Joint deep calibration of the 4-factor PDV model

Near infrared spectral analysis is a kind of indirect measuring technology, that is, calibration model is established by using sample spectra and the reference value of the interested concentration, and then predict the interested concentration of the unknown sample by using its spectrum and calibration model. Generally, the quality of calibration model is evaluated by the deviation between the predicted value and the reference value of the measured concentration. Reference value measurement error of the calibration set samples and model error are the main error source of multivariate calibration. When there was reference value measurement error, the regression model would cause great prediction deviation. However in practical application, apparent prediction error caused by reference value measurement error is always considered as the faultiness of calibration model. So it is falsely regarded that the prediction error of multivariate calibration model would be greater than that of reference method. In this paper, effect of reference value measurement error on prediction error was analyzed, and the relationship among apparent prediction error, actual prediction error and reference value measurement error was discussed. At the same time, the applicability and rationality of a computation formula was also discussed, which could estimate the true prediction error in a more precise way. It was proved that the prediction accuracy of an excellent calibration model was greater than that of the reference method by analyzing the reference value measurement error. An experiment of glucose aqueous solution samples was given in this paper. In this example, the effect of reference value measurement error on the prediction error of multivariate calibration model was investigated, and the estimating formula of the corrected prediction error was evaluated. The conclusion could be universally applied to not only multivariate calibration MLR, PCR, PLS and so on, but also in much more complex samples.

Statistical calibration of forecasts from numerical weather prediction (NWP) models aims to produce forecasts that are unbiased, reliable in ensemble spread, and as skillful as possible. We suggest that the calibrated forecasts should also be coherent in climatology, including seasonality, consistent with observations. This is especially important when forecasts approach climatology as forecast skill becomes low, such as at long lead times. However, it is challenging to achieve these aims when data available to establish sophisticated calibration models are limited. Many NWP models have only a short period of archived data, typically one year or less, when they become officially operational. In this paper, we introduce a seasonally coherent calibration (SCC) model for working effectively with limited archived NWP data. Detailed rationale and mathematical formulations are presented. In the development of the model, three issues are resolved. These are 1) constructing a calibration model that is sophisticated enough to allow for seasonal variation in the statistical characteristics of raw forecasts and observations, 2) bringing climatology that is representative of long-term statistics into the calibration model, and 3) reducing the number of model parameters through sensible reparameterization to make the model workable with short NWP dataset. A case study is conducted to examine model assumptions and evaluate model performance. We find that the model assumptions are sound, and the developed SCC model produces well-calibrated forecasts.

Development of a combined tension–torsion experiment for calibration of ductile fracture models under conditions of low triaxiality

We consider modeling approaches to characterize solute transport in porous media, integrating them into a unique theoretical and experimental framework for model evaluation and data interpretation. To date, development of (conservative and reactive chemical) transport models and formulation of model calibration methods grounded on sensitivity‐based collection of measurements have been pursued in parallel. Key questions that remain include: For a given set of measurements, which conceptual picture of the transport processes, as embodied in a mathematical model or models, is most appropriate? What are the most valuable space‐time locations for solute concentration measurements, depending on the model selected? How is model parameter uncertainty propagated to model output, and how does this propagation affect model calibration? We address these questions by merging parallel streams of research—model formulation, reduction, calibration, sensitivity analysis, and discrimination—offering our view on an emerging framework that guides (i) selection of an appropriate number and location of time‐dependent concentration measurements given a transport model and (ii) assessment (through discrimination criteria) of the relative benefit of applying any particular model from a set of several models. Our strategy is to employ metrics to quantify the relative contribution of each uncertain model parameter to the variability of the model output. We evaluate these metrics through construction of a surrogate (or “meta”) transport model that has the additional benefit of enabling sensitivity analysis and model calibration at a highly reduced computational cost. We demonstrate the applicability of this framework, focusing on transport of reactive chemicals in laboratory‐scale porous media.

