摘要
提出了基于小波包和模糊聚类分析的连杆轴承故障的噪声诊断方法。在EQ6100型发动机上预先模拟连杆轴承故障,根据发动机故障时变、非平稳的特点,运用小波包对发动机噪声信号进行特征提取并削减了背景噪声的影响。选取时域上5个参数作为评价故障的特征指标。通过对模糊聚类理论方法的分析比较,引入模糊C聚类划分理论及方法对噪声信号的指标样本进行分类,得到最优分类矩阵和聚类中心,从而建立了故障的标准类型样本。通过对新测取的噪声信号样本进行检验,证明该方法能有效地判断待检样本的类型,诊断连杆轴承故障。
A new method based on the wavelet packet and fuzzy C-clusters about connecting rod bearing fault was put forward. The faults of wear were previously designed on connecting rod bearing of EQ6100 model gasoline engine. Due to signal from the engine noise was time-varied and nonstationary, wavelet packet was suitable to reduce background noise and extract fault characteristics. Five parameters from the time domain were used as value indexes. According to analysis and comparison to the fuzzy cluster theory and methods, we believe that fuzzy C-cluster is a good method to fault recognition. Using it, we got best-classified matrix and center matrix, which could construct for four kinds of normal fault models. Case research proved it's effective to recognize new data of fault model and diagnose the faults of connecting rod bearing.
出处
《农业机械学报》
EI
CAS
CSCD
北大核心
2005年第6期87-91,共5页
Transactions of the Chinese Society for Agricultural Machinery
