EMD-AR和GRNN算法下的航空液压泵多模态故障诊断分析  

Multimodal Fault Diagnosis Analysis of Aviation Hydraulic Pump Based on EMD-AR Model and GRNN Algorithm

作　　者：郭文军 张自来 陈丽君[1] GUO Wen-jun;ZHANG Zi-lai;CHEN Li-jun(Aviation Key Laboratory of Science and Technology, Aero Electromechanical System Integration, Nanjing Engineering Institute of Aircraft System, Nanjing, Jiangsu 210000)

机构地区：[1]南京机电液压工程研究中心航空机电系统综合航空科技重点实验室,江苏南京210000

出　　处：《液压与气动》2022年第6期151-158,共8页Chinese Hydraulics & Pneumatics

摘　　要：针对新一代飞机高综合化、高复杂度和高耦合性导致的传统推理故障诊断策略难以满足现代维修保障需求的问题,开展基于广义回归神经网络(Generalized Regression Neural Network,GRNN)的飞机液压泵智能化故障诊断研究。构建经验模态分解(Empirical Mode Decomposition,EMD)与自回归(Autoregressive,AR)相融合的深度特征提取方法,提升原始信号的隐层故障特征筛选能力;再将增强后的隐层特征与GRNN相结合进行神经网络训练,提升智能诊断模型的识别精度。实验结果表明,EMD-AR-GRNN智能诊断模型能快速、准确地诊断出液压泵各故障模态,对保障设备的安全运行,提升系统可靠性具有重要的意义。The new generation aircraft is developing towards high integration,high complexity and high coupling.As a result,the traditional inference-based fault diagnosis strategy is difficult to meet the modern maintenance support needs.Therefore,it is necessary to carry out research on the intelligent fault diagnosis of aviation hydraulic pumps based on the general regression neural network(GRNN).A fault feature extraction method based on the combination of empirical mode decomposition(EMD)and autoregressive(AR)model is adopted to enhance the hidden layer feature extraction ability of fault signals;then the hidden layer features after feature preprocessing are combined with GRNN for neural Network training to improve the recognition accuracy of the intelligent diagnosis model.The experimental results prove that the proposed EMD-AR-GRNN intelligent diagnosis model can quickly and accurately diagnose the failure modes of the hydraulic pump,which is of great significance for ensuring the safe operation of the equipment and improving the reliability of the system.

关 键 词：EMD-AR 多模态 液压泵 故障诊断 神经网络 

分 类 号：TH137[机械工程—机械制造及自动化]