基于复杂性测度的轴承振动状态监控

Condition Monitoring for Vibration of Bearings Based on Complexity Measurement

针对轴承运动稳定性和早期故障振动状态监控,提出了基于复杂性测度的状态监控方法,通过直接度量信号波形复杂度实现轴承的状态监控和早期微弱故障预警。从理论角度对受迫Duffing系统和Logsitic映射的复杂性进行了度量,验证多值粗粒化Lempel-Ziv复杂度和近似熵对动力系统的周期运动和混沌运动定量度量能力;并从试验角度对轴承早期微弱故障进行了监控。研究表明,多值粗粒化Lempel-Ziv复杂度和近似熵可以实现周期运动和混沌运动的定量判定,对早期微弱故障有较强的预警能力,可作为轴承振动状态监控的参量。

In order to monitor motion stability and vibration condition of early fault for bearings,a condition monitoring method is proposed based on complexity measurement. The condition monitoring and early weak fault warning of the bearings are realized by measuring complexity of signal waveform directly. The complexity of forced Duffing system and Logistic mapping are measured theoretically,the quantitative measurement ability of multiple-valued coarse graining Lempel-Ziv complexity and approximate entropy to periodic and chaotic motion of power system is verified. The early weak fault of the bearings is monitored experimentally. The researches show that the multiple-valued coarse graining Lempel-Ziv complexity and approximate entropy are characterized by quantitative determination of periodic and chaotic motion and strong early warning ability for early weak fault,which are able to serve as parameters of condition monitoring for vibration of the bearings.