Calibration Framework Research Articles

Modified Volume-Delay Function Based on Traffic Fundamental Diagram: A Practical Calibration Framework for Estimating Congested and Uncongested Conditions

Modified Volume-Delay Function Based on Traffic Fundamental Diagram: A Practical Calibration Framework for Estimating Congested and Uncongested Conditions

Comparison of Parallel Genetic Algorithm and Particle Swarm Optimization for Parameter Calibration in Hydrological Simulation

ABSTRACT Parameter calibration is an important part of hydrological simulation and affects the final simulation results. In this paper, we introduce heuristic optimization algorithms, genetic algorithm (GA) to cope with the complexity of the parameter calibration problem, and use particle swarm optimization algorithm (PSO) as a comparison. For large-scale hydrological simulations, we use a multilevel parallel parameter calibration framework to make full use of processor resources, and accelerate the process of solving high-dimensional parameter calibration. Further, we test and apply the experiments on domestic supercomputers. The results of parameter calibration with GA and PSO can basically reach the ideal value of 0.65 and above, with PSO achieving a speedup of 58.52 on TianHe-2 supercomputer. The experimental results indicate that using a parallel implementation on multicore CPUs makes high-dimensional parameter calibration in large-scale hydrological simulation possible. Moreover, our comparison of the two algorithms shows that the GA obtains better calibration results, and the PSO has a more pronounced acceleration effect.

Development of low-cost air quality stations for next-generation monitoring networks: calibration and validation of NO2 and O3 sensors

Abstract. A pre-deployment calibration and a field validation of two low-cost (LC) stations equipped with O3 and NO2 metal oxide sensors were addressed. Pre-deployment calibration was performed after developing and implementing a comprehensive calibration framework including several supervised learning models, such as univariate linear and non-linear algorithms, and multiple linear and non-linear algorithms. Univariate linear models included linear and robust regression, while univariate non-linear models included a support vector machine, random forest, and gradient boosting. Multiple models consisted of both parametric and non-parametric algorithms. Internal temperature, relative humidity, and gaseous interference compounds proved to be the most suitable predictors for multiple models, as they helped effectively mitigate the impact of environmental conditions and pollutant cross-sensitivity on sensor accuracy. A feature analysis, implementing dominance analysis, feature permutations, and the SHapley Additive exPlanations method, was also performed to provide further insight into the role played by each individual predictor and its impact on sensor performances. This study demonstrated that while multiple random forest (MRF) returned a higher accuracy than multiple linear regression (MLR), it did not accurately represent physical models beyond the pre-deployment calibration dataset, so a linear approach may overall be a more suitable solution. Furthermore, as well as being less computationally demanding and generally more suitable for non-experts, parametric models such as MLR have a defined equation that also includes a few parameters, which allows easy adjustments for possible changes over time. Thus, drift correction or periodic automatable recalibration operations can be easily scheduled, which is particularly relevant for NO2 and O3 metal oxide sensors. As demonstrated in this study, they performed well with the same linear model form but required unique parameter values due to intersensor variability.

Performance evaluation of MOMA (MOment MAtching) – a remote network calibration technique for PM2.5 and PM10 sensors

Abstract. We evaluate the potential of using a previously developed remote calibration framework we name MOMA (MOment MAtching) to improve the data quality in particulate matter (PM) sensors deployed in hierarchical networks. MOMA assumes that a network of reference instruments can be used as “proxies” to calibrate the sensors given that the probability distribution over time of the data at the proxy site is similar to that at a sensor site. We use the reference network to test the suitability of proxies selected based on distance versus proxies selected based on land use similarity. The performance of MOMA for PM sensors is tested with sensors co-located with reference instruments across three Southern Californian regions, representing a range of land uses, topography and meteorology, and calibrated against a distant proxy reference. We compare two calibration approaches: one where calibration parameters get calculated and applied at monthly intervals and one which uses a drift detection framework for calibration. We demonstrate that MOMA improves the accuracy of the data when compared against the co-located reference data. The improvement was more visible for PM10 and when using the drift detection approach. We also highlight that sensor drift was associated with variations in particle composition rather than instrumental factors, explaining the better performance of the drift detection approach if wind conditions and associated PM sources varied within a month.

