In order to solve the problem that the traditional model is difficult to accurately simulate the urban rail traction power supply system with time-varying characteristics

the shortcomings of traditional modeling in the aspects of preset load

fixed parameters and off-line calculation were summarized and analyzed

and a hybrid driven modeling method based on digital twin technology was proposed in this paper. Combined with the actual characteristics of urban rail traction power supply system

the operation structure and calculation algorithm of digital twin model were designed. An information interaction system was established to input the load of the system model accurately. The closed-loop correction algorithm based on swarm intelligence optimization was adopted to realize the active correction of model parameters. Theoretical analysis and simulation results show that the global optimization correction of system parameters can be realized by the closed-loop correction algorithm based on the swarm intelligence optimization. The hybrid correction strategy of subsystem parameter identification and global parameter correction helps to reduce the optimization range of parameters

which will reduce the calculation time cost of the swarm intelligence optimization in system parameter correction and ensure the response speed of the model to parameter changes. The power flow calculation results of the digital twin model with the ability of accurate load input and active parameter correction are highly close to the sampling data. Compared with the traditional model

the accuracy of power flow calculation is greatly improved

which has practical significance for the operation decision and design optimization of traction power supply system.