海军航空大学学报2025，Vol.40Issue(5)：697-705,716,10.DOI:10.7682/j.issn.2097-1427.2025.05.005
基于改进VB粒子滤波的航空发动机气路故障诊断
Gas Path Fault Diagnosis for Aeroengine Based on Improved VB Particle Filter
摘要
Abstract
Aimed at the situation where the noise characteristics of an aeroengine are unknown under abnormal value in-terference,a particle filter method based on improved variational Bayes(Improved VB-PF)is proposed.This algorithm models the system process noise and measurement noise by mutually independent student's t-distributions to ensure the robustness of the algorithm.The unknown noise parameters are derived using variational Bayes,and the quality of sam-pling particles is improved by the particle swarm optimization algorithm to achieve the estimation of engine health pa-rameters.In the case of complex noise characteristics with abnormal value interference,the proposed method has en-hanced the accuracy and robustness of the fault diagnosis for the gas path of the aeroengine.关键词
航空发动机/气路分析/变分贝叶斯/粒子滤波/学生t分布Key words
aeroengine/path analysis/variational Bayes/particle filter/student's t distribution分类
航空航天引用本文复制引用
刘政,于兵强,魏芳,周鑫,黄金泉,鲁峰..基于改进VB粒子滤波的航空发动机气路故障诊断[J].海军航空大学学报,2025,40(5):697-705,716,10.基金项目
国家自然科学基金(52302472) （52302472）
空天动力基金项目(KDJJ20240201006) （KDJJ20240201006）
