Structural Monitoring Strategies of Buildings Using Inertial Measurement Units for Drift Ratio Estimation, Mode‐Shape Derivatives, and Damage Diagnosis

Structural health monitoring has evolved toward data‐rich frameworks that enable continuous, autonomous diagnosis of structural integrity. However, conventional vibration‐based methods, which rely solely on accelerometers, often struggle to infer displacement‐related indicators, for example, interstory drift ratios and exhibit limited sensitivity to distributed damage. To overcome these limitations, this study proposes inertial measurement unit (IMU)–based structural health monitoring strategies for estimating interstory drift ratios, reconstructing continuous mode‐shape derivatives, and diagnosing structural damage of buildings. First, the feasibility and accuracy of IMU‐based interstory drift estimation are experimentally evaluated by integrating measured angular velocity to obtain rotational responses under seismic excitation. Shaking table experiments on a five‐story model building demonstrate that IMUs can capture both translational and rotational responses with high accuracy, achieving a root‐mean‐square error below 0.04 m/s ² and less than 5% drift deviation compared with reference displacement sensors. Building upon this experimental validation, a modal reconstruction method is proposed to reconstruct continuous mode‐shape slopes and curvatures from sparse IMU measurements through polynomial fitting with structural boundary conditions to enhance the spatial resolution of modal features. The reconstructed derivatives are further decomposed into shear‐ and flexure‐dominant components and incorporated into an Integrated Strain Energy Damage Index (ISEDI). A sequential filtering strategy is then introduced to reduce interference among multiple damaged regions and improve damage localization and severity assessment. Verification through combined experimental and numerical studies on shear‐, flexure‐, and hybrid‐dominant structures confirms that the proposed framework achieves over 95% damage localization accuracy and robust drift estimation, and its effectiveness and generalizability for real‐world SHM applications are demonstrated.