis （MF-DFA） and semi-supervised Mahalanobis fuzzy c-means （SSA-FCM）. First of all, the sliding window technology is introduced to improve the deficiency in the time series data partitioning of traditional MF-DFA method and enhance the calculation accuracy of the MF-DFA method. And then, the improved MF-DFA method is used to calculate four kinds of muhi-fractal spectrum parameters of hydraulic pump. The presenting capability of five fractal spectrum parameters on hydraulic pump degradation state is analyzed, and as a result the singular index ao and the mean of the fluctuation of generalized Hurst index, Att（q） are selected as the degradation features. Finally, SSA-FCM method is adopted to achieve the hydraulic pump degradation state identification. The hydraulic pump real test data are taken as examples to verify the effectiveness of the proposed method.