| It is one of the important factors to result in the power station to stop theboiler that boiler tube leaks. if can discover small leakages of the boiler and monitor the developing trend of leakages, we can arrange to shut down the boiler in advance, working well to maintain the preparation, shorten the time of maintain, canning reduce economic loss resulted from disarranged stopping of boilers. On basis of tallying up current technology of boiler leakage detection, studying acoustic characteristics of the leakage noise, analyzing background noise in normal running process and factors that affect the traveling of acoustic wave ,this paper putting forward to use the wavelet packet transform and spectral analysis approaches to handle the noise signal data, which has good localized features in both the time domain and the frequency domain , then the adoption distinguishes rate more the analysis the technique, for on time accurate the judgment leaks the condition provides the technique according to. |