Abstract: This paper focuses on the inversion of ladder network parameters for transformer windings in the context of diagnosing winding deformations. The goal is to utilize parameter identification algorithms and measurable frequency response analysis (FRA) data under high-frequency excitation. To begin with, the paper determines the test arrangement for the driving-point admittance (DPA) function, which is crucial for accurate modeling and achieving the simplest network topology. The combination of Genetic Algorithm and Iteration Algorithm is then employed to identify the network parameters. The network mathematical model based on the U-I relationship, and the feasible region of network parameters derived using the winding equivalent inductance and capacitance components, are systematically established. The paper also determines the high frequency (HF) region of the DPA data and evaluates the computation efficiency and accuracy of the proposed method. The obtained network parameters for an actual transformer meet all relevant constraints. Furthermore, the corresponding DPA curve within the frequency range of 800kHz~2MHz exhibits a high degree of fitting with the measured data. The objective function is significantly reduced from 442.11 to 90.44, indicating a descent rate of 79.54%. These results validate the effectiveness and precision of the proposed method.