The dynamic temperature information of transformer windings can effectively measure their load capacity and service life. Therefore

this paper investigates the evolution law and deduction calculation model of the dynamic hot-spot temperature rise in oil-immersed transformer windings. First

based on multi-physics coupling simulations involving electromagnetic

fluid

and thermal fields

the steady-state temperature distribution of transformer windings under different load rates was obtained

and the validity of the simulation model was verified through temperature-rise tests. Then

the evolution law of the dynamic hot-spot temperature rise was analyzed by simulation

and the relationship between the dynamic hot-spot temperature rise and the load rate was studied. Finally

the differential evolution algorithm was employed to predict the temperature-rise curve

and a deduction calculation model for deducing the hot-spot temperature rise was proposed. The results show that the maximum error between the multi-physics coupling temperature simulation results and the measured results is less than 5%. The hot-spot temperature-rise value increases with the duration of the load current application following a saturated exponential function

while the steady-state temperature-rise value increases with the load rate according to a power function. The dynamic temperature-rise time constant decreases with the load rate following a power function. The maximum error percentage of the steady-state hot-spot temperature-rise value obtained using the proposed deduction calculation model is only 2.83%

and the maximum error percentage of the temperature-rise time constant is only 4.76%. The research findings provide a feasible method for deducing temperature-rise data under high load rates from low load rate temperature-rise tests

offering a technical reference for addressing the challenges of conducting on-site transformer temperature-rise tests

which are difficult to implement and time-consuming.