Distributed optical fiber acoustic sensing (DAS) technology has a wide range of applications in the oil field, but its original DAS massive data need pre-processing and noise reduction processing to get more accurate reflection of the actual situation of the DAS data. In this paper, the principles of DAS data preprocessing methods, Fast Fourier transform and noise reduction filtering methods, de-trending filtering and Butterworth, Eastern Cape filtering are studied, and the DAS data processing method is established, combined with the example well test, the results of DAS data spectrum and energy spectrum extraction can be comparable to the commercial software Ariane extraction results. The successful test of this method provides an effective technical means for the application of distributed optical fiber acoustic sensing (Das) technology in fracturing monitoring and fluid production monitoring of oil and gas wells, the technology level of fracturing monitoring and fracturing effect evaluation of horizontal wells in our country has been improved, it fills the gap of distributed fiber optics for acoustic sensing (DAS) data interpretation in China.