Abstract: There are many kinds of transmission line faults, and the traditional diagnosis method focuses on fault location, which is difficult to realize the accurate differentiation of multiple types of faults. So this paper proposes a fault diagnosis method based on adaptive weighted feature fusion and mantis optimized fuzzy decision-tree (DBO-FDT). Firstly, the time complexity, nonlinear and amplitude change characteristics of the fault voltage current signal are extracted by arrangement entropy, fuzzy entropy and envelope entropy according to the entropy of the entropy. Secondly, the FDT classifier, where the fuzzy membership parameter of FDT is optimized by DBO to avoid the subgoodness of manual setting; finally, the algorithm is verified in the fault simulation model of 110 kV transmission line. The simulation results show that the proposed method can achieve reliable diagnosis for 10 types of faults such as single phase grounding, two phase short circuit, two phase short circuit grounding and three phase short circuit, and the average identification accuracy is up to 99%. The identification accuracy was improved by at least 2.9% compared with the single feature diagnosis method. At least 3.9% better than the existing machine learning algorithms such as FDT, PSO-FDT, GA-FDT) and DT.