Abstract: The train control and service network requires that the train control services and passenger services be transmitted over the same network. In order to ensure the safe operation of the train, the passenger service traffic from and to passenger compartments must be monitored while the traffic characteristics of passenger services are a key basis for effective supervisory control. So, a LAN in the lab was used to simulate the train compartment environment, traffic data of several common passenger service applications was collected using WireShark and some data from the public traffic data set was also extracted as a supplement to build a traffic data set of passenger services. Related Python toolkits were used to fit the arrival interval of passenger service message to classic probability distribution models and the fitting experimental results were compared and analyzed. Based on this, mathematical models of the arrival interval of passenger service messages were established and the credibility of the models was verified by calculating the evaluation indicators. The research results are of great significance to the planning and design, traffic supervisory control and system simulation of the train control and passenger service network.