The small current grounding fault line selection due to the low accuracy of line selection and fault line selection problem for a longer time, directly affects the stable operation of the power distribution network, aiming at this problem, in this paper, using big data intelligent analysis algorithm, using the historical data of the small current grounding fault, the depth of convolution network structure, the statistical properties of the application of network model and parameters for training; Heuristic optimization algorithm is used to accelerate the convergence of the algorithm and reduce the line selection time. According to the established intelligent network model and the actual measurement history data of a distribution system in Shanxi province, the small current grounding line selection system can quickly and accurately select the line successfully, thus enhancing the effect of practical application of the system.