Abstract: Working environment in underground coal mine is harsh, and gas drainage pipelines are vulnerable to collision, coal falling and other injuries, resulting in gas leakage. When a large amount of air in tunnel enters pipe network system, the concentration of gas drainage in pipe network may be much lower than the concentration at the drilling hole. For the above problems, a self-renewing leakage detection technology of coal mine gas drainage pipe network system based on multi-Gaussian beam model was proposed. Multi-Gaussian beam model is adopted to strengthen the processing of the sound of the leakage points, main leakage types and noise sources of underground gas drainage pipe network system are analyzed, and leakage model and noise model are established.The collected sound samples are compared with pre-stored models to determine whether there is gas leakage, and the sound samples that occur more than 30% in the use environmen are automatically stored as gas leakage models to realize automatic model update and improve leakage detection accuracy. YJL40 leakage detector is developed based on self-renewing leakage detection technology, its main components include probe, metal hose, host and alarm. Self-renewing leakage detection technology and corresponding products are applied in high and low negative pressure drainage system in Gaojiazhuang Coal Mine for leakage detection of totals 7 585 m of pipelines, after the detected leakage points are effectively sealed, the gas volume fraction in drainage terminal is increased by 37.1% and 28% respectively, verifying the effectiveness of the self-renewing leakage detection technology.