Abstract
Two damage identification techniques for structures based on change in modal data and frequency response functions (FRFs) are presented. A learning coefficient was introduced to prevent the solution from divergence. The technique aims to identify damage only from single input and single output (SISO) data using single sensor and single actuator. To save time and labor of experiments, we present an idea of using modal data easily measurable, and changing the excitation frequencies at wide range instead of changing the position of devices. Through numerical simulations, two techniques were both verified. It is confirmed that combining both data improves the identification accuracy even if measured modal data.
