基于EMD能量熵和相关向量机的燃机涡轮叶片故障诊断方法

摘要

应用温度这一特征数值对燃气轮机的涡轮叶片进行故障诊断及分析，建立叶片的分析模型，从而可以分析叶片的冷却效果。然后利用经验模态分解（ EMD）将非平稳的故障信号分解成若干个平稳信号，即固有模态函数（ IMF）之和。计算不同频段上信号的能量熵，将其作为特征参数。采用相关向量机（ RVM ）来进行进一步的故障诊断。仿真实验表明，本文所研究的方法能有效地进行燃气轮机故障诊断。%In this paper, temperature features are used for turbine blades faults diagnosis and analysis. The tempera⁃ture model of turbine blades are set up to analyze the cooling efficiency of turbine blades. Then the empirical mode decomposition( EMD) method is applied to decompose the non⁃stationary fault signals into several stationary sig⁃nals, namely the sum of intrinsic mode functions. The feature parameters are obtained by calculating different fre⁃quency band energy entropy. We can also use the relevant vector machine to diagnose the faults of turbine blades. Experimental results show the proposed method is an effective method for turbine blade diagnosis.