Optimal Sensor Placement for Time-Dependent Systems: Application to Wind Studies around Buildings
01 Mar 2016-Journal of Computing in Civil Engineering (American Society of Civil Engineers)-Vol. 30, Iss: 2, pp 04015024
TL;DR: In this article, the authors proposed a methodology for systematic sensor placement for situations when no measurement data are available and knowledge of the wind environment around buildings is limited, based on CFD simulation predictions at plausible locations.
Abstract: Warm climates pose challenges to building energy consumption and pedestrian comfort. Knowledge of the wind flow around buildings can help address these issues through improving natural ventilation, energy use, and outdoor thermal comfort. Computational fluid dynamics (CFD) simulations are widely used to predict wind flow around buildings, despite the large discrepancies that often occur between model predictions and actual measurements. Wind speed and direction exhibit a high degree of variability that adds uncertainties in modeling and measurements. Although some studies focus on methods to evaluate and minimize modeling uncertainties, sensor placement has been mostly based on subjective judgment and intuition; no systematic methodology is available to identify optimal sensor locations prior to field measurement. This work proposes a methodology for systematic sensor placement for situations when no measurement data are available and knowledge of the wind environment around buildings is limited. Sequential sensor placement algorithms and criteria are used to identify sensor configurations based on CFD simulation predictions at plausible locations. Optimal sensor configurations are compared for their ability to improve wind speed predictions at another location where no measurements are taken. The methodology is applied to two full-scale building systems of varying size. Results show that the methodology can be applied prior to field measurement to identify optimal configurations of a limited number of sensors that improve wind speed predictions at unmeasured locations. (C) 2015 American Society of Civil Engineers.
Citations
More filters
[...]
01 Jan 1997
TL;DR: It is shown that in the scope of a particular but quite general functional class it is possible to automatically choose the best equation form to represent a physical process, avoiding the hard work of analytical model definition that the designer should do.
Abstract: In this paper we propose an original modeling method. Our work can be considered as somewhere between classical modeling methods and learning based methods. We show that in the scope of a particular but quite general functional class it is possible to automatically choose the best equation form to represent a physical process, avoiding the hard work of analytical model definition that the designer should do. Our method uses an experimental protocol where the parameters identification is limited to one entry dimension problems, reducing the amount of data required by the model acquisition. The models generated by our method can be easily differentiated, corrected and reused. The method can be particularly useful in robotics where the functional form the method hands can be easily found in many kinds of problems.
79 citations
[...]
TL;DR: The model-falsification strategy for data interpretation provides engineers with an easy-to-understand tool that is compatible with the context of the built environment.
Abstract: Sensing in the built environment has the potential to reduce asset management expenditure and contribute to extending useful service life. In the built environment, measurements are usually performed indirectly; effects are measured remote from their causes. Modelling approximations from many sources, such as boundary conditions, geometrical simplifications and numerical assumptions result in important systematic uncertainties that modify correlation values between measurement points. In addition, conservative behavior models that were employed - justifiably during the design stage, prior to construction - are generally inadequate when explaining measurements of real behavior. This paper summarizes the special context of sensor data interpretation for asset management in the built environment. Nearly twenty years of research results from several doctoral thesis and fourteen full-scale case studies in four countries are summarized. Originally inspired from research into model based diagnosis, work on multiple model identification evolved into a methodology for probabilistic model falsification. Throughout the research, parallel studies developed strategies for measurement system design. Recent comparisons with Bayesian model updating have shown that while traditional applications Bayesian methods are precise and accurate when all is known, they are not robust in the presence of approximate models. Finally, details of the full-scale case studies that have been used to develop model falsification are briefly described. The model-falsification strategy for data interpretation provides engineers with an easy-to-understand tool that is compatible with the context of the built environment.
52 citations
Cites background from "Optimal Sensor Placement for Time-D..."
[...]
[...]
TL;DR: A sensor placement strategy using joint entropy is able to lead to predictions of wind characteristics around buildings and capture short-term wind variability more effectively than sequential strategies, which maximize entropy.