In Situ Calibration of Antenna Arrays for Positioning With 5G Networks

Owing to the ubiquity of cellular communication signals, positioning with the fifth-generation (5G) signal has emerged as a promising solution in global navigation satellite system-denied areas. Unfortunately, although the widely employed antenna arrays in 5G remote radio units (RRUs) facilitate the measurement of the direction of arrival (DOA), DOA-based positioning performance is severely degraded by array errors. This article proposes an in situ calibration framework with a user terminal (UT) transmitting 5G reference signals at several known positions in the actual operating environment and the accessible RRUs estimating their array errors from these reference signals. Further, since sub-6 GHz small-cell RRUs deployed for indoor coverage generally have small-aperture antenna arrays, while 5G signals have plentiful bandwidth resources, this work segregates the multipath components via super-resolution delay estimation based on the maximum likelihood criteria. This differs significantly from existing in situ calibration works which resolve multipaths in the spatial domain. The superiority of the proposed method is first verified by numerical simulations. We then demonstrate via field test with commercial 5G equipment that, a reduction of 46.7% for <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$1{\text -}\sigma$</tex-math> </inline-formula> DOA estimation error can be achieved by in situ calibration using the proposed method.

Zero-shot Scene Graph Generation via Triplet Calibration and Reduction

Scene Graph Generation (SGG) plays a pivotal role in downstream vision-language tasks. Existing SGG methods typically suffer from poor compositional generalizations on unseen triplets. They are generally trained on incompletely annotated scene graphs that contain dominant triplets and tend to bias toward these seen triplets during inference. To address this issue, we propose a Triplet Calibration and Reduction (T-CAR) framework in this article. In our framework, a triplet calibration loss is first presented to regularize the representations of diverse triplets and to simultaneously excavate the unseen triplets in incompletely annotated training scene graphs. Moreover, the unseen space of scene graphs is usually several times larger than the seen space, since it contains a huge number of unrealistic compositions. Thus, we propose an unseen space reduction loss to shift the attention of excavation to reasonable unseen compositions to facilitate the model training. Finally, we propose a contextual encoder to improve the compositional generalizations of unseen triplets by explicitly modeling the relative spatial relations between subjects and objects. Extensive experiments show that our approach achieves consistent improvements for zero-shot SGG over state-of-the-art methods. The code is available at https://github.com/jkli1998/T-CAR .

A modular Bayesian approach to model calibration of low‐fidelity simulation models

AbstractLow‐fidelity computational models have been widely used for the computation of complex engineered systems to greatly save the computational time while sacrificing the computational precision. Model calibration can be implemented for low‐fidelity simulation models to remedy the errors of simulation results. This article proposes a modular Bayesian updating framework that considers epistemic uncertainty to achieve model calibration of low‐fidelity simulation models. The proposed framework mainly contains two steps: (1) A model calibration framework based on Bayesian theory is proposed that can simultaneously quantify the model form uncertainty, model parameter uncertainty, and the experimental measurement uncertainty, (2) the proposed method updates a low‐fidelity surrogate model via a Metropolis–Hastings (M–H) algorithm by using high‐fidelity experimental data. The proposed method greatly improves the prediction accuracy of the simulation model and enhances the computational efficiency. A mathematical example is used to testify the accuracy of the developed method, while the aerodynamics simulation of the NACA0012 airfoil is leveraged to demonstrate the engineering application. The results show that the posterior prediction mean is in better agreement with the experimental mean than the prior prediction mean, suggesting that the modular Bayesian updating method improves the prediction accuracy of low‐fidelity simulation models.

