Abstract: Due to the poor diagnostic accuracy and high mean square error of existing diagnostic methods, this study focuses on the insulation fault diagnosis of power equipment in high-voltage transmission line systems based on an improved Stackelberg game model. Install sensors to collect real-time operational data of equipment and extract insulation fault characteristics of power equipment. Construct a Stackelberg game model, introduce a payoff function, and achieve the optimal matching of insulation fault diagnosis features for power equipment. Calculate the strategy update for normal and fault states based on the current best matching situation. By comparing with the threshold, the diagnosis of the device status is completed. The test results show that the diagnostic accuracy of the experimental group is 99%, indicating a high level of diagnostic precision; The mean square error of the experimental group is always less than 0.05%, which is the smallest value among the three groups, and has extremely high accuracy in diagnosing the insulation performance indicators of power equipment.