A model-free data-interpretation approach for long-term monitoring of bridges

This paper presents a new model-free data-interpretation approach for damage detection of bridges using long-term monitoring data. The approach combines two model-free methods: Moving Principal Component Analysis (MPCA) and Robust Regression Analysis (RRA). The objective is to integrate complementary advantages of these methods in order to improve performance in terms of damage detectability and time to detection. To illustrate the applicability of the proposed approach, a railway truss bridge in Zangenberg (Germany) is selected as a case study. The performance of the approach is compared with that of using MPCA and RRA alone. Results demonstrate that the combined approach performs better than MPCA and RRA in terms of damage detectability and time to detection.