| 周 良1 ，王华伟 1 ，许珊珊 2 ，王清薇 1 ，王颖1.基于滑油磨粒聚类分析的航空发动机故障诊断[J].航空发动机,2024,50(6):113-119 |
| 基于滑油磨粒聚类分析的航空发动机故障诊断 |
| Aeroengine Fault Diagnosis Based on Cluster Analysis of Lubricant Wear Particles |
| |
| DOI： |
| 中文关键词: 聚类识别 故障诊断 状态监测 粒子群算法 支持向量机 航空发动机 |
| 英文关键词:identification clustering fault diagnosis condition monitoring particle swarm optimization support vector machine aeroengine |
| 基金项目:国家自然科学基金（U1833110）资助 |
|
| 摘要点击次数: 130 |
| 全文下载次数: 75 |
| 中文摘要: |
| 为了有效分析滑油状态，实现基于滑油状态分析的航空发动机关键部件磨损状况监测，开展了基于滑油磨粒聚类分析
的航空发动机故障诊断研究。利用主成分分析算法提取滑油磨粒的关键特征参数；结合关键特征参数，利用支持向量机模型进行
滑油磨粒的聚类识别。并通过对改进粒子群算法进行分析，利用自适应权重粒子群算法和异步收缩因子粒子群算法对支持向量
机模型进行优化，以提高聚类的准确性。以滑油光谱数据为例，进行了滑油磨粒聚类试验。试验结果表明：通过2种改进粒子群
优化的支持向量机模型仿真，可实现航空发动机滑油中磨粒的聚类识别，识别精度高达99.7%，且监测和诊断结果与实际故障一
致。研究结果可为航空发动机的滑油状态监测和故障诊断提供依据。 |
| 英文摘要: |
| In order to effectively analyze the lubricating oil status and realize the wear status monitoring of key components of
aeroengine based on lubricating oil status analysis, aeroengine fault diagnosis research based on lubricating oil abrasive particle clustering
analysis was carried out. The principal component analysis algorithm is used to extract the key feature parameters of lubricating oil abrasive
particles. Combined with the key feature parameters, the support vector machine model is used to perform cluster identification of lubricat?
ing oil abrasive particles. By analyzing the improved particle swarm algorithm, the support vector machine model is optimized using the
adaptive weight particle swarm algorithm and the asynchronous shrinkage factor particle swarm algorithm to improve the accuracy of
clustering. Taking lubricating oil spectral data as an example, a lubricating oil abrasive particle clustering test was conducted. The test
results show that through two improved particle swarm optimization support vector machine model simulations, the cluster identification of
abrasive particles in aeroengine lubricating oil can be achieved, with an identification accuracy as high as 99.7%, and the monitoring and
diagnosis results are consistent with actual faults. The results provide a basis for oil condition monitoring and fault diagnosis of aeroengine. |
| 查看全文 查看/发表评论 下载PDF阅读器 |