Seismic Induced Damage Detection through Parallel Estimation of Force and Parameter using Improved Interacting Particle-Kalman Filter

Abstract

Standard filtering techniques for structural parameter estimation assume that the input force either is known exactly or can be replicated using a known white Gaussian model. Unfortunately for structures subjected to seismic excitation, the input time history is unknown and also no previously known representative model is available. This invalidates the aforementioned idealization. To identify seismic induced damage in such structures using filtering techniques, a novel algorithm is proposed to estimate the force as additional state in parallel to the system parameters. Two concurrent filters are employed for parameters and force respectively. For the parameters, interacting Particle-Kalman filter [1] is employed targeting systems with correlated noise. Alongside a second filter is employed to estimate the seismic force acting on the structure. The proposal is numerically validated on a sixteen degrees-of-freedom mass-spring-damper system. The estimation results confirm the applicability of the proposed algorithm.

Integrating structural control, health monitoring, and energy harvesting for smart cities

“…Accordingly, the residuals are judged by a decision rule to identify the damage. The residuals can be obtained using bond graph approaches (Chatti et al, 2014), diagnosis observers (Yang et al, 2015), parity relations (Zhong et al, 2015), or parameter estimation (Sen et al, 2018).…”

Estimation of local failure in tensegrity using Interacting Particle-Ensemble Kalman Filter

“…An IP-EnKF approach has been adopted for this in which the PF approaches the non-linear health parameter estimation while the EnKF estimates the non-linearly evolving system states x k (as per Equation ( 13)). IP-EnKF can therefore be considered as an improvisation of IPKF [52] wherein EnKF replaces KF to extend the reach of the algorithm to non-linear systems. It further facilitates with the option to parallelize the entire computation.…”

“…Estimation of the HIs can further be approached either jointly [36] or conditionally [17,51] with respect to the real system states. The relative efficiency of the conditional over joint estimation approach has already been corroborated in several articles [13] and upon further introduction of interacting strategies by [32], the focus has strongly shifted to the use of individual filters for states or parameter estimation, like in Interacting Particle-Kalman filter (IPKF) [52,71], Dual KF [9], Dual EKF (DEKF) [51], etc.…”

Strain‐based joint damage estimation approach robust to unknown non‐stationary input force

“…With the Bayesian filtering‐based stochastic estimation approach, the system can be estimated efficiently even in the presence of system and measurement uncertainties as long as the uncertainty models are available. Successful application of this approach for SHM problems has been reported in abundance 15–21 …”

“…This approach has further been improved for computational efficiency by Criniére et al 43 and for a system with correlated noise by Sen et al 44 This approach assumes stationarity and Gaussianity for input forcing which most often gets violated, especially for the cases with wind/wave/seismic excitation. An improved version of the mentioned IPKF algorithm includes an additional force filter to handle non‐stationary forcing types, 18 which also increases the computational time. There are studies that focus on input force estimation filters 45,46 for time varying systems with input as an additional state 47–51 .…”