Abstract: The secondary circuit of substation and the operating state of system have certain uncertainties caused by the equipment itself and physical connection,and it is difficult to locate the fault in the secondary system. This paper proposes a fault location method for the secondary system in smart substations based on connection state location. Firstly,the topological connection relationship of the secondary system equipment is abstracted into a matrix description,and the state parameters of the secondary system are used as the corresponding matrix elements. Then the secondary system connection state is located with a matrix algorithm to determine the matrix elements related to the fault. After locating the connection state,the fuzzy radial basis（RBF）neural network is used to make the fault correlation matrix elements corresponding to the fault conditions in a fault set,and the fault search and location are carried out. The results of the fault example show that the proposed fault location method for the secondary system has high fault location accuracy.