Article
Chemistry, Analytical
Tulay Ercan, Costas Papadimitriou
Summary: A framework for optimal sensor placement for virtual sensing is proposed based on modal expansion technique and information theory. The framework maximizes a utility function to reduce uncertainty in predicted quantities of interest at virtual sensing locations, considering uncertainties in structural model and modeling error parameters. The Gaussian nature of the response is utilized to derive analytical expressions for the utility function, highlighting the importance of robustness to errors and uncertainties.
Article
Engineering, Mechanical
Zifan Zhang, Chang Peng, Guangjun Wang, Zengye Ju, Long Ma
Summary: This study proposes a new method for optimal sensor placement of composite virtual strain sensing, which can obtain the global unbiased estimation of modal coordinates. It utilizes a Bayesian probabilistic model and K-L divergence to achieve this goal. The proposed method also introduces a new variance determination method to improve the stability of the solution. The effectiveness of the method is demonstrated through an example involving a laminate plate.
MECHANICAL SYSTEMS AND SIGNAL PROCESSING
(2023)
Article
Engineering, Mechanical
Yichao Yang, Mayank Chadha, Zhen Hu, Michael D. Todd
Summary: This paper introduces a novel framework for optimal sensor placement design in structural health monitoring using Bayes risk as the objective function. The framework considers external and internal costs, making it applicable to various SHM designs. Through an example problem, the effectiveness and comprehensiveness of the framework are demonstrated, along with discussions on challenges such as computationally expensive models and uncertainty quantification.
MECHANICAL SYSTEMS AND SIGNAL PROCESSING
(2022)
Article
Multidisciplinary Sciences
Limin Zhang, Jing Dong, Junfang Zhang, Junzi Yang
Summary: This paper introduces a novel variational inference method that combines Bayesian and gradient descent techniques. A modified Stein variational inference algorithm is proposed to make the gradient descent of Kullback-Leibler divergence more random. The suggested technique is validated using four data sets, and its performance is evaluated using statistical measures such as parameter estimate classification accuracy, F1, and NRMSE.
Article
Acoustics
Zifan Zhang, Chang Peng, Guangjun Wang, Zengye Ju, Long Ma
Summary: In this study, optimal sensor placement (OSP) for strain sensing of a high-speed EMU beam is investigated using the information gain and modal expansion method within the Bayesian framework. The role of prediction error in OSP and strain reconstruction is studied in detail, and two new prediction errors are proposed. The method is validated through a case study of a beam and full-scale beam monitoring systems based on Fiber Bragg grating (FBG) sensors.
JOURNAL OF SOUND AND VIBRATION
(2023)
Article
Engineering, Mechanical
Yichao Yang, Mayank Chadha, Zhen Hu, Manuel A. Vega, Matthew D. Parno, Michael D. Todd
Summary: This study proposes a new approach to optimal sensor design for structural health monitoring using a modified f-divergence objective functional to infer the unknown and uncertain damage state parameter. Risk-adjustment is made using functions that weigh the importance of acquiring useful information for a given true value of the state parameter, based on the loss of boundary contact between a navigation lock miter gate and the supporting wall quoin block.
MECHANICAL SYSTEMS AND SIGNAL PROCESSING
(2021)
Article
Engineering, Mechanical
Tulay Ercan, Omid Sedehi, Lambros S. Katafygiotis, Costas Papadimitriou
Summary: An optimal sensor placement (OSP) framework for virtual sensing using the augmented Kalman Filter (AKF) technique is proposed based on information and utility theory. The framework considers uncertainties in the structural model and modelling error parameters, and maximizes the utility function through heuristic sequential sensor placement (SSP) strategies and genetic algorithms (GA). The study highlights the importance of accounting for robustness to errors and uncertainties in selecting the optimal sensor configuration using a Finite Element (FE) model.
MECHANICAL SYSTEMS AND SIGNAL PROCESSING
(2023)
Article
Engineering, Mechanical
C. Argyris, C. Papadimitriou, G. Samaey, G. Lombaert
Summary: A Bayesian framework for optimal sensor placement based on model optimization is presented to minimize uncertainty in predicting a specific quantity of interest. Emphasizing prediction inference over parameter inference, the method aims to reduce uncertainty in key parameters for accurate predictions. By using the determinant to measure uncertainty and evaluating covariance matrices through Monte Carlo sampling, the approach differs from traditional methods and is more suitable for prediction inference.
