The research on the temperature field of the transformer winding under dynamic variable load is of great significance for ensuring the safe operation of power systems. This paper takes an SSP-360000/500 transformer as the research object, establishes a 2-D simulation model, and uses the finite volume method to analyze the high voltage winding and low voltage winding temperature distribution of the transformer under dynamic variable load. The simulation calculation results have been fully verified by the experimental data to make the successful prediction of the overall temperature and hot spot temperature position of forced oil circulation transformers with a guided structure. The results show that the most significant temperature raise occurs at the secondary end of the winding. In the case of dynamic variable load. The temperature raise of the winding becomes larger as the load increases, but before the rated load is in the stable temperature rise range, it can run safely for a long time. However, during overload operation, the average temperature raise of the high voltage winding may exceed its limit, the insulation material is damaged.
The end-winding insulation structure of stator windings in pumped-storage generator units is complex, with pronounced electric field concentration under out-of-phase conditions, making them critical concerns in insulation design and condition-based maintenance. Although the finite element method (FEM) offers reliable accuracy, the strong nonlinearity of the anti-corona layer results in a computation time exceeding 104 seconds per single solution, rendering it impractical for parameter optimization and rapid on-site assessment. This paper proposes a fast prediction method for end-region potential distribution based on a deep neural network (DNN). Taking a 334 MW unit as the research object, a three-dimensional electroquasistatic finite element model with six stator coils is established and validated through power-frequency withstand voltage and ultraviolet imaging experiments. Training samples are generated via design of experiments (DoE), and a multilayer DNN surrogate model with a 7-dimensional input (comprising 3D spatial coordinates and four physical parameters) and a 1-dimensional output is constructed to directly reconstruct the spatial potential field at the end region. The results demonstrate that the surrogate model achieves a maximum relative error of less than 2% along the entire path compared with the high-fidelity FEM solutions, with a single prediction time of approximately 38 s—representing a speedup factor of approximately 272—while also exhibiting good generalization capability. This method provides a feasible technical approach for rapid reconstruction of end-region field distribution and optimization of insulation structures.