Output-only measurement-based parameter identification of dynamic systems subjected to random load processes

En savoir plus sur le livre

This work introduces a novel output-only measurement method for identifying modal parameters of structures under natural loads like wind, ocean waves, traffic, or human activity. It emphasizes the dynamic excitation caused by wind turbulence and wind-induced ocean waves, modeled as stationary Gaussian random processes. Unlike existing techniques that treat unmeasured loads as white noise, this approach incorporates statistical data from wind fluctuations near the structure, enhancing identification results and enabling the estimation of unmeasured load processes. The identification problem is addressed using the H-fractional spectral moment (H-FSM) decomposition of the transfer function, which effectively represents Gaussian random processes with known power spectral density (PSD) as outputs of linear fractional differential equations with white noise inputs. The method's efficiency and accuracy are enhanced through an alternative fractional operator, making it suitable for both short and long memory processes. The study compares widely used wind and ocean wave model spectra, providing closed-form H-FSMs for straightforward process simulation. A state-space representation of correlated Gaussian processes is developed, eliminating the need for PSD factorization or optimization. This model integrates with state-space parameter identification algorithms, such as the extended Kalman filter, to estimate stiffness and damping in sys