MECHANICAL SYSTEMS AND SIGNAL PROCESSING
(2021)
Article
Engineering, Mechanical
Burak Bagirgan, Azin Mehrjoo, Babak Moaveni, Costas Papadimitriou, Usman Khan, Jason Rife
Summary: This paper proposes an iterative optimal sensor placement (OSP) framework for structural identification and model updating using mobile sensors. The OSP is performed iteratively to minimize information entropy in estimating the model's updating parameters. A forward sequential sensor placement algorithm is used to solve the OSP problem at each iteration. The proposed framework is applied to a numerical case study and shows better model updating results compared to using an optimal static sensor configuration with a larger number of sensors.
MECHANICAL SYSTEMS AND SIGNAL PROCESSING
(2023)
Article
Engineering, Mechanical
Azin Mehrjoo, Mingming Song, Babak Moaveni, Costas Papadimitriou, Eric Hines
Summary: This paper proposes an optimal sensor placement (OSP) framework using information theory, which combines a Bayesian OSP method with modal expansion to minimize information entropy about quantities of interest (QoI) without knowledge of input excitation. The framework takes into account variations in sensor installation cost and has been numerically evaluated using a realistic model of an offshore wind turbine. The results demonstrate that the OSP framework is an effective tool for decision-makers and can offer valuable insights when considering cost constraints.
MECHANICAL SYSTEMS AND SIGNAL PROCESSING
(2022)
Article
Engineering, Mechanical
Sahil Bansal, Sai Hung Cheung
Summary: This study proposes an optimal Bayesian sensor placement approach for updating linear structural models, involving two stages of identifying modal parameters and updating model parameters, selecting the sensor configuration that maximizes the expected information gain.
MECHANICAL SYSTEMS AND SIGNAL PROCESSING
(2022)
Article
Chemistry, Multidisciplinary
Qiao Hu, Yangkun Zhang, Xingju Xie, Wenbin Su, Yangyang Li, Liuhao Shan, Xiaojie Yu
Summary: This paper presents an optimal sensor placement method for vibration signal acquisition in the field of industrial robot health monitoring and fault diagnosis. The method derives an evaluation function for sensor placement based on the general formula of Bayes and relative entropy and uses a modal confidence matrix to express the redundancy of sensor placement. It describes the optimal placement of vibration sensors as a discrete variable optimization problem and provides a theoretical basis for industrial robots to acquire vibration data effectively.
APPLIED SCIENCES-BASEL
(2022)
Article
Computer Science, Artificial Intelligence
Shuya Nagayasu, Sumio Watanbe
Summary: In this study, by theoretical derivation, we investigate the asymptotic behaviors of the generalization loss and the free energy in Bayesian inference when there are multiple optimal probability distributions, revealing differences from conventional asymptotic analysis.
Article
Engineering, Mechanical
Vesa Nieminen, Jussi Sopanen
Summary: Sensor placement is a crucial factor in determining the quality and accuracy of virtual sensing. This study proposes a two-phase optimization method for triaxial accelerometers, using minimum variance criterion and a measure of redundancy of information. The method successfully avoids spatial correlation and clustering of sensor locations and introduces a weighting proposal based on modal displacement to enhance selection of sensor positions in noisy environments. The effectiveness of the method is demonstrated through numerical models and laboratory experiments. The average error in response reconstruction was found to be 1.4% of the maximum measured response amplitude. This method is particularly suitable for large finite element models of industrial-scale structures with fine meshes.
MECHANICAL SYSTEMS AND SIGNAL PROCESSING
(2023)
Article
Construction & Building Technology
Ka-Veng Yuen, Xiao-Han Hao, Sin-Chi Kuok
Summary: In this paper, a robust sensor placement methodology considering sensor failure is proposed, which introduces the concept of robust information entropy and a tailor-made heuristic sequential search algorithm to enhance efficiency. The designed robust sensor configuration is illustrated through designing sensors for a 20-story shear building and a space truss, and validated using a case study of the Canton Tower's in-field measurements.
STRUCTURAL CONTROL & HEALTH MONITORING
(2022)
Article
Construction & Building Technology
Victor Flores Terrazas, Omid Sedehi, Costas Papadimitriou, Lambros S. Katafygiotis
Summary: This paper proposes a framework for monitoring fatigue in linear components of a structure using limited vibration data, estimating dynamic responses through virtual sensing techniques and computing fatigue damage. The approach allows for realistic fatigue damage information based on actual operational conditions.