Development of the Social Force Model Considering Pedestrian Characteristics and Behavior

Use of pedestrian facilities like sidewalks is influenced by pedestrian characteristics like age, gender, and purpose, for example, walking in social groups, shopping. Helbing and Molnár’s base social force model (SFM) has proven reliable at simulating pedestrian behavior, However, only a small number of simulation models, including SFM and its variations, have concentrated on modeling pedestrian walking while taking into account their traits and behavior. This study’s goal was to simulate bidirectional pedestrian walking behavior by taking into account various pedestrian characteristics (age, gender, behavior) by introducing a novel pedestrian simulation model: the social force model considering pedestrian characteristics and behavior (SFMPCB). Pedestrian gender alone was considered as a modification in our earlier study. The current study also implements a multistep global calibration framework based on a genetic algorithm that takes into account the bidirectional flow of pedestrians from various sidewalks. For calibration and validation, videographic data of pedestrian traffic from four sidewalks in Mumbai, India, were employed. The SFMPCB could be utilized to simulate pedestrian walking on sidewalks reflecting different characteristics and behavior. The quality of service and sidewalk design could also be better studied by utilizing the SFMPCB by generating pedestrian equivalent factors for pedestrian characteristics, similar to passenger car units.

Issues in Implementing Regression Calibration Analyses.

Regression calibration is a popular approach for correcting biases in estimated regression parameters when exposure variables are measured with error. This approach involves building a calibration equation to estimate the value of the unknown true exposure given the error-prone measurement and other covariates. The estimated, or calibrated, exposure is then substituted for the unknown true exposure in the health outcome regression model. When used properly, regression calibration can greatly reduce the bias induced by exposure measurement error. Here, we first provide an overview of the statistical framework for regression calibration, specifically discussing how a special type of error, called Berkson error, arises in the estimated exposure. We then present practical issues to consider when applying regression calibration, including: 1) how to develop the calibration equation and which covariates to include; 2) valid ways to calculate standard errors of estimated regression coefficients; and 3) problems arising if one of the covariates in the calibration model is a mediator of the relationship between the exposure and outcome. Throughout, we provide illustrative examples using data from the Hispanic Community Health Study/Study of Latinos (United States, 2008-2011) and simulations. We conclude with recommendations for how to perform regression calibration.

Signatures-and-sensitivity-based multi-criteria variational calibration for distributed hydrological modeling applied to Mediterranean floods

Classical calibration methods in hydrology typically rely on a single cost function computed on long-term streamflow series. Even when hydrological models achieve acceptable scores in NSE and KGE, imbalances can still arise between overall model performance and its ability to simulate flood events, particularly flash floods. Multi-scale signatures, which refer to hydrological signatures computed at different temporal and/or spatial scales, and distributed flood modeling, which accounts for spatial variability in input variables and model parameters, are important concepts in hydrological modeling. In this study, the potential of using multi-scale signatures is explored to enhance multi-criteria calibration methods for spatially distributed flood modeling, which remains considerable challenges. We present a novel signatures and sensitivity-based calibration approach implemented into a variational data assimilation algorithm capable to deal with high dimensional spatially distributed hydrological optimization problems. It is tested on 141 flash flood pronecatchments mostly located in the French Mediterranean region. Our approach involves computing several signatures, including flood event signatures, using an automated flood segmentation algorithm. We select suitable signatures for constraining the model based on their global sensitivity with the input parameters through global signature-based sensitivity analysis (GSSA). We then perform two multi-criteria calibration strategies using the selected signatures, including a single-objective optimization approach, which transforms the multi-criteria problem into a single-objective function, and a multi-objective optimization approach, which uses a simple additive weighting method to select an optimal solution from the Pareto set. Our results show significant improvements in both calibration and temporal validation metrics, especially for flood signatures, demonstrating the robustness and delicacy of our signatures-based calibration framework for enhancing flash flood forecasting systems.

