Operational modal damping identification based on compressive sensing

Abstract Modal damping is a crucial parameter for structural condition and damage assessment. The modal identification using incomplete measurements can be realized by techniques based on Compressive Sensing (CS), which can reduce the amount of data for transmission and improve computational efficiency. However, techniques, such as sparse decomposition with prior information (SDPI), may hardly realize modal damping identification with satisfied precision under ambient vibration. To improve both the efficiency and precision, a novel CS-based modal damping identification method is presented, called Damping Identification by Sparse Decomposition (DISD). A random compression sampling scheme, random sampling with subsamples, is proposed to realize DISD. To enhance the damping identification accuracy, a new formula is proposed. Studies on numerical examples and real monitoring data from the Tsing Ma bridge were conducted for verification. The effectiveness of the presented DISD method is compared with that of the CS-based and traditional methods. Both the numerical and the real-monitoring cases show that DISD has improved modal damping results. Therefore, the proposed method is an efficient operational modal analysis tool and has the potential to be adopted in civil engineering practice.