The invention discloses an intelligent fault analysis method based on CBTC system network data, and the method comprises the steps: employing a port mirroring technology in a switch cascade state, andachieving the obtaining of the CBTC system data; in the fault identification and alarm stage, if a query type fault alarm mode is adopted, sending a problem related to the current state to the CBTC system, and combining data fed back by the CBTC system to find a potential fault, and give out an alarm; and if an active fault alarm mode is adopted, extracting fault code information and auxiliary information generated by each subsystem logic part from the CBTC system data, determining a fault source in combination with a pre-defined fault label library, and giving alarm or prompt information. According to the method, on the basis of completing the real-time interface data monitoring of the CBTC system, the data is comprehensively analyzed, the fault is accurately positioned, and an alarm prompt is given, so that the integration and intelligence test level of the system is improved, the analysis time is reduced, and the labor intensity of test personnel is also reduced.