Aiming to improve the reliability of rotor fault diagnosis for induction motor, this paper proposed a method of rotor fault diagnosis for induction motor based on wavelet packet energy analysis and signal fusion. The frequency spectrum fusion of stator current signal and vibration signal is used as the diagnosis basis of rotor broken bar fault and air gap eccentric fault. First, the signal is decomposed by wavelet packet to obtain the corresponding energy distribution under different wavelet packet frequency band nodes and compare it with the normal motor signal. Then, the wavelet packet node reconstruction is carried out for the signal frequency band with abnormal energy. Finally, the fault characteristic frequency is identified by the fast Fourier transform to diagnose whether the motor fault occurs. Through simulation analysis, the effectiveness and practicability of the method are verified, and it is of great significance for the accurate monitoring of motor running state.