基于快速排序绝对值特性全连接分类器的轴承故障诊断

Bearing fault identification based on fast sorting absolute feature fully-connected classifier

工业生产中的机械设备故障是一个反复出现的问题。一个微小的故障会对机械设备造成不可逆转的损伤。严重的机械设备故障会导致企业生产部门停产,进而导致企业的经济效益下降。提供准确及时地故障诊断能最大程度避免轴承故障的出现。提出一种快速排序绝对值特性全连接分类器(fast sorted absolute feature Fully-connected classifier,FSAFCC)用于轴承的故障识别。FSAFCC有两个部分。FSAFCC第1部分是数据预处理部分。数据预处理部分是对采集到的原始信号进行随机时刻截取、取绝对值和排序的操作。FSAFCC的第2部分是利用全连接分类器进行轴承的故障分类。FSAFCC在数据预处理部分只需要进行取绝对值和重新排序的计算;FSAFCC的数据预处理部分计算步骤简单,而且计算量很小,所以FSAFCC的运算速度很快。FSAFCC方法在美国西储大学轴承故障数据集,轴向柱塞液压泵数据集和自吸离心泵数据集中进行实验,实验结果表明,FSAFCC的方法在轴承故障诊断领域不但有着很高的准确度,而且所用时间很短。

The failure of mechanical equipment in industrial production is a recurring problem.A small failure could cause irreversible damage to mechanical equipment.Severe mechanical equipment failure can lead to production stoppages in the production sector,resulting in a decline in the economic efficiency of the enterprise.Accurate and timely fault diagnosis maximally mitigates the occurrence of bearing failures.This paper proposes a fast sorted absolute feature fullyconnected classifier(FSAFCC)for fault identification of bearings.The FSAFCC has two parts.The first part of FSAFCC is the data preprocessing part.The data preprocessing part is the operation of intercepting at random moments,calculating absolute values,and sorting the acquired raw signals.The second part of FSAFCC is the fault classification of bearings utilizing a fully-connected classifier.The FSAFCC only needs to sort absolute values in the data preprocessing part.The data preprocessing part of FSAFCC has simple calculation steps with fast calculation speeds.The proposed FSAFCC experiments in the bearing failure,axial piston hydraulic pump,and self-priming centrifugal pump datasets;the experimental results show that the FSAFCC achieves high fault diagnosis accuracy with short training time.