Noise reduction for oil pipeline pressure time series based on wavelet filtering technology

Since the pressure fluctuation signals are masked by noise and can not be detected, the oil pipeline leakage detection method is not satisfactory on small leakage problem, without more accurate warning and location. This paper uses discrete wavelet transform (DWT), lifting wavelet transform (lifting WT) and undecimated wavelet transform (UWT) to denoise the noisy signals collected in the pipeline leakage scene, and analyzes the simulation results respectively. The results show that the signal denoising methods based on lifting WT and UWT are much more effective than that on DWT. Through the conclusion, it is easier to find and locate the singular point while applying lifting WT and UWT.