ObjectiveVital signs-based models are complicated by repeated measures per patient and frequently missing data. This paper investigated the impacts of common vital signs modeling assumptions during clinical deterioration prediction model development. Study Design and SettingElectronic medical record (EMR) data from five Australian hospitals (1 January 2019–31 December 2020) were used. Summary statistics for each observation's prior vital signs were created. Missing data patterns were investigated using boosted decision trees, then imputed with common methods. Two example models predicting in-hospital mortality were developed, as follows: logistic regression and eXtreme Gradient Boosting. Model discrimination and calibration were assessed using the C-statistic and nonparametric calibration plots. ResultsThe data contained 5,620,641 observations from 342,149 admissions. Missing vitals were associated with observation frequency, vital sign variability, and patient consciousness. Summary statistics improved discrimination slightly for logistic regression and markedly for eXtreme Gradient Boosting. Imputation method led to notable differences in model discrimination and calibration. Model calibration was generally poor. ConclusionSummary statistics and imputation methods can improve model discrimination and reduce bias during model development, but it is questionable whether these differences are clinically significant. Researchers should consider why data are missing during model development and how this may impact clinical utility.

Calibrated energy models are used for measurement and verification of building retrofit projects, predictions of savings from energy conservation measures, and commissioning building systems (both prior to occupancy and during real-time model based performance monitoring, controls and diagnostics). This paper presents a systematic and automated way to calibrate a building energy model. Efficient parameter sampling is used to analyze more than two thousand model parameters and identify which of these are critical (most important) for model tuning. The parameters that most affect the building’s energy end-use are selected and automatically refined to calibrate the model by applying an analytic meta-model based optimization. Real-time data from an office building, including weather and energy meter data in 2010, was used for the model calibration, while 2011 data was used for the model verification. The modeling process, calibration and verification results, as well as implementation issues encountered throughout the model calibration process from a user’s perspective are discussed. The total facility and plug electricity consumption predictions from the calibrated model match the actual measured monthly data within ±5%. The calibrated model gives 2.80% of Coefficient of Variation of Root Mean Squared Error (CV (RMSE)) and −2.31% of Normalized Mean Bias Error (NMBE) for the whole building monthly electricity use, which is acceptable based on the ASHRAE Guideline 14–2002. In this work we use EnergyPlus as a modeling tool, while the method can be used with other modeling tools equally as well.

Objective This study aimed to cross-validate two comparable Weibull models of prediction of age at natural menopause from two cohorts, the Scheffer, van Rooij, de Vet (SRV) cohort and the Tehran Lipid and Glucose Study (TLGS) cohort. It summarizes advantages and disadvantages of the models and underlines the need for achieving correct time dependency in dynamic variables like anti-Müllerian hormone.Methods Models were fitted in the original datasets and then applied to the cross-validation datasets. The discriminatory capacity of each model was assessed by calculating C-statistics for the models in their own data and in the cross-validation data. Calibration of the models on the cross-validation data was assessed by measuring the slope, intercept and Weibull shape parameter.Results The C-statistic for the SRV model on the SRV data was 0.7 (95% confidence interval (CI) 0.7–0.8) and on the TLGS data it was 0.8 (95% CI 0.8–0.9). For the TLGS model on the TLGS data, it was 0.9 (95% CI 0.8–0.9) and on the SRV data it was 0.7 (95% CI 0.6–0.8). After calibration of the SRV model on the TLGS data, the slope was 1, the intercept -0.3 and the shape parameter 1.1. The TLGS model on the SRV data had a slope of 0.3, an intercept of 12.7 and a shape parameter of 0.6.Conclusions Both models discriminate well between women that enter menopause early or late during follow-up. While the SRV model showed good agreement between the predicted risk of entering menopause and the observed proportion of women who entered menopause during follow-up (calibration) in the cross-validation dataset, the TLGS model showed poor calibration.

