Abstract: Single monitoring indicators in structural damage identification are often susceptible to environmental noise and measurement errors, leading to limitations in localization accuracy. To address this issue, this study proposes a data fusion approach based on full-mode shape vectors and their corresponding dynamic components to improve the reliability of damage detection. Specifically, displacement responses under moving loads are first collected via a sensor network arranged on the bridge deck. The proper orthogonal decomposition method is then applied to extract the full modal information of the structure. By integrating the modal shape vectors and dynamic components, a new damage index (DI) is constructed. To validate the effectiveness of the proposed method, a simply supported beam bridge model is established, and numerical simulations are conducted that systematically examine various damage scenarios, including different damage locations, severity levels, and loading velocities. The results indicate that DI can effectively identify damage locations under various scenarios, with an average localization accuracy exceeding 95%. It is capable of detecting damage with a severity of less than 10% and shows strong robustness the changes in moving velocity. This paper confirms the advantage of complete mode shape data fusion in damage identification, providing an effective tool for structural health monitoring of bridges.