航空物流传送设备中轴承故障稀疏特征提取

Sparse Feature Extraction of Fault Bearing in Aviation Logistics Transmission Equipment

航空物流传送设备中的轴承由于长期受外侵灰尘影响,其内外环极易发生故障;利用计算机采集轴承的振动信号并进行故障特征提取是轴承故障诊断的常用方法;提出了基于稀疏分解的轴承故障特征提取方法;首先,分析轴承故障特征稀疏提取原理;然后,构造参数化Gabor字典,利用遗传算法对故障特征成分进行匹配追踪(Matching Pursuit,简称MP),以峭度值最大原则作为迭代结束条件;最后,重构提取的特征成分,进行包络谱分析,得出故障类型;对仿真数据和轴承振动数据的测试结果表明,所提方法能有效提取轴承故障特征成分,为航空物流传送设备中轴承的故障监测提供了一种有效方法。

Due to the long-term influence of dust from outside intrusion,it often happens faults in inner race or outer race of bearings in aviation logistics transmission equipment.That acquisition of vibration signals and extraction of the fault feature is a common method for bearing fault diagnosis.A fault feature extraction method for bearing based on sparse decomposition is proposed.Firstly,the mechanism of sparse extraction is analyzed.Then,a parameterized dictionary is predefined and Matching Pursuit （MP） optimized by Genetic Algorithm （GA） is utilized to extract the bearing fault feature,whose iterating termination condition is determined by the kurtosis of approximation signal.Finally,the feature component is reconstructed and the fault type can be determined according to its envelope spectrum.The test results of simulate data and bearing vibration data demonstrate that the proposed method can effectively extract the fault feature component and meets the demand of real-time bearing condition monitor in aviation logistics transmission equipment.