As the main equipment of the power system, the fault diagnosis of power transformers will directly affect the reliability of the entire power grid. Therefore, in most cases, traditional fault diagnosis methods based on a single data feature may encounter accuracy and stability issues when dealing with complex fault scenarios. This article proposes a fault diagnosis method for power transformers based on the fusion of recorded data and expert systems. This fault diagnosis method utilizes the expert's experience and knowledge, as well as the multi data feature fault characteristics contained in the recorded data, to accurately diagnose faults through intelligent fusion technology. This article designs relevant experiments to verify the performance of the proposed fault diagnosis method, and compares the differences in system stability between traditional fault diagnosis methods and diagnostic methods based on recorded data and expert system fusion. The experimental results show that the fault diagnosis method proposed in this paper is superior to traditional fault diagnosis methods based on single data features when comparing system stability. In the experiment, the average system stability of the experimental group was about 99.3%, while the average system stability of the control group was about 96.5%.