摘要

In pipeline leak detection, interference and environment noise directly influence the effectiveness of leak detection and the accuracy of leakage location. It is essential to suppress noises in leak detection signals and improve the SNR of leak detection signals for ensuring reliable detection and accurate location. According to the fact that the autocorrelation lengths vary with the bandwidths of signals and noises, this paper proposes a new adaptive noise suppression method, which uses adaptive linear enhancer and predictor to suppress narrow band and wide band noises, respectively. In this method, no priori-knowledge is necessary for adaptively suppressing the noise that has different bandwidth with useful signals. After the proposed processing, the signal-to-noise ratios of the data are improved by more than 4 dB. After processing the actual leak detection signals, the leak location errors are reduced by more than 0.3 m in average; and the fault detection rate is also reduced.

  • 出版日期2012

全文