基于信息熵的油液监测数据特征及磨损故障诊断研究

应用信息熵及模糊熵聚类算法对内燃机油液监测数据进行处理，得到了表征设备磨损状态的主特征量，并根据系统输出的数据序列间的Shannon互信息，较为简化地表征系统内部因素间的相互联系程度，提出了将互信息作为监测设备磨损状态变化的重要指标的监测方法，可以准确而敏锐地表征系统的变化，为应用光谱分析等方法监测设备磨损的状态、检测可能的故障及故障定位提供了一种新的分析手段。通过应用到实例分析和故障诊断中，验证了该方法的有效性。

Study on the Characteristic of Oil Monitoring Data and Wear Fault Diagnosis Based on Information Entropy

This paper presents the utilization of Shannon Entropy and fuzzy Entropy Clustering on oil monitoring data processing in internal combustion engine. Based on the results of the clusters, the main characteristic elements of spectral analysis are obtained. Through calculating the mutual information according to Shannon's Information Theory, it provides a new analyzing approach for the main wear components and a new measure to monitor wear state and to identify possible fault location. By analyzing two examples, the algorithm and effectiveness are validated for acquiring the oil monitoring information.