【摘要】 对涡轮增压器压气机喘振在线检测进行了研究，利用压气机出口动态压力信号，在时域和频域分析了稳态工况到喘振工况过程的特征；通过小波变换模极大值和李普希兹指数判别信号中的突变点，同时对信号进行小波滤波，在此基础上采用一种改进方差法实时检测喘振。结果表明，该方法可在半个喘振周期内检测出压气机的喘振现象，与方差法相比信噪比提高了20.5%,且对不同型号或不同工况下的涡轮增压器压气机喘振检测具有良好的通用性。更多还原

【Abstract】 The on-line detection of turbocharger compressor surge was studied. The characteristics of steady condition to surge condition were analyzed in time and frequency domain based on the dynamic pressure signal of compressor outlet. The mutation points in the signal were identified by using the wavelet transform modulus maxima and Lipschitz exponent, and the wavelet filter of signal and then the real-time detection of surge with an improved variance method were conducted. The results show that the method can detect the surge phenomenon of compressor in the half of surge period and the signal-to-noise ratio increases by 20.5% compared with the variance method. Accordingly, the method has good universality for the detection of turbocharger compressor surge under different types or working conditions.更多还原