Group Sparse Optimization Based-Compressive Sensing of Vibration Data Using Wireless Sensors for Structural Health Monitoring

For most of vibration signals of civil infrastructures have sparse characteristic, namely, only a few modes contribute to the vibration of the structures. Additionally, the measured vibration data by the sensors placed on different locations of structure almost has same sparse structure in the frequency domain. Based on this group sparsity of the vibration data of structure, the group sparse optimization based compressive sensing (CS) for wireless sensors are proposed. Different from the Nyquist sampling theorem, the data is first acquired by non-uniform low rate random sampling method according to the CS theory. Then, the group sparse optimization algorithm is developed to reconstruction the original data from incomplete measurements. The field tests on Xiamen Haicang Bridge with wireless sensors are carried out to illustrate the ability of the proposed approach. The results show that even using 10% random sampling data, the original data can be reconstructed by the proposed group sparse optimization method with small reconstruction error.