Evaluation of Transformer Health State in Distribution Network Based on Data-Model Combination and Fuzzy Comprehensive Evaluation

This article presents a methodology for assessing the health status of transformers in distribution networks. The proposed approach integrates numerical modeling and fuzzy comprehensive evaluation in order to address the limitations of existing evaluation methods, such as high ambiguity and subjective weighting. To establish a comprehensive dataset, fault data and normal operation data of transformers are collected. Subsequently, specific evaluation indicators for transformers in distribution networks are selected. Membership degree models and fuzzy comprehensive evaluation weights are then determined. To enhance the accuracy of the model, an optimization objective function is formulated by incorporating the deviation between predicted and actual lifespans. This function is utilized to update the parameters of the membership degree model and the evaluation weights. Through reverse optimization using the dataset, the model parameters and evaluation weights are adjusted accordingly. Finally, the health status of the transformers is assessed by combining the model parameters with the optimized weights. The proposed methodology is further validated by analyzing dissolved gas data obtained from transformer oils in the distribution network. A comparative analysis with traditional three-ratio and fuzzy comprehensive evaluation method based on entropy weight method results demonstrates that this approach significantly enhances the objectivity and accuracy of health status assessment for transformers in distribution networks.