Aiming at the current challenges of enormous scale, complex structure, difficult control and frequent accidents of city gas high-pressure pipeline network, there are still three aspects of difficulties in the risk monitoring and control of China’s city gas high-pressure pipeline network, namely, rough data, shallow assessment, and lack of power. This paper proposes an intelligent management system for gas pipelines based on C/S model and J2EE enterprise-level framework, in which the failure warning models of gas leakage, Gaussian plume diffusion, and fire and explosion are established. And the Kalman filter algorithm improved by DS evidence theory is used for intelligent fusion of Multi-source data, analyzing and screening the unified adequate information on data types, extracting state characteristics, classifying warning levels, and developing an integrated and visualized pipeline remote diagnosis and warning platform. In the simulation of the intelligent management system of gas pipeline, when the wind speed is 1.5m/s in winter, the ground surface is a safe area within 12.15m of the gas pipeline. When the maximum wind speed is 10m/s, the upper limit distance of the gas leading to fire and explosion is only 2.43m, and the hazardous range of the gas pipeline jet fire is within 12.69m. Relying on the gas high-pressure pipeline network in L city for practical experiments and applications, it provides technical support and decision-making basis for the construction of intelligent pipeline network, comprehensively improves the risk control capability of city gas high-pressure pipeline network, and has reference significance for the risk control of national city gas high-pressure pipeline network.