Abstract: ［Objective］ Traction transformers typically employ a combination of forced oil circulation and fan-forced convection for heat dissipation. The cooling performance of this method significantly impacts the transformer′s lifespan, making it essential to study the heat dissipation process of forced oil and air-cooling in traction transformers in depth. ［Method］ Based on a one-dimensional temperature field assumption, a distributed parameter model is developed for the transformer and oil cooler components, then integrated with a centralized parameter model for the oil pump, pipelines, and other components, forming a complete system model. Numerical simulations of this model are conducted using Matlab′s Simulink toolkit to investigate the dynamic temperature behaviors of transformer winding and ester oil under actual alternating power conditions. Additionally, PID (proportional-integral-derivative) control technology is tentatively introduced into the transformer cooling system. To verify the model′s accuracy, the calculated results are compared with experimental values reported in the literature. ［Result & Conclusion］ Under typical dynamic operating conditions, the winding temperature of traction transformer exhibits a steady-state characteristics during low-load phases, whereas it rises rapidly over time during high-load phases. By employing the PID method to control the operation of oil cooler fan, the winding temperature can be maintained within the allowable high-temperature range, while the fan power consumption is reduced by approximately 13% compared to traditional on-off control strategies.