Modeling the biaxial, rate-dependent response of filament-wound FRP tubes

This work studies the rate-dependent mechanical behavior of filament-wound fiber-reinforced polymer (FRP) composite pipes. Commercially available tubes with a filament winding angle of ±55° were tested under cyclic axial compression for four loading rates. Stress relaxation under constant strain was observed as well as a dependence of stress on the strain rate. A novel modeling methodology is presented to capture the nonlinear cyclic response, including the viscoelastic behavior of the epoxy matrix and the interaction of axial and hoop strains. This is accomplished by defining an element configuration with separate elements for the epoxy matrix and the glass fibers. The nonlinear and viscoelastic behavior is incorporated using the generalized Maxwell model. A machine learning (ML) calibration framework is adapted for this study and used to calibrate the nonlinear and viscoelastic properties for the analytical model using a convolutional neural network (CNN). The CNN is trained to identify and understand the interdependencies among the model parameters. The calibrated model parameters are used to simulate the experimentally measured response of the FRP tubes and were found to be applicable across the range of strain rates. The proposed modeling methodology accurately predicted the axial stress and hoop strain time histories as well as the rate-dependent stress relaxation during constant axial strains. The accuracy capturing the measured stress-strain responses demonstrated the synthetic dataset was adequate for training the CNN without requiring additional experimental data.

A calibration method for SWMM to mitigate the impact of the structure defect without considering runoff on building walls

In this study, we propose a hypothesis that an automatic calibration framework can address modeling uncertainties in the Storm Water Management Model (SWMM) due to structural defects that result in the inability of the model to account for runoff generated on building walls from wind-driven rain. To test this hypothesis, we introduce a rainfall error model into the calibration framework to indirectly consider the effects of inclined wind-driven rain on building walls. We couple the optimization algorithm Differential Evolution Adaptive Metropolis (DREAM) with SWMM using newly developed API functions. To demonstrate the effectiveness of the framework, we conduct a case study in Guangzhou, China and assess the impacts of rainfall uncertainty on model parameter estimations and simulated runoff boundaries. The results show that the framework can improve the average Nash–Sutcliffe index of selected events by more than 5%. It also captures peak flow more accurately. This framework contributes to the theory of SWMM calibration by accounting for structural defects and considering rainfall uncertainty.

Real-Time Monitoring of Weather Radar Network Calibration and Antenna Pointing

Abstract We present an integrated framework that leverages multiple weather radar calibration and monitoring techniques to provide real-time diagnostics on reflectivity calibration, antenna pointing, and dual-polarization moments. This framework uses a volume-matching technique to track the absolute calibration of radar reflectivity with respect to the Global Precipitation Measurement (GPM) spaceborne radar, the relative calibration adjustment (RCA) technique to track relative changes in the radar calibration constant, the solar calibration technique to track daily change in solar power and antenna pointing error, and techniques that track properties of light-rain medium to monitor the differential reflectivity and dual-polarization moments. This framework allows for an evaluation of various calibration and monitoring techniques. For example, we found that a change in the RCA is highly correlated to a change in absolute calibration, with respect to GPM, if a change in antenna pointing can first be ruled out. It is currently monitoring 67+ radars from the Australian radar network. Because of the diverse and evolving nature of the Australian radar network, flexibility and modularity are at the core of the calibration framework. The framework can tailor its diagnostics to the specific characteristics of a radar (band, beamwidth, etc.). Because of its modularity, it can be expanded with new techniques to provide additional diagnostics (e.g., monitoring of radar sensitivity). The results are presented in an interactive dashboard at different level of details for a wide and diverse audience (radar engineers, researchers, forecasters, and management), and it is operational at the Australian Bureau of Meteorology. Significance Statement Weather radars, like all instruments, require maintenance and upgrades. Rainfall measurements are highly variable and sensitive to change, and this can lead to inconsistencies within a radar network. Calibration is the process to counteract those inconsistencies. Any calibration requires a fixed standard to which the changed/upgraded radar can be compared. The SCAR calibration framework presented herein makes use of several standards to retrieve a full set of diagnostics about the radar data. We apply these techniques over the entire Australian weather radar network and demonstrate that, by using this integrated approach, absolute calibration can be achieved to within 1 dBZ of reflectivity, antenna pointing can be monitored within 0.1°, and the various measurements of the radars can be quality controlled.

