Abstract: Open-circuit faults of IGBT switching transistors in photovoltaic grid-connected inverters can lead to deterioration of power quality, reduction in system efficiency, and seriously threaten the stability of grid connection. Aiming at the problems of insufficient generalization ability and low feature discrimination of existing diagnostic methods under dynamic photovoltaic operating conditions, an open-circuit fault diagnosis method for switching transistors based on random forest is proposed. Firstly, the three-phase current signals output by the inverter are collected, and after normalization preprocessing, an 18-dimensional feature vector including time-domain features (root mean square, peak value, peak-to-peak value, mean value) and frequency-domain features (fundamental wave amplitude, total harmonic distortion) is extracted to comprehensively characterize fault characteristics. Subsequently, a random forest diagnostic model is constructed, the robustness of the model is improved through Bootstrap resampling and random feature subspace strategies, and key parameters such as the number of decision trees and maximum depth are optimized using the grid search method. Finally, a simulation model is built on the MATLAB/Simulink platform, and the effectiveness of the method is verified by experiments. The results show that the diagnostic accuracy of this method for 21 types of switching transistor open-circuit faults and normal operating conditions reaches 100%, and the single-sample diagnosis time is only 0.08 ms. Compared with traditional methods such as BP (Back Propagation) neural network, SVM (Support Vector Machine), and KNN (K-Nearest Neighbor), it has stronger anti-interference ability and real-time performance, and can meet the engineering requirements of online fault diagnosis for photovoltaic grid-connected inverters.