Aiming at the problem that it is difficult to detect the slight inter-turn short circuit fault of transformer winding

a method of inter-turn short circuit fault diagnosis based on the characteristic curve of RSO（ Repetitive Surge Oscillograph） is proposed. The propagation process of the pulse signal in the transformer winding is analyzed

and the distribution parameter circuit model of the transformer winding reflecting the propagation process is established. The variation law of wave impedance after inter-turn short circuit is derived. By inputting a low-voltage pulse at one end of the transformer winding

the response characteristic curve can be collected at the other end

and the characteristic curve is obtained by combining the two response characteristic curves before and after the inter-turn short circuit. Whether the inter-turn short circuit fault occurs or not is determined by the characteristic curve is raised or not. The characteristic curve’s protrusion degree reflects the severity of the intern-turn short circuit fault.Combined with artificial neural network algorithm

the fault identification rate of the proposed method is analyzed

and the simulation sample analysis shows that fault identification rate of the proposed method is up to 95%. Simulation and case analysis show the feasibility and accuracy of RSO method for diagnosing transformer winding inter-turn short circuit fault.