A Multilevel Diagnosis Method for Large Transformer Fault Modes under Multi-sensor Information Fusion

The conventional multi-level diagnosis method for fault mode of large transformers mainly uses INGO (Impacted Northern Goshak Optimization) random differential perturbation to improve the Northern Goshawk algorithm to build a quasi segment model, which is vulnerable to the influence of NGO search and update behavior, resulting in abnormal diagnosis alarms. Therefore, a multi-level diagnosis method for fault mode of large transformers under multi-sensor information fusion is proposed. That is to say, using multi-sensor information fusion to build a multi-level diagnosis framework of large-scale transformer fault mode, and design a multi-level diagnosis algorithm of large-scale transformer fault mode, thus completing the multi-level diagnosis of large-scale transformer fault mode. The experimental results show that the designed fault mode multi-level diagnosis method for large transformers can effectively complete the alarm at different diagnosis times, and each diagnostic index is good, which proves that the designed fault mode multi-level diagnosis method has good diagnosis effect, reliability, and certain application value, and has made certain contributions to reducing the operation safety risk of large transformers.