A Novel Fault Diagnosis Method for a Power Transformer Based on Multi-Scale Approximate Entropy and Optimized...

Area of Science:

• Electrical Engineering
• Materials Science
• Data Science

Background:

• Traditional transformer fault diagnosis methods struggle with limited gas characteristic components and high misjudgment rates.
• Oil-immersed transformers require reliable methods for prompt fault detection to prevent failures.

Purpose of the Study:

• To develop an advanced transformer fault diagnosis model.
• To improve the accuracy and efficiency of detecting faults in oil-immersed transformers.

Main Methods:

• Proposed a transformer fault diagnosis model integrating multi-scale approximate entropy and optimized convolutional neural networks (CNNs).
• Utilized an improved sparrow search algorithm (ISSA) to optimize CNN parameters, creating the ISSA-CNN model.
• Analyzed dissolved gas components and calculated multi-scale approximate entropy for different fault modes, using entropy values as input features.

Main Results:

• Multi-scale approximate entropy effectively characterizes dissolved gas components in transformer oil.
• The proposed ISSA-CNN model demonstrated significantly improved diagnostic efficiency.
• Comparative analysis confirmed the superiority of the ISSA-CNN model over BPNN, ELM, and standard CNNs.

Conclusions:

• The ISSA-CNN model offers a superior approach for transformer fault diagnosis.
• The integration of multi-scale approximate entropy enhances the characterization of dissolved gases for fault detection.
• This method provides a more accurate and efficient solution for maintaining the reliability of oil-immersed transformers.