Abstract: The single-phase grounding faults of active distribution network are usually weak and complicated, easily contributing to the distortion of fault information collected at the measurement point. Therefore, it is difficult to position a fault locally. Taking into consideration the application of intelligent algorithms in the location of inter-phases short circuit faults in distribution network, a grounding fault location method of active distribution network based on wide area distribution of transient zero-sequence current is proposed, which is later proved less affected by information distortion and has more accurate location. The transient zero sequence current direction is extracted by the Closing Opening Difference Operation (CODO) calculation based on the zero sequence current signals collected at the feeder measurement points when there is a ground fault in the active distribution network. Given the distribution characteristics of transient zero sequence current and the application of shortest path theory in the graph theory, the suspected fault area is determined, reducing the size of solution space for fault location. Then a new fitness function for fault location is proposed, which makes the proposed method more sensitive to the distorted information. An improved binary ocean predator algorithm (IBMPA) is used to find out the fault feeders in the solution space, and the chaotic mapping, adaptive t-distribution, and group learning strategies are used to improve its ability to jump out of the local optimization and the global searching speed, succeeding in accurate fault location. Finally, simulation tests are carried out on the IEEE119 active distribution network. The results indicate that this method improves the accuracy and speed of grounding fault location in the distribution network effectively. Meanwhile, it has a higher fault tolerance for the distortion information.