STRUCTURAL CONTROL & HEALTH MONITORING
(2022)
Article
Engineering, Multidisciplinary
Xinyu Jia, Omid Sedehi, Costas Papadimitriou, Lambros S. Katafygiotis, Babak Moaveni
Summary: A new time-domain probabilistic technique based on hierarchical Bayesian modeling framework is proposed for calibration and uncertainty quantification of hysteretic type nonlinearities of dynamical systems. The technique introduces probabilistic hyper models for material hysteretic model parameters and prediction error variance parameters, considering both the uncertainty of the model parameters and the prediction error uncertainty. The technique employs a new asymptotic approximation to simplify the process of nonlinear model updating and reduce the computational burden. Numerical examples demonstrate the accuracy and performance of the proposed method.
COMPUTER METHODS IN APPLIED MECHANICS AND ENGINEERING
(2022)
Article
Engineering, Mechanical
Mansureh-Sadat Nabiyan, Hamed Ebrahimian, Babak Moaveni, Costas Papadimitriou
Summary: An adaptive recursive Bayesian inference framework is developed in this study to estimate model parameters and the statistical characteristics of the prediction error. The proposed approach allows for better consideration of time-variant model uncertainties in the model updating process.
JOURNAL OF ENGINEERING MECHANICS
(2022)
Article
Engineering, Mechanical
Daniz Teymouri, Omid Sedehi, Lambros S. Katafygiotis, Costas Papadimitriou
Summary: This study proposes a joint input-state estimation and virtual sensing method based on Bayesian probability theory, focusing on data-driven uncertainty quantification and propagation. By introducing a random walk model for input forces and including input pseudo-observations, singularity problems are overcome and a Bayesian Expectation-Maximization (BEM) strategy is established for parameter estimation.
MECHANICAL SYSTEMS AND SIGNAL PROCESSING
(2022)
Article
Engineering, Mechanical
Azin Mehrjoo, Mingming Song, Babak Moaveni, Costas Papadimitriou, Eric Hines
Summary: This paper proposes an optimal sensor placement (OSP) framework using information theory, which combines a Bayesian OSP method with modal expansion to minimize information entropy about quantities of interest (QoI) without knowledge of input excitation. The framework takes into account variations in sensor installation cost and has been numerically evaluated using a realistic model of an offshore wind turbine. The results demonstrate that the OSP framework is an effective tool for decision-makers and can offer valuable insights when considering cost constraints.
MECHANICAL SYSTEMS AND SIGNAL PROCESSING
(2022)
Article
Engineering, Mechanical
Xinyu Jia, Omid Sedehi, Costas Papadimitriou, Lambros S. Katafygiotis, Babak Moaveni
Summary: This paper introduces the hierarchical Bayesian modeling (HBM) framework for uncertainty quantification and propagation in structural dynamics inverse problems. The framework is developed further for model inference based on modal features and incorporates asymptotic approximations to simplify computation. The proposed framework is beneficial for propagating uncertainty and providing reasonable uncertainty bounds for both structural and prediction error parameters.
MECHANICAL SYSTEMS AND SIGNAL PROCESSING
(2022)
Article
Acoustics
Sergio Cantero-Chinchilla, Adriano T. Fabro, Han Meng, Wang-Ji Yan, Costas Papadimitriou, Dimitrios Chronopoulos
Summary: This work proposes a strategy for optimal design of mechanical metastructure considering uncertainties arising from additive manufacturing. The variability of material properties introduced by the additive manufacturing process is experimentally obtained, and a transfer matrix approach is employed to predict the structural receptance of the metastructure. The mass ratio of the metastructure is optimized for maximizing vibration attenuation, and a cost function is introduced to favor designs with least complexity. Results show that even small variability in material properties can affect the robustness of the optimal design for locally resonant metastructures.
JOURNAL OF SOUND AND VIBRATION
(2022)
Article
Engineering, Mechanical
Menghao Ping, Xinyu Jia, Costas Papadimitriou, Xu Han, Chao Jiang
Summary: A new Bayesian modeling framework is proposed to account for the uncertainties in model parameters arising from various factors. The framework incorporates uncertainty using a single level hierarchy with Normal distributions. The likelihood function is constructed based on the discrepancy between model predictions and measurements, and the posterior PDF of model parameters depends on the lower two moments of the respective PDFs.
