Since the traditional butterfly optimization algorithm may （BOA） have the problems of poor global search ability and slow convergence speed， while structural frequency is insensitive to the damage， a structural damage identification method based on improved butterfly optimization algorithm （IBOA） and wavelet packet energy curvature is proposed in the present study. The improved butterfly optimization algorithm is initially proposed by introducing clustering competitive learning mechanism and chaotic elite learning mechanism into the traditional butterfly optimization algorithm. With faster convergence speed and higher calculation accuracy， the proposed improved algorithm is able to achieve a good balance between local search and global search. In addition， the wavelet packet energy curvature is evaluated in the objective function to further improve the accuracy of identification results. The effectiveness of the proposed method is verified by numerical examples of simply-supported beam and 8-DOF spring-mass experiment. The results show that the proposed method can accurately identify the damage location and extent of the structure， even if the environmental noise and model error are considered.