An adaptive-noise Augmented Kalman Filter approach for input-state estimation in structural dynamics

Vettori, Silvia; Lorenzo, Emilio Di; Peeters, Bart; Łuczak, Marcin; Chatzi, Eleni

doi:10.1016/j.ymssp.2022.109654

Cited by 26 publications

(12 citation statements)

References 52 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…When dealing with the tuning of Kalman-type filters, researchers have developed methods for the online identification of noise parameters to improve the estimation accuracy (Kontoroupi & Smyth, 2016;Yuen & Kuok, 2016). While the majority of adaptive noise tuning for recursive Bayesian estimators was focused on an online estimation of covariance of modeling error and measurement noise, Vettori et al (2023) recently proposed an adaptive tuning of the covariance of an unknown force in an AKF framework.…”

Section: Akf Versus the Present Methodsmentioning

confidence: 99%

“…In this example, the tuning of the AKF focuses on minimizing the observation error rather than the input error, as the estimation quality would significantly deteriorate if the input error were to be minimized (Vettori et al, 2023). The error 𝛿 versus 𝑄 𝑥 and 𝑄 𝑓 for the AKF is plotted in Figure 2 in which the sensitivity to the variation of 𝑄 𝑥 is negligible compared to that of 𝑄 𝑓 .…”

Section: Akf Versus the Present Methodsmentioning

confidence: 99%

“…The record of the 1976 Friuli earthquake is chosen as the base excitation of the tower. The frequencies of the earthquake range from 1.57 to 6.28 rad/s (Aoudia et al., 2000). The observations include displacement measurements at the first, sixth, 22nd, and the 34th floors as well as velocity measurements at the first, fifth, 15th, 20th, 25th, 30th, 35th, and the 39th floors.…”

Section: Numerical Examplesmentioning

confidence: 99%

See 2 more Smart Citations

A linear Bayesian filter for input and state estimation of structural systems

Hassanabadi

Liu

Azam

et al. 2023

Computer aided Civil Eng

View full text Add to dashboard Cite

This paper proposes a linear recursive Bayesian filter for minimum variance unbiased joint input and state estimation of structural systems. Unlike the augmented Kalman filter (AKF), the proposed filter falls within the category of Bayesian filters in which unknown inputs are estimated without attributing any fictitious input model or statistics. Also, in contrast with the existing algorithms in the latter category, such as the Gillijns and De Moor Filters (GDFs), the developed filter applies to systems with and without direct feedthrough, in particular, systems with a rank-deficient feedforward matrix. Because of the latter features, the filter is referred to as universal filter (UF) for convenience. The numerical examples show that the UF performs better than the AKF. Due to its structure, the UF does not require the tuning of the hyperparameters for inputs, and therefore the problematic instability of the AKF is not encountered in the case of a large modeling error variance of the input. For systems with direct feedthrough, the error and covariance propagation terms differ due to the distinct state space.Consequently, the UF can enhance estimations due to the well-conditionedness of the relevant inversion problem. Moreover, the UF can deal with systems with rank-deficient feedforward matrix where these systems are not covered by GDFs.

show abstract

Section: Akf Versus the Present Methodsmentioning

confidence: 99%

Section: Akf Versus the Present Methodsmentioning

confidence: 99%

Section: Numerical Examplesmentioning

confidence: 99%

See 1 more Smart Citation

A linear Bayesian filter for input and state estimation of structural systems

Hassanabadi

Liu

Azam

et al. 2023

Computer aided Civil Eng

View full text Add to dashboard Cite

show abstract

“…To derive the augmented state-space model, the discrete-time state equation (1) is supplemented with an equation relating the unknown parameter vector

at time

and time

. Due to the lack of prior information on the time evolution of uncertain parameters, a discrete-time random-walk model is adopted to describe their dynamics [ 36 , 39 ]:

where the noise vector

, which accounts for the parameter increment, is assumed to be a zero-mean multivariate normal distribution with associated covariance matrix

. Equation (3) allows the time trend of unknown parameters to be estimated through an appropriate calibration of the covariance

.…”

Section: Digital Twin Architecturementioning

confidence: 99%

Digital twin with augmented state extended Kalman filters for forecasting electric power consumption of industrial production systems

Baldassarre,

Dion,

Peyret

et al. 2024

Heliyon

View full text Add to dashboard Cite

“…In this section, ANAKF [29] is used for pre-processing the animation image. ANAKF is used to remove the noise from the image.…”

Section: B Pre-processing Using An Adaptive-noise Augmented Kalman Fi...mentioning

confidence: 99%

Virtual Reality Animation Interaction Design using Bayesian Physics-Informed Neural Network with Archimedes Optimization Algorithm Based Scene Modelling

Pengyuan Zhu

2024

jes

View full text Add to dashboard Cite

Virtual Reality Animation Interaction Design is a dynamic field at the intersection of technology and creativity, where immersive experiences come to life through the fusion of animation and interactive design. There are some challenges involves in design of virtual Reality animation. The main challenge is animation image error involves in the design. To overcome this issue, present a Virtual Reality Animation Interaction Design Using Bayesian Physics-Informed Neural Network with Archimedes Optimization Algorithm Based scene modelling (VRAID-BPINN-AOA) is proposed. Initially, the animation images are collected from animation dataset. Then, the animation images are fed to pre-processing segment. In pre-processing segment, the noise of the image is removing layer by layer by utilizing adaptive-noise Augmented Kalman Filter (ANAKF). Then the pre-processed animation image is given for Feature extraction process. In Feature extraction, spatiotemporal features such as Object Position, velocity, Optical flow and Crowd Density are extracted by utilizing Multi-Level 2-D Quantum Wavelet Transform (ML2DQWT). Finally the extracted feature attributes are given to Bayesian physics-informed neural network (BPINN) for the prediction of error in the animation images. In general, BPINN does not express some adaption of optimization strategies for determining optimal parameters to promise accurate prediction of error. Therefore, Archimedes Optimization Algorithm (AOA) is proposed to optimize the parameter of BPINN .The proposed technique is implemented and efficacy of VRAID-BPINN-AOA technique is assessed by support of numerous performances such as Bit Error Rate, Design Rate, End to End Delay, Latency, and Transmission Rate. Proposed VRAID-BPINN-AOA method attains 28.56%, 26.67% and 25.67% lower Bit error rate and26.42%, 25.67% and 23.67% lower Design rate are compared with existing such as Virtual Interactive Animation Design with Restricted Boltzmann machine (VI-AD-RBM),Generative deep learning for visual animation in landscapes design using Generative adversarial network(GDL-VALD-GAN), and Hand interface using deep learning in immersive virtual reality using convolutional neural network(HID-LVIR-CNN)respectively.

show abstract

An adaptive-noise Augmented Kalman Filter approach for input-state estimation in structural dynamics

Cited by 26 publications

References 52 publications

A linear Bayesian filter for input and state estimation of structural systems

A linear Bayesian filter for input and state estimation of structural systems

Digital twin with augmented state extended Kalman filters for forecasting electric power consumption of industrial production systems

Virtual Reality Animation Interaction Design using Bayesian Physics-Informed Neural Network with Archimedes Optimization Algorithm Based Scene Modelling

Contact Info

Product

Resources

About