Online input, state, and response estimation for building structures under earthquakes using limited acceleration measurements

Tuning of Kalman Filter Noise Parameters for Uncertainty Quantification in Input-State Estimation of Mdof Systems

“…( 3) but with the direct feedthrough term J being set to zero. The details of this formulation can be found in [3,13].…”

Absolute Displacement-Based Formulation for Peak Inter-Story Drift Identification of Shear Structures Using Only One Accelerometer

“…These studies focused on the structural parameters or seismic wave, wherein the relative displacement can be obtained from the estimated absolute responses and input. On the other hand, the relative responses can be directly used to form the state vector based on the general equation of motion, which is formulated in a relative coordinate system, so that there is no direct feedthrough term in the state-space equation [ 23 , 24 , 25 ]. In addition, some researchers combined other responses data (e.g., displacement or strain) with acceleration responses to evaluate the structures [ 26 , 27 , 28 , 29 ].…”

“…Thus, the identification algorithms that use limited output measurements [ 19 , 20 , 21 , 22 , 23 , 24 , 25 ] have been developed to simplify the sensing systems of Zhao et al’s method [ 17 , 18 ] (it requires sensors to be deployed on all floors). Nevertheless, these methods [ 19 , 20 , 21 , 22 , 23 , 24 , 25 ] still entail more than one accelerometer to provide sufficient output data, and in particular, the methods in the literature [ 22 , 23 , 26 , 27 , 28 , 29 ] require simultaneous measurements from accelerometers and displacement or strain sensors for data fusion. To simplify SHM systems further, this study develops an identification algorithm for determining the relative displacement from measurement made by one accelerometer.…”

Estimation of maximum drift of multi‐degree‐of‐freedom shear structures with unknown parameters using only one accelerometer

“…Thus, we chose to use modal coordinates to circumvent having to identify the structural stiffness, which is different from what was done in Lei et al 21,22 Moreover, the reliability of the estimation should be taken into account in the case of strong ground excitations and extremely limited output. The response evaluation is affected by modeling error, 27,28 and Li et al 23 indicated that the estimation accuracy when using modal coordinates is vulnerable to the error in the measured natural frequencies, so the precise model parameters were used in those studies. However, when the natural frequencies are determined from the responses to the wind or small earthquake loads, they will be overestimated because of the stiffening effect of non‐structural components, 39 and they may shift during a large event because the connections between structural and non‐structural members loosen or fail, whereas the mode shapes do not significantly change 40 .…”

“…24 developed a modal Kalman filter to identify structural systems and unknown inputs without using collocated acceleration measurements, and further generalized algorithms were presented for the estimation from the absolute modal responses. 25 Other researchers have tried to use the relative responses instead of the absolute responses in the state vector and develop a state-space equation that does not have a direct feedthrough term based on the general equation of motion, [26][27][28] and it is also able to make the estimation by using the filters of the class of inputaugmented state-space models [29][30][31][32][33] and some of minimum-variance unbiased estimators 34 when the direct feedthrough term becomes zero. In addition, Hu et al 35,36 merged the responses of displacements, inclinations, and accelerations in the observation vector to reconstruct the structural responses for high-rise buildings under unknown seismic loading.…”