Abstract: Good prediction of the behavior of wind around buildings improves designs for natural ventilation in warm climates. However wind modeling is complex, predictions are often inaccurate due to the large uncertainties in parameter values. The goal of this work is to enhance wind prediction around buildings using measurements through implementing a multiple-model system-identification approach. The success of system-identification approaches depends directly upon the location and number of sensors. Therefore, this research proposes a methodology for optimal sensor configuration based on hierarchical sensor placement involving calculations of prediction-value joint entropy. Computational Fluid Dynamics (CFD) models are generated to create a discrete population of possible wind-flow predictions, which are then used to identify optimal sensor locations. Optimal sensor configurations are revealed using the proposed methodology and considering the effect of systematic and spatially distributed modeling errors, as well as the common information between sensor locations. The methodology is applied to a full-scale case study and optimum configurations are evaluated for their ability to falsify models and improve predictions at locations where no measurements have been taken. It is concluded that a sensor placement strategy using joint entropy is able to lead to predictions of wind characteristics around buildings and capture short-term wind variability more effectively than sequential strategies, which maximize entropy.
31 citations
[...]
01 Sep 2015
TL;DR: This paper demonstrates the implementation of CFD to Unity3D conversion and weather data visualization and proposes visualizations based on game engine technology to bridge the gap between architect and engineer.
Abstract: In architecture and urban design, urban climate on is a strong design criterion for outdoor thermal comfort and building's energy performance. Evaluating the effect of buildings on the local climate and vice versa is done by computational fluid dynamics (CFD) methods. The results from CFD are typically visualized through post-processing software closely related to pre-processing and simulation software. The built-in functions are made for engineers and thus, it lacks user-friendliness for real-time exploration of results for architects. To bridge the gap between architect and engineer we propose visualizations based on game engine technology. This paper demonstrates the implementation of CFD to Unity3D conversion and weather data visualization.
29 citations
Cites methods from "Optimal Sensor Placement for Time-D..."
[...]
[...]
TL;DR: This study presents a measurement system design methodology to identify the best sensor locations and sensor types using information from static-load tests and shows that a carefully-configured measurement strategy that includes multiple sensor types and several load tests maximizes information gain.
Abstract: Assessing ageing infrastructure is a critical challenge for civil engineers due to the difficulty in the estimation and integration of uncertainties in structural models. Field measurements are increasingly used to improve knowledge of the real behavior of a structure; this activity is called structural identification. Error-domain model falsification (EDMF) is an easy-to-use model-based structural-identification methodology which robustly accommodates systematic uncertainties originating from sources such as boundary conditions, numerical modelling and model fidelity, as well as aleatory uncertainties from sources such as measurement error and material parameter-value estimations. In most practical applications of structural identification, sensors are placed using engineering judgment and experience. However, since sensor placement is fundamental to the success of structural identification, a more rational and systematic method is justified. This study presents a measurement system design methodology to identify the best sensor locations and sensor types using information from static-load tests. More specifically, three static-load tests were studied for the sensor system design using three types of sensors for a performance evaluation of a full-scale bridge in Singapore. Several sensor placement strategies are compared using joint entropy as an information-gain metric. A modified version of the hierarchical algorithm for sensor placement is proposed to take into account mutual information between load tests. It is shown that a carefully-configured measurement strategy that includes multiple sensor types and several load tests maximizes information gain.
24 citations
Cites methods from "Optimal Sensor Placement for Time-D..."
[...]
References
More filters
Book•
[...]
31 Mar 1995
TL;DR: A comparison of Numerical Data with Test Results and Iterative Methods Using Modal Data for Model Updating shows that the former is more accurate than the latter.
Abstract: Preface. 1. Introduction. 2. Finite Element Modelling. 3. Vibration Testing. 4. Comparing Numerical Data with Test Results. 5. Estimation Techniques. 6. Parameters for Model Updating. 7. Direct Methods Using Modal Data. 8. Iterative Methods Using Modal Data. 9. Methods Using Frequency Domain Data. 10. Case Study: an Automobile Body M. Brughmans, J. Leuridan, K. Blauwkamp. 11. Discussion and Recommendations. Index.
2,100 citations
"Optimal Sensor Placement for Time-D..." refers methods in this paper
[...]
Journal Article•
[...]
TL;DR: It is proved that the problem of finding the configuration that maximizes mutual information is NP-complete, and a polynomial-time approximation is described that is within (1-1/e) of the optimum by exploiting the submodularity of mutual information.
