开展设备性能退化评估研究,是制定主动设备维护策略、降低设备维护费用的基础。在设备性能退化过程中,信号成份会逐渐复杂化。本文提出利用小波包熵监测信号的复杂性变化,从而为设备性能退化评估提供可靠的特征向量。为了研究性能退化过程中振动信号的小波包熵的变化规律,使用裂纹转子动力学模型模拟了转子裂纹逐渐增加的过程,并使用仿真数据计算了各个状态下的小波包能量熵和小波包奇异值熵值。分析结果表明,随着转子性能退化程度的加深,小波包熵值逐渐增加,且对于性能恶化的突变较为敏感。
In order to monitor the equipment performance degradation, the wavelet packet entropy is studied. Equipment signal component complexity will increase during the process of equipment performance degradation. Our purpose is to use wavelet packet entropy to monitor this change. Two kinds of wavelet packet entropy, namely wavelet packet energy entropy and wavelet packet singular value entropy, have been defined and studied in this paper. To determine the sensitivity of these two kinds of wavelet packet entropy to the signal component change, the simulation data sets of a cracked rotor model based on the simple hinge crack model is used. By the decreasing stiffness coefficient of the cracked rotor model, a series of rotor degradation data sets are obtained. Then the two kinds of wavelet packet entropy are calculated on each data set. The analysis results have shown that the wavelet packet entropy is sensitive to the signal change. By using wavelet packet entropy as the feature, the equipment performance degradation can be assessed accurately.