A New Method for Detecting and Evaluating the Insulation State of Transformers Based on Frequency Response Analysis

The insulation status of transformers is crucial for the safety of power systems. As the operation time of transformers increases, the aging of insulation materials will affect their performance. Therefore, evaluating the insulation status of transformers is of great significance. In this paper, the transformer's hybrid multi-conductor transmission line model is studied, and the transformer is simulated by using resistance, inductance and capacitance elements to reflect the changes of its geometric structure and insulation state. It is pointed out that the component parameters in this model are closely related to insulation aging, so this paper proposes a method to extract the model component parameters and evaluate the insulation state of transformer by using Frequency Response Analysis (FRA). In this paper, the parameter extraction problem is transformed into an optimization problem, and Particle Swarm Optimization algorithm (PSO) is introduced to solve this problem. The results show that the model proposed in this paper can fit two curves and accurately extract the model parameter values. The research in this paper provides theoretical and practical support for evaluating transformer insulation state through frequency response curve.