Abstract: When monitoring spatial phenomena, which can often be modeled as Gaussian processes (GPs), choosing sensor locations is a fundamental task. There are several common strategies to address this task, for example, geometry or disk models, placing sensors at the points of highest entropy (variance) in the GP model, and A-, D-, or E-optimal design. In this paper, we tackle the combinatorial optimization problem of maximizing the mutual information between the chosen locations and the locations which are not selected. We prove that the problem of finding the configuration that maximizes mutual information is NP-complete. To address this issue, we describe a polynomial-time approximation that is within (1-1/e) of the optimum by exploiting the submodularity of mutual information. We also show how submodularity can be used to obtain online bounds, and design branch and bound search procedures. We then extend our algorithm to exploit lazy evaluations and local structure in the GP, yielding significant speedups. We also extend our approach to find placements which are robust against node failures and uncertainties in the model. These extensions are again associated with rigorous theoretical approximation guarantees, exploiting the submodularity of the objective function. We demonstrate the advantages of our approach towards optimizing mutual information in a very extensive empirical study on two real-world data sets.
1,463 citations
[...]
TL;DR: In this paper, the concept of the variance function for an experimental design is introduced, and the problem of selecting practically useful designs is discussed, and in this connection, the notion of variance function is introduced.
Abstract: Suppose that a relationship $\eta = \varphi(\xi_1, \xi_2, \cdots, \xi_k)$ exists between a response $\eta$ and the levels $\xi_1, \xi_2, \cdots, \xi_k$ of $k$ quantitative variables or factors, and that nothing is assumed about the function $\varphi$ except that, within a limited region of immediate interest in the space of the variables, it can be adequately represented by a polynomial of degree $d$. A $k$-dimensional experimental design of order $d$ is a set of $N$ points in the $k$-dimensional space of the variables so chosen that, using the data generated by making one observation at each of the points, all the coefficients in the $d$th degree polynomial can be estimated. The problem of selecting practically useful designs is discussed, and in this connection the concept of the variance function for an experimental design is introduced. Reasons are advanced for preferring designs having a "spherical" or nearly "spherical" variance function. Such designs insure that the estimated response has a constant variance at all points which are the same distance from the center of the design. Designs having this property are called rotatable designs. When such arrangements are submitted to rotation about the fixed center, the variances and covariances of the estimated coefficients in the fitted series remain constant. Rotatable designs having satisfactory variance functions are given for $d = 1, 2$; and $k = 2, 3, \cdots, \infty$. Blocking arrangements are derived. The simplification in the form of the confidence region for a stationary point resulting from the use of a second order rotatable design is discussed.
1,278 citations
"Optimal Sensor Placement for Time-D..." refers methods in this paper
[...]
[...]
TL;DR: In this paper, the authors present guidelines for using computational fluid dynamics (CFD) techniques for predicting pedestrian wind environment around buildings in the design stage, based on cross-comparison between CFD predictions, wind tunnel test results and field measurements.
Abstract: Significant improvements of computer facilities and computational fluid dynamics (CFD) software in recent years have enabled prediction and assessment of the pedestrian wind environment around buildings in the design stage. Therefore, guidelines are required that summarize important points in using the CFD technique for this purpose. This paper describes guidelines proposed by the Working Group of the Architectural Institute of Japan (AIJ). The feature of these guidelines is that they are based on cross-comparison between CFD predictions, wind tunnel test results and field measurements for seven test cases used to investigate the influence of many kinds of computational conditions for various flow fields.
1,274 citations
[...]
TL;DR: In this paper, the authors focus on the simulation of a neutrally stratified, fully developed, horizontally homogeneous ABL over uniformly rough, flat terrain and discuss the problem and its negative consequences.
Abstract: Accurate Computational Fluid Dynamics (CFD) simulations of atmospheric boundary layer (ABL) flow are essential for a wide variety of atmospheric studies including pollutant dispersion and deposition. The accuracy of such simulations can be seriously compromised when wall-function roughness modifications based on experimental data for sand-grain roughened pipes and channels are applied at the bottom of the computational domain. This type of roughness modification is currently present in many CFD codes including Fluent 6.2 and Ansys CFX 10.0, previously called CFX-5. The problems typically manifest themselves as unintended streamwise gradients in the vertical mean wind speed and turbulence profiles as they travel through the computational domain. These gradients can be held responsible—at least partly—for the discrepancies that are sometimes found between seemingly identical CFD simulations performed with different CFD codes and between CFD simulations and measurements. This paper discusses the problem by focusing on the simulation of a neutrally stratified, fully developed, horizontally homogeneous ABL over uniformly rough, flat terrain. The problem and its negative consequences are discussed and suggestions to improve the CFD simulations are made.
997 citations
"Optimal Sensor Placement for Time-D..." refers background in this paper
[...]