Structures are susceptible to external impacts under long-term service, resulting in various types of damage. Online accurate assessment of the severity of damage is the basis for formulating subsequent maintenance and reinforcement plans. In this work, an online damage identification method based on the Adaptive Extended Kalman Filter (AEKF) is proposed. Initially, the vibration signals of a concrete-filled steel tubular (CFST) test structure subject to multiple lateral impacts are processed, and signals before and after damage inception are spliced to track damage evolution. Subsequently, the natural frequencies extracted from the signals before and after damage inception, and the amplitude of the damage itself are integrated into the state vector, to build a nonlinear state transfer and observation model and allow estimation of the dynamic flexural stiffness of the structure. To further improve the problem solution in the presence of signal losses caused by detachment or breakage of the sensors when damage occurs, the reconstruction of missing signals is accomplished by way of the weighted Matrix Pencil (MP), which ensures the continuity and stability of the AEKF filtering process. By comparing the results with the real damage state, the proposed method is shown to effectively track the gradual reduction of the flexural stiffness, and verifies the feasibility of the proposed method to provide a reliable support for online monitoring and damage assessment.