Continuous structural health monitoring with a modified dual Kalman filter applied with optical fiber sensing data

The advent of new sensing technologies has led to an increased use of sensors for the monitoring of systems and structures. By integrating data from these sensors with numerical models, a customized digital twin of the struc ture can be constructed and updated, enabling damage detection, localization, and control, possibly in real-time. In this context, the present study utilizes real experimental data from distributed optical fiber sensors within a modified dual Kalman filter (MDKF) algorithm for damage identification. The MDKF enhances the observation metric of the classical Kalman filter by employing the modified constitutive relation error (mCRE), a powerful identification functional with a strong physical sense that is effective even in the presence of noisy measure ments. This results in a robust and bias-aware method, coupling variational and sequential approaches, essential for developing effective continuous health monitoring on engineering structures. In the paper, the algorithm is applied to identify on-the-fly damaged zones in a beam subjected to a four-point bending test and equipped with distributed optical fiber sensors. The results are compared with those obtained using the mCRE method alone and are validated by means of Digital Image Correlation (DIC) data.