Abstract: In view of the existing problems of poor signal feature extraction performance and high misdetection rate, the ultrasonic leakage detection method of coal mine gas drainage pipeline is proposed. According to the practical application scenario of coal mine gas drainage pipeline, the appropriate ultrasonic detector is selected and arranged, and on this basis, the ultrasonic signals are obtained in real time, and the temporal domain features, frequency domain features, time and frequency domain features and statistical characteristics are extracted. The support vector machine is used to determine whether there is leakage phenomenon in coal mine gas drainage pipeline, and the time difference method is used to locate the leakage position of coal mine gas drainage pipeline and quantify the gas leakage strength, so as to realize the effective detection of coal mine gas drainage pipeline leakage. The experimental data show that the skewness characteristics extracted by the design method are consistent with the actual characteristic data, and the minimum detection rate of gas leakage reaches 3%.