An intelligent fault diagnosis assistant for power transformer based on fine-tuning large language model

This paper presents a power transformer equipment fault diagnosis assistant based on fine tuning large language model (LLM), aiming at improving the efficiency and accuracy of transformer fault diagnosis through natural language processing technology. By constructing the instruction data set in a specific field, the pre-training language model is supervised and fine-tuned (SFT), so that it can understand and generate professional diagnosis suggestions related to transformer faults. The experimental results show that this method shows high accuracy and practicability in the task of fault classification and diagnosis suggestion generation, and provides an intelligent auxiliary tool for power equipment operation and maintenance.