基于自适应免疫检测器的变速器轴承故障诊断研究

Transmission bearing fault diagnosis based on adaptive immune detector

针对变速器轴承振动信号不平稳、特征难以提取等问题,以对振动信号的幅值和频率的变化不敏感的5种无量纲参数为特征参数,将这5种特征参数分为自我模式和非我模式,并基于人工免疫原理设计了一种自适应免疫检测器来克服传统变速器轴承故障诊断方法的局限,重新定义了自适应变异和抑制阈。在变速器轴承实验台上,首先对变速器轴承6种状态下的振动加速度信息进行了测量;然后计算了这些振动信息的无量纲指标;再用所得到的5种无量纲特征参数训练免疫检测器,使之成为成熟检测器;最后采用训练好的成熟检测器对变速器轴承的故障状态进行了诊断。试验结果表明:利用自适应免疫检测器进行变速器轴承故障诊断是可行的,而且该检测器具有较强的故障识别能力。

Aiming at the non-stationary characteristics of transmission bearing fault vibration signals and the difficulty to extract the feature, the dimensionless parameters are not sensitive to be changed in the amplitude and frequency of vibration signal of as the charaeteristic param- eter,and the eharacteristic parameters were divided into self-model and non-self-model. An adaptive immune detector was designed to over- come the limitations of conventional diagnosis method based on the artificial immune principle. The adaptive mutation and suppression thresh- old were redefined. In the transmission bearing test rig, firstly, vibration acceleration information of bearing under the six states were meas- ured. Secondly, five dimensionless characteristic parameters were obtained by calculating the dimensionless index of vibration infurmation. And then, the detectors generation by utilizing those feature parameters were matured to train the proposed detector. Finally, the five bearing faults were diagnosed by mature detectors. The experimental results indicate that, the adaptive immune detector presented to diagnose trans- mission bearing faults is feasible and effective.