Time delay estimation of furnace tube leakage acoustic signal based on generalized cross correlation

Time delay estimation (TDE) is a critical step in accurately locating boiler tube leakage. To deal with the problem of selecting the weighting function of generalized cross correlation (GCC) in boiler environment, the gas leak detection platform is built in the laboratory, the characteristics of the gas leak acoustic signal are analyzed, and the TDE performance of different weighting functions is compared between laboratory environment and simulated boiler environment in this paper. The results show that Roth, Phase Transform (PHAT), Smoothed Coherence Transform (SCOT) and Hassab Boucher (HB) outperform Cross Correlation (CC), Maximum Likelihood (ML) and Wiener Processor (WP) in the indoor environment. For the boiler environment, the TDE of SCOT weighting function is the most accurate when the signal-to-noise ratio (SNR) is -1 OdB to 15 dB. The TDE of PH AT and HB weighting functions is accurate when the SNR is greater than 0 dB, and the TDE of Roth weighting function is accurate when the SNR is greater than 10 dB. These findings provide a reference for selecting the weighting function for TDE of furnace tube gas leakage in the boiler environment.