The result of fault location is affected by the type of fault. In order to locate fault accurately, it’s necessary to identify the type of the fault and then match the corresponding location algorithm. First, the correlation dimension in space domain is proposed to describe the complexity of the change of electrical quantity after fault, and the difference square root in the time domain is proposed to describe the degree of distortion of electric quantity after fault. Then, the entropy weight method is used to fuse the two indexes, according to the amount of information contained in the two features. Finally, the fault type is recognized based on the fusion feature, the corresponding location algorithm is matched for different types of faults to minimize the impact of fault impedance. The simulation shows that the proposed adaptive fault location method can acquire the result accurately, and is of universal applicability. This work is supported by Outstanding Young Talents Project of Jiangxi Province (No. 20162BCB23046), The Key R & D Projects of Jiangxi Province (No. 20161BBH80033), Research and Development Project of Jiangxi Education Department (No. GJJ160498), and Science and Technology Support Program of Nanchang (No. 2013HZCG013).