A matrix-solving hand-eye calibration method considering robot kinematic errors

Hand-eye calibration is one of the practical solutions for enhancing the robotic machining accuracy of complex freeform surfaces. The hand-eye parameters identification, however, is inevitably affected by the robot kinematic errors. In this paper, a matrix-solving hand-eye calibration method is proposed to overcome the challenge by taking into consideration the position constraint based robot kinematics identification. In the calibration framework, the calibration of hand-eye parameters is performed for solving the rotation matrix and translation vector, and the corresponding calibration strategies are proposed based on the axial translation under robot end-effector coordinate frame and the robot relocalization, respectively. Then the hand-eye parameters are corrected in the form of matrix compensation by using the differential matrix of the robot pose errors. A uniform sampling method for robot pose screening is further designed to automatically acquire spherical point clouds by the laser scanner when the robot poses change. Based on these steps, the combined identification model of robot kinematic parameter errors and hand-eye parameters is developed according to the position constraint that the position of the criterion sphere in the robot base coordinate frame is invariant. By virtue of the least square method, the kinematic parameter errors of the robot are calculated to compensate the hand-eye calibration accuracy. The calibration experiments are investigated by robotic scanning of the criterion sphere and two typical complex workpiece of car bodywork and blade, and the results are particularly compared with the uncompensated method and the state-of-the-art methods, showing the effectiveness and applicability of the proposed method.

Bayesian uncertainty quantification framework for wake model calibration and validation with historical wind farm power data

AbstractThe expected growth in wind energy capacity requires efficient and accurate models for wind farm layout optimization, control, and annual energy predictions. Although analytical wake models are widely used for these applications, several model components must be better understood to improve their accuracy. To this end, we propose a Bayesian uncertainty quantification framework for physics‐guided data‐driven model enhancement. The framework incorporates turbulence‐related aleatoric uncertainty in historical wind farm data, epistemic uncertainty in the empirical parameters, and systematic uncertainty due to unmodeled physics. We apply the framework to the wake expansion parametrization in the Gaussian wake model and employ historical power data of the Westermost Rough Offshore Wind Farm. We find that the framework successfully distinguishes the three sources of uncertainty in the joint posterior distribution of the parameters. On the one hand, the framework allows for wake model calibration by selecting the maximum a posteriori estimators for the empirical parameters. On the other hand, it facilitates model validation by separating the measurement error and the model error distribution. In addition, the model adequacy and the effect of unmodeled physics are assessable via the posterior parameter uncertainty and correlations. Consequently, we believe that the Bayesian uncertainty quantification framework can be used to calibrate and validate existing and upcoming physics‐guided models.

Correction: An asynchronous parallel high-throughput model calibration framework for crystal plasticity finite element constitutive models

Correction: An asynchronous parallel high-throughput model calibration framework for crystal plasticity finite element constitutive models

