Abstract: Aiming at the problems of low reliability, large calculation amount and complex algorithm implementation when a single-phase grounding fault occurs in the resonance grounding system, a distributed intelligent fault line selection method for distribution network based on cloud computing is proposed. First, by constructing an intelligent fault diagnosis cloud model of the distribution network, the large amount of data processing and complex line selection algorithms are pushed to the "cloud". Second, in the distributed intelligent line selection system, the zero sequence voltage mutation and 15% of the phase voltage changes are used as the start of the intelligent unit to improve the reliability of the start of the intelligent unit and the quickness of the line selection. Then, the characteristic frequency band containing the main characteristics of the fault current is discussed in detail, and the transient zero-sequence current collected in the regional intelligent unit is transformed into wavelets to select the characteristic frequency band. Furthermore, through the collaborative analysis between the distributed intelligent units, the signal amplitude and phase information on the characteristic frequency band are combined to select the fault line. Finally, it is verified by a calculation example. The results show that the proposed method is not affected by complex fault conditions It has low communication pressure but high line selection accuracy, which provides a reference for the realization of fast line selection for single-phase grounding faults in resonance grounding systems.