The primary and secondary fusion equipment ignores the extraction of cable impulse current characteristics, resulting in large error in the identification results of ground fault and low identification accuracy. This paper proposes a ground fault identification method for the primary and secondary fusion equipment based on the consideration of three-phase asymmetric load. Firstly, calculate the three-phase asymmetric load, determine the ground fault identification node of the primary and secondary fusion complete equipment, establish the ground fault identification tree according to the identification node results, and build the artificial intelligent ground fault identification model, thus completing the simple fault diagnosis; Then the method is optimized, that is, the ground fault of the primary and secondary fusion equipment is finely classified, and the node characteristic wave is selected to realize the ground fault identification of the primary and secondary fusion equipment. The experimental results show that after optimization, the error result of this method reaches a stable state, the recognition effect is good, the error value is at least 0.08, the detection efficiency is high, and the maximum can reach 97%, and the diagnosis result is the same as the expected output result.