Deterministic and probabilistic-based model updating of aging steel bridges

Numerical modeling is a very useful tool in different fields of bridge engineering, such as load-carrying capacity assessment or structural health monitoring. Developing a reliable computational model that accurately represents the actual bridge mechanical behavior entails advanced FEM-based modeling complemented by a comprehensive experimental campaign that provides the necessary supporting information and allows validating simulation outcomes. This paper proposes a unified approach aimed at the experimental characterization and FE model updating of aging steel bridges. It first involves the realization of an extensive experimental campaign aimed at the bridge's geometrical, material, and dynamic behavior characterization. Then, a model calibration framework is developed, where deterministic (optimization) and probabilistic (Bayesian inference) approaches are employed, and techniques such as global variance-based sensitivity analysis and Kriging-based surrogate modeling are further implemented in order to enhance the identification process and reduce the overall computational burden. The methodology has been validated in a historical riveted steel bridge in O Barqueiro, north of Galicia, Spain. The results show a good agreement in the identified model parameter values and a noticeable correlation between numerical and experimental modal properties, with an average relative error in frequencies of 0.34% and 0.44% for the deterministic and probabilistic approaches and an average MAC (Modal Assurance Criterion) ratio of 0.96.

Calibration of patient-specific boundary conditions for coupled CFD models of the aorta derived from 4D Flow-MRI.

Introduction: Patient-specific computational fluid dynamics (CFD) models permit analysis of complex intra-aortic hemodynamics in patients with aortic dissection (AD), where vessel morphology and disease severity are highly individualized. The simulated blood flow regime within these models is sensitive to the prescribed boundary conditions (BCs), so accurate BC selection is fundamental to achieve clinically relevant results. Methods: This study presents a novel reduced-order computational framework for the iterative flow-based calibration of 3-Element Windkessel Model (3EWM) parameters to generate patient-specific BCs. These parameters were calibrated using time-resolved flow information derived from retrospective four-dimensional flow magnetic resonance imaging (4D Flow-MRI). For a healthy and dissected case, blood flow was then investigated numerically in a fully coupled zero dimensional-three dimensional (0D-3D) numerical framework, where the vessel geometries were reconstructed from medical images. Calibration of the 3EWM parameters was automated and required ~3.5 min per branch. Results: With prescription of the calibrated BCs, the computed near-wall hemodynamics (time-averaged wall shear stress, oscillatory shear index) and perfusion distribution were consistent with clinical measurements and previous literature, yielding physiologically relevant results. BC calibration was particularly important in the AD case, where the complex flow regime was captured only after BC calibration. Discussion: This calibration methodology can therefore be applied in clinical cases where branch flow rates are known, for example, via 4D Flow-MRI or ultrasound, to generate patient-specific BCs for CFD models. It is then possible to elucidate, on a case-by-case basis, the highly individualized hemodynamics which occur due to geometric variations in aortic pathology high spatiotemporal resolution through CFD.

A multi-objective calibration framework for capturing the behavioral patterns of autonomously-driven vehicles

Calibration of car-following (CF) models is considered a very important task towards reproduction of individual vehicle behaviors and collective traffic phenomena. In most works, calibration is performed on trajectories of leading and following vehicles either on speed or spacing quantities, with the combination of the two plus the acceleration quantity to be considered as the most appropriate according to the literature. With the advent of adaptive cruise control (ACC) technology, there are intrinsic behavioral differences between human- and ACC-driven vehicles that are visible in experimental observations. Some examples include constant headway for ACC-enabled vehicles, string instability, hysteretic behavior, human-like response time and others. Since calibration is performed only on basic vehicle dynamics, i.e., the spacing, speed and acceleration, even after proper parametrization we cannot be certain that CF models will be able to reproduce some or all of the above phenomena and to what extend. The aim of this work is to propose a multi-objective calibration framework validated on empirical observations of car platoons. Furthermore, it investigates if, and to what extent, the traffic dynamics and behavioral patterns of ACC-driven vehicles can be reproduced. Two state-of-the-art CF models are tested with the empirical dataset. The results indicate that the proposed framework leads to acceptable errors and comparable performance, facilitating the models to capture phenomena in a way that is not possible until now, using the calibration on basic vehicle dynamics.

An asynchronous parallel high-throughput model calibration framework for crystal plasticity finite element constitutive models

An asynchronous parallel high-throughput model calibration framework for crystal plasticity finite element constitutive models