Aerothermal heating is the primary driver for launch vehicle external thermal protection system (TPS) design, except for the smaller areas exposed to high plume radiation heating rates. Aerothermal heating models are frequently calibrated via the data from cylindrical, in-flight, flush-mounted surface heat flux gauges. These sensors, typically copper or aluminum Schmidt-Boelter gauges taking advantage of the one-dimensional Fourier’s law of heat conduction, are used to measure the incident heat flux. This instrumentation, when surrounded by relatively low-conductivity materials (such as cork), can have wall temperatures significantly lower than the surrounding wall temperature. As a result of this substantial disturbance to the convective heating profile, the heat flux indicated by the gauge tends to be considerably higher (potentially by factors higher than 2) than it would have been had the calorimeter not been there. The contributors to this include thermal boundary layer disturbances and radial conductive heat transfer from the hotter insulation. These measurements, if uncorrected, can lead to highly conservative TPS designs. An effort to quantify the radial conduction effects will be presented here. A three-dimensional computational thermal math model has been developed, and includes the details of a Schmidt-Boelter gauge. These details of the model are presented. Calibration of the model was performed via comparison with data collected on flat plates exposed to an aerothermal environment in the Marshall Space Flight Center (MSFC) Improved Hot Gas Facility (IHGF). The analysis includes a study of the sensitivity of results and corrections to gauge design parameters. Analyzed parameters include gauge orientation (i.e. the direction of the high and low thermopile bead lines) with respect to flow direction, the axial location of the thermopile beads and the separation between the high and low beads, and the contact conductance from the calorimeter to the surrounding material.

