多尺度能量熵与优化极限学习机的航空液压管路故障诊断方法

针对航空液压管路故障特征难以提取问题,考虑到航空液压系统中振动信号存在非平稳性以及非线性等特点,提出了一种基于局部均值分解(Local Mean Decomposition,LMD)的多尺度能量熵(Multi-scale Energy Entropy,MEE)和麻雀搜索算法(Sparrow Search Algorithm,SSA)优化极限学习机(Extreme Learning Machine,ELM)的航空液压管路故障诊断方法。首先,采用局域均值分解方法将采集的振动信号自适应分解;其次,综合考虑相关系数-能量比准则,选取最佳PF分量;最后,计算最佳分量的多尺度能量熵,选取合适的尺度因子并将其对应的能量熵值作为特征向量,输入到麻雀搜索算法优化的极限学习机网络模型进行学习训练,实现对航空液压管路的故障进行分类识别。结果表明:该方法能够有效地实现对航空液压管路故障类型的准确识别,为区分航空液压管路故障提供了一种可行的诊断思路。

Fault Diagnosis Method of Aviation Hydraulic Pipeline Based on Multi-scale Energy Entropy and Optimized Extreme Learning Machine

A fault diagnosis approach of aviation hydraulic pipeline based on the combination of Sparrow Search Algorithm (SSA) for Optimizing extreme learning machine (ELM) and multi-scale energy entropy (MEE) of local mean decomposition (LMD), which considers the non-stationary and nonlinear characteristics of vibration signals in the aviation hydraulic system is proposed to solve the problem of fault feature extraction of aviation hydraulic pipeline. Firstly, the local mean decomposition method is adopted to decompose adaptively the collected vibration signals. Secondly, the best PF component was selected according to the correlation coefficient-energy ratio criterion. Finally, calculating the optimal amount of multi-scale energy entropy, and select the appropriate scale factor and its corresponding energy entropy as a feature vector the input to the extreme learning machine model to study and implement of aviation hydraulic line fault classification recognition. The results show that this method can effectively realize the accurate identification of the fault types of aviation hydraulic pipeline, and provides a feasible diagnostic idea for distinguishing the fault types of aviation hydraulic pipeline.