Inter-turn short circuit fault in doubly-fed generator is a common electrical fault that directly affects the normal operation of the wind turbine and endangers the safety of the grid connection. This paper proposes an inter-turn short circuit fault identification technology based on multi-flux sensor data fusion. A simulation experiment platform for the inter-turn short circuit fault of a doubly-fed generator with a power of 100kW is used, consisting of a three-phase wound rotor induction machine, monitoring systems, frequency converters, and related control a drive motor. This experimental system can genuinely simulate the actual working conditions of the machine when the machine is short-circuited, and the experimental simulation can be closer to the operation process of the double-fed induction generator. The signal detection system is used to collect the external magnetic flux leakage signal of the motor. We place magnetic flux sensors in the axial and radial positions and extract the RMS value of each magnetic flux sensor as the characteristic value. Use DS evidence theory to fuse data from magnetic flux sensors to determine the operating status of the generator. This method can effectively detect the short circuit fault of rotor windings.