MECHANICAL SYSTEMS AND SIGNAL PROCESSING
(2022)
Article
Engineering, Multidisciplinary
Olga Markogiannaki, Alexandros Arailopoulos, Dimitrios Giagopoulos, Costas Papadimitriou
Summary: This paper presents a model-based Damage Detection Framework for truss structural systems, which utilizes vibration measurements to detect damage. The framework provides accurate damage localization and quantification, and can be achieved using a limited number of sensors for unknown input excitation.
STRUCTURAL HEALTH MONITORING-AN INTERNATIONAL JOURNAL
(2023)
Article
Engineering, Mechanical
Victor Flores Terrazas, Omid Sedehi, Costas Papadimitriou, Lambros S. Katafygiotis
Summary: A Bayesian framework is proposed to generate probabilistic stress-life fatigue curves and identify parameters of a multiaxial fatigue model using experimental data. Classical and hierarchical Bayesian inference strategies are presented, along with analytical expressions for calculating the joint posterior distributions. An example demonstrates the application of the proposed hierarchical Bayesian inference framework and its comparison to a deterministic approach. This probabilistic treatment enables uncertainty propagation for reliability analysis and design using existing fatigue models.
INTERNATIONAL JOURNAL OF FATIGUE
(2022)
Article
Engineering, Mechanical
Wang-Ji Yan, Teng-Teng Hao, Ka-Veng Yuen, Costas Papadimitriou
Summary: In this study, a new transmissibility-like index was proposed for monitoring the Gross Vehicle Weights (GVWs) of heavy vehicles. An influence line-free algorithm was used to estimate the GVW of arbitrary vehicles, and the Bridge Weigh-In-Motion (B-WIM) problem was formulated in the framework of Bayesian inference. The efficiency and accuracy of the proposed method were demonstrated through numerical examples and experimental verification.
MECHANICAL SYSTEMS AND SIGNAL PROCESSING
(2022)
Article
Engineering, Mechanical
Daniz Teymouri, Omid Sedehi, Lambros S. S. Katafygiotis, Costas Papadimitriou
Summary: This paper proposes a new Bayesian approach to estimate input and state in linear structures using response-only measurements. The approach utilizes a modally reduced state-space model to overcome the dimensionality of dynamical responses in complex structures. It also replaces unknown physical forces with equivalent modal forces, which is advantageous when the magnitude and location of input forces are unknown. The proposed method achieves accurate estimations and reasonable uncertainty bounds for the dynamical state and input forces.
JOURNAL OF ENGINEERING MECHANICS
(2023)
Article
Engineering, Mechanical
Tulay Ercan, Omid Sedehi, Lambros S. Katafygiotis, Costas Papadimitriou
Summary: An optimal sensor placement (OSP) framework for virtual sensing using the augmented Kalman Filter (AKF) technique is proposed based on information and utility theory. The framework considers uncertainties in the structural model and modelling error parameters, and maximizes the utility function through heuristic sequential sensor placement (SSP) strategies and genetic algorithms (GA). The study highlights the importance of accounting for robustness to errors and uncertainties in selecting the optimal sensor configuration using a Finite Element (FE) model.
MECHANICAL SYSTEMS AND SIGNAL PROCESSING
(2023)
Article
Engineering, Mechanical
Burak Bagirgan, Azin Mehrjoo, Babak Moaveni, Costas Papadimitriou, Usman Khan, Jason Rife
Summary: This paper proposes an iterative optimal sensor placement (OSP) framework for structural identification and model updating using mobile sensors. The OSP is performed iteratively to minimize information entropy in estimating the model's updating parameters. A forward sequential sensor placement algorithm is used to solve the OSP problem at each iteration. The proposed framework is applied to a numerical case study and shows better model updating results compared to using an optimal static sensor configuration with a larger number of sensors.
MECHANICAL SYSTEMS AND SIGNAL PROCESSING
(2023)
Article
Engineering, Multidisciplinary
Antonina M. Kosikova, Omid Sedehi, Costas Papadimitriou, Lambros S. Katafygiotis
Summary: This paper investigates Bayesian model updating based on Gaussian Process models by reformulating the problem and proposing a new kernel function selection method, aiming to balance fitting accuracy, generalizability, and model parsimony. Computational issues are addressed using Laplace approximation and sampling techniques, and numerical and experimental examples are provided to demonstrate the accuracy and robustness of the proposed framework.
COMPUTER METHODS IN APPLIED MECHANICS AND ENGINEERING
(2023)