Calibration of the HSPF model with a new coupled FTABLE generation method

The seismic vulnerability of urban assets exposed to earthquake hazard represents a growing concern in the engineering field due to the potential risk of collapse for the buildings, and the associated devastating consequences for the inhabitants. This concern has led to the development of novel strategies for the assessment of the seismic vulnerability of the existing buildings. An accurate evaluation of the seismic vulnerability constitutes the point of departure for the implementation of the necessary mitigation strategies aimed at increasing the seismic resilience of the cities. Seismic vulnerability assessment (SVA) at large-scale requires robust knowledge of the existing structural typologies, often biased by the intrinsic uncertainties related to local construction materials and techniques, especially in territories composed of historical and vernacular constructions. A propaedeutic stage of SVA is the definition of building taxonomies to characterize the structural features of the buildings and their main sources of vulnerabilities. This difficult task requires the availability of consistent technical data, which are commonly scarce in cities with historical constructions. This doctoral thesis contributes to the development of suitable methodologies for the systematic surveying of urban buildings in historical urban centres, by proposing four complete survey forms for timber, masonry, reinforced concrete and steel/iron structural typologies. This tool aims to improve and optimize the complex process of collecting and gathering data about the structural characteristics of the existing building stock. This thesis also contributed to the calibration of simplified numerical models for the assessment of the seismic vulnerability of traditional buildings composed of timber frame and masonry wall structures. The nonlinear behaviour is simulated through the adoption of Lumped Plasticity Models (LPM) for timber frames, and continuum Finite Element (FE) models for masonry walls. The calibration of these models is pursued based on rigorous comparisons with numerous experimental tests available in the scientific literature. The developed simplified numerical models for timber frame and masonry wall structures are used to assess the seismic vulnerability of two existing buildings located in the historical centre of the city of Valparaíso, Chile. The numerical models are employed afterwards to analyse the different structural configurations of timber-masonry buildings of the study area. The examined structural configurations differ in the organization of the resisting systems, number of storeys, and presence of structural irregularities. The different pushover analyses show the influence of the aforementioned parameters on the seismic behaviour of the buildings. This research proposes a tentative set of scores and weights for the application of the Vulnerability Index Method (VIM) to traditional buildings composed of timber frame and masonry wall structures based on a numerical investigation. The resulting VIM form is calibrated by applying the N2 method to determine the seismic performance of the representative typologies within the study area. This study is based on the results of a numerical study, unlike other available VIM forms available in the existing literature are based on past post-earthquake surveys of damaged buildings. The thesis ends in a prospective study on the seismic vulnerability by VIM of 111 buildings located in three neighbourhoods of the historical centre of Valparaíso. This is an area of special interest, as it includes two of the main evacuation routes for the city in case of a tsunami, and congregate different typologies of heritage buildings, made of timber, masonry, and reinforced concrete. The obtained results are used to create seismic vulnerability maps in a GIS environment that may be used for future works aimed at the definition of proper policies for the implementation of mitigation strategies. La vulnerabilidad sísmica de zonas urbanas expuestas a terremotos representa una creciente preocupación en el campo de la ingeniería debido al potencial riesgo de colapso de los edificios, y las consecuencias devastadoras que afectan a sus habitantes. Esta inquietud ha motivado el desarrollo de nuevas estrategias para la evaluación de la vulnerabilidad sísmica de los edificios existentes en centros urbanos. Un correcto estudio de la vulnerabilidad sísmica constituye el punto de partida para la correcta implementación de estrategias de mitigación enfocadas a aumentar la resiliencia sísmica de las ciudades. Los estudios de vulnerabilidad sísmica a escala urbana requieren un alto nivel de conocimiento respecto a las tipologías constructivas existentes, información comúnmente sesgada por las incertidumbres intrínsecas de los edificios, asociadas a los materiales y técnicas de construcción locales, especialmente en territorios compuestos por construcciones históricas o de naturaleza vernácula. Una etapa propedéutica de este tipo de estudios es la definición de una taxonomía de edificios, que permita caracterizar los sistemas estructurales existentes y detectar sus principales fuentes de vulnerabilidad. Esta tesis doctoral contribuye al desarrollo de una metodología para el levantamiento sistemático de edificios históricos en centros urbanos, basada en la formulación de cuatro formularios de encuestas para tipologías estructurales de entramado de madera, mampostería, hormigón armado y acero/hierro forjado. Otra importante contribución de esta tesis se centra en la calibración de modelos numéricos simplificados y orientados a la práctica para la evaluación de la vulnerabilidad sísmica de edificios tradicionales compuestos por muros de entramados de madera y mampostería. El comportamiento no lineal dichas estructuras es simulado mediante el uso de Modelos de Plasticidad Concentrada para muros de entramado de madera, y modelos continuos de Elementos Finitos para muros de mampostería. La calibración de estos modelos se lleva a cabo en base a rigurosas comparaciones entre diferentes ensayos experimentales disponibles en la literatura científica. Los modelos propuestos se utilizan para la evaluar la vulnerabilidad sísmica de dos edificios existentes situados en el centro histórico de la ciudad de Valparaíso, Chile. Este caso de estudio se caracteriza por su singular entorno urbano, donde predominan los sistemas estructurales compuestos por muros de entramados de madera y mampostería. Los modelos numéricos se utilizan para analizar diferentes configuraciones estructurales de los edificios de entramado de madera y mampostería localizados en tres barrios históricos del centro de la ciudad de Valparaíso. Este estudio examina diferentes configuraciones estructurales que difieren en la organización de los sistemas resistentes, número de plantas y la presencia de irregularidades estructurales. La siguiente etapa de esta investigación propone una calibración preliminar del Método del Índice de Vulnerabilidad (MIV) para edificios de entramado de madera y mampostería. Una nueva ficha de vulnerabilidad se calibra en base a una investigación numérica y en la aplicación del Método N2 para determinar el rendimiento sísmico de los edificios. Esta tesis culmina con un estudio prospectivo de la vulnerabilidad sísmica aplicando el MIV a 111 edificios ubicados en tres barrios del centro histórico de Valparaíso. Esta incluye dos de las principales rutas de evacuación de la ciudad en caso de tsunami y alberga diferentes tipologías de edificios patrimoniales, hechos de entramados de madera, mampostería y hormigón armado. Los resultados obtenidos se utilizan para crear mapas de vulnerabilidad sísmica en el entorno GIS que pueden utilizarse para llevar a cabo planes futuros para la implementación de estrategias de mitigación de la vulnerabilidad.

