LAUSR

Accurate aging state evaluation of traction transformer insulation at the hotspot region is crucial for ensuring the stable operation of the railway traffic system. Given this issue, this research begins by collecting the polarization currents of oil-paper insulation. Then, the modified dielectric response model is proposed to simulate the collected polarization currents. To extract the dielectric feature parameters related to the insulating paper’s aging state, the cooperative co-evolutionary algorithm (CCEA) modified by niche mechanisms (NMs) is proposed. During the iteration process, the inversion computation technique is devised to reduce the number of feature parameters to be optimized. Futhermore, the dielectric feature parameters database is established by the surface fitting technique, Finally, the database is employed to evaluate the aging state of hotspot insulating paper under both laboratory conditions and field transformers and correct the influence of testing temperature. The relative errors between the calculated and measured DP in verification experiments are under 4%, and the corresponding aging state labels are exactly the same. In this context, the study is anticipated to provide a precise assessment of the aging degree of traction transformer insulation at the hotspot region.