Identifying dissolved gas characteristic parameters in transformer oil based on trapezoidal model

Dissolved gas analysis(DGA) in oil is one of the most effective methods for diagnosing internal insulation faults in transformers. Transformer oil possesses stable chemical properties at high temperatures and is primarily used for insulation and cooling in power equipment, requiring high dielectric strength and low viscosity. However, during prolonged operation, factors such as oxidation, contamination, and high temperatures can cause degradation and decomposition of transformer oil, leading to insulation defects. If not detected and repaired promptly, these issues can result in failures. This paper employs the identification of dissolved gas characteristic parameters in transformer oil based on a trapezoidal model to enhance the accuracy of transformer fault discrimination based on dissolved gas analysis. The approach was tested using dissolved gas data from transformers in a specific region in Northeast China, yielding the expected results. After more than 25 iterations, the accuracy of transformer fault detection reached 78%-80%.