Studying different theoretical properties of epidemiological models has been widely addressed, while numerical studies and especially the calibration of models, which are often complicated and loaded with a high number of unknown parameters, against measured data have received less attention. In this paper, we describe how a combination of simulated data and Markov Chain Monte Carlo (MCMC) methods can be used to study the identifiability of model parameters with different type of measurements. Three known models are used as case studies to illustrate the importance of parameter identifiability: a basic SIR model, an influenza model with vaccination and treatment and a HIV–Malaria co-infection model. The analysis reveals that calibration of complex models commonly studied in mathematical epidemiology, such as the HIV–Malaria co-dynamics model, can be difficult or impossible, even if the system would be fully observed. The presented approach provides a tool for design and optimization of real-life field campaigns of collecting data, as well as for model selection.

An integrated data-model system to support monitoring and assessment of marine systems

Traffic simulation models have been widely used to evaluate design alternatives and to help decision-makers to select best design option for prevailing traffic conditions. This study focuses on application of microsimulation model to the performance assessment of Tight Urban Diamond Interchange (TUDI) located in a congested urban setting with population more than 9 million and current transport demand up to 13.5 million daily motorized trips. Geometric and operational data was collected by conducting multiple site visits. Traffic volume data showed the heterogeneous nature of traffic. Microsimulation model; VISSIM was applied and appropriateness of this model and the proposed methodology was assessed based on maximum queue length as Measure of Effectiveness (MOE). Calibration of model was done in two stages: system calibration and operational calibration. System calibration was done by reflecting the actual geometric and control conditions in model. While operational calibration was done by conducting Sensitivity Analysis (SA). MOE values from calibrated model were compared with the field values of maximum queue length. This study revealed that SA helps in selecting the most appropriate parameters and their values. Results from VISSIM show the cumulative difference of 32 % from the field observed values.

AimsIn the SUSTAIN 6 cardiovascular outcomes trial, once-weekly semaglutide was associated with a statistically significant reduction in major adverse cardiovascular events compared with placebo. To date, no studies have assessed how accurately existing diabetes models predict the outcomes observed in SUSTAIN 6. The aims of this analysis were to investigate the performance of the IQVIA Core Diabetes Model when used to predict the SUSTAIN 6 trial outcomes, to calibrate the model such that projected outcomes reflected observed outcomes, and to examine the impact of calibration on the cost-effectiveness of once-weekly semaglutide from a UK healthcare payer perspective.MethodsThe IQVIA Core Diabetes Model was calibrated to ensure that the projected non-fatal stroke event rates reflected the non-fatal stroke event rates observed in SUSTAIN 6 over a two-year time horizon. Cost-effectiveness analyses of once-weekly semaglutide versus placebo plus standard of care were conducted over a lifetime horizon using the uncalibrated and calibrated models to assess the impact on cost-effectiveness outcomes.ResultsTo replicate the non-fatal stroke event rate in SUSTAIN 6, calibration of the model through the application of relative risks for stroke of 1.07 and 1.65 with once-weekly semaglutide and placebo, respectively, was required. In the long-term cost-effectiveness analysis, the uncalibrated model projected an incremental cost-effectiveness ratio for once-weekly semaglutide versus placebo plus standard of care of GBP 22,262 per quality-adjusted life year (QALY) gained, which fell to GBP 17,594 per QALY gained when the calibrated model was used.ConclusionsThe requirement for calibration to replicate the outcomes observed in SUSTAIN 6 suggests that the reductions in risk of cardiovascular complications observed with once-weekly semaglutide cannot be solely explained by differences in conventional risk factors. Accurate estimation of the risk of diabetes-related complications using methods such as calibration is important to ensure accurate cost-effectiveness analyses are conducted.

Study on calibration model of soil water content based on actively heated fiber-optic FBG method in the in-situ test