Measurement Uncertainty Research Articles

Machine Learning‐Based Hydrofacies Classification: Effects of Noise and Regularization

AbstractThis study utilizes an unsupervised ML approach, the expectation‐maximization (EM) algorithm using Gaussian Mixture Models (GMM), to integrate near‐surface geophysics measurements for hydrofacies classification. We examined the impact of noise and noise estimation on classification across two synthetic models with varying lateral heterogeneity, simulating and inverting resistivity and seismic data with noise levels ranging from minimal to very high. The algorithm proved robust in accurately reconstructing hydrofacies when noise was correctly estimated or not significantly underestimated, showing minimal misclassification in shallow hydrofacies. However, severe underestimation of noise during inversion led to increased misclassifications and artifact‐laden hydrofacies models, especially in shallow regions. Higher lateral heterogeneity lessened the negative impact of noise, slightly improving algorithm performance when noise was correctly estimated. We also explored the influence of geophysical measurement uncertainties on classification uncertainty through hydrofacies probability maps, noting the greatest impact when noise was underestimated. Additionally, we investigated the effect of the regularization trade‐off parameter on the hydrofacies classification and show how the performance of the algorithm can be evaluated in the absence of ground truth data using the average silhouette score of the classification with data obtained from a basement complex field site. We found that moderate regularization (λ = 200) yielded the best hydrofacies model, as indicated by the highest average silhouette score. Our findings underscore the effectiveness of unsupervised ML for facies classification and emphasizes the critical role of accurate noise characterization in geophysical data processing for enhancing the integration of subsurface heterogeneity information into hydrological models.

Using a research-based assessment instrument to explore undergraduate students’ proficiencies around measurement uncertainty in physics lab contexts

Concepts and practices surrounding measurement uncertainty are vital knowledge for physicists and are often emphasized in undergraduate physics laboratory courses. We have previously developed a research-based assessment instrument—the Survey of Physics Reasoning on Uncertainty Concepts in Experiments (SPRUCE)—to examine student proficiency with measurement uncertainty along a variety of axes, including sources of uncertainty, handling of uncertainty, and distributions and repeated measurements. We present here initial results from the assessment representing over 1500 students from 20 institutions. We analyze students’ performance pre- and postinstruction in lab courses and examine how instruction impacts students with different majors and gender. We find that students typically excel in certain areas, such as reporting the mean of a distribution as their result, while they struggle in other areas, such as comparing measurements with uncertainty and correctly propagating errors using formulas. Additionally, we find that the importance that an instructor places in certain areas of measurement uncertainty is uncorrelated with student performance in those areas. Published by the American Physical Society 2024

Attenuation Measurement Using the IF Substitution Method in the Frequency Range of 9 kHz to 10 MHz

Using the intermediate frequency (IF) substitution method, this study establishes a low-frequency attenuation measurement standard operating in the 9 kHz to 10 MHz frequency range. We designed and fabricated a low-frequency attenuation measuring instrument using a mixer that converts the input radio frequency signal into an output IF of 1 kHz. When varying the attenuation of a variable attenuator under test, an attenuation of up to 100 dB was measured. The IF output of the mixer was measured using a spectrum analyzer as the receiver. The receiver's performance was evaluated using an inductive voltage divider whose measurement capability is traceable to the electrical measurement standard. The measurement uncertainty, which accounts for uncertainty sources such as the performance of the receiver, mismatch, and mixer nonlinearity, was evaluated to be 0.005 dB to 0.058 dB (k = 2) in the attenuation range of 10 dB to 100 dB.

Preparation and certification of benzoic acid reference material for calorimetry analysis

Benzoic acid reference material is indispensable for bomb calorimeter instrument calibration and validation of gross calorific value (GCV) analysis of any substance. In this work, we demonstrated the preparation of benzoic acid reference material through a homogeneity study, round-robin analysis, and stability study. Two-factor analysis of variance (ANOVA) test for gross calorific value in the randomly selected sub-samples of benzoic acid exhibits a lower FTS value than the Fcrit value, indicating that the samples are sufficiently homogeneous. The calculated uncertainty of between-bottle (ubb) and uncertainty of homogeneity (uhom) for GCV of benzoic acid in the sub-samples were found as 1.82 and 4.42 ​cal/g respectively. We found that the observed homogeneity (uhom (Finding)) value is lower than the assumed homogeneity (uhom (Assume)) value for the prepared benzoic acid reference material. Overall observations confirm that the sub-samples are sufficiently homogeneous. Moreover, the round-robin analysis/or inter-laboratory comparison analysis was conducted to assign the gross calorific value and determine the characterization uncertainty. The seventh order of Grubbs' analysis was done using robust estimator Alogoritm A to assign the GCV of benzoic acid. Finally, the measurement uncertainty of the assigned GCV of benzoic acid was calculated with the combined uncertainties from various sources.

Feature selection based on neighborhood complementary entropy for heterogeneous data

In practical applications, there exists a large amount of heterogeneous data consisting of a mixture of numerical and categorical data, which are more complex and difficult to handle than a single class. In recent years, feature selection for heterogeneous data has received more and more attention from scholars. However, the current research on feature selection for heterogeneous data seldom takes into account the fineness of the granularity structure and the complementary information contained in the complements of each granularity. Addressing the shortcomings of existing efforts, this paper introduces complementary entropy into the neighborhood rough set model to define an uncertainty measurement system in heterogeneous data, and constructs the corresponding feature selection algorithm. First, the neighborhood complementary entropy and its associated uncertainty measures based on neighborhood relation are defined, and the relevant theoretical properties of these uncertainty measures are investigated. Then, based on the proposed uncertainty measures, a significance function is defined to assess the importance of candidate features. Next, a feature selection method for heterogeneous data is designed using neighborhood complementary entropy. Finally, the feasibility of the method proposed in this paper is demonstrated through an experimental comparative analysis with various existing feature selection methods.

A 10 µH Inductance Standard in PCB Technology with Enhanced Protection against Magnetic Fields

Low-frequency working standards of inductance are generally uniformly wound toroids on a ceramic core. Planar inductors made using printed circuit board (PCB) technology are simple and cheap to manufacture in comparison to inductors wound on toroid cores, but they are significantly prone to the influence of external magnetic fields. In this paper, we propose the design of a PCB inductance working standard of 10 μH, consisting of a duplex system of planar PCB coils, electrostatic shielding, and an enclosure. Alongside an electromagnetic analysis and design procedure, the measurements on the manufactured prototype included the generated magnetic field, the thermal time constant of the enclosure, temperature coefficients, and its error analysis. The measurements show negligible generated magnetic fields (<1.68 nT at 7 cm, 49 mA, 10 kHz). The minimum thermal time constant of the enclosure is 1270 s and the temperature coefficient of resistance is 0.00384 1/℃. The presented method of temperature coefficient measurement using a climate chamber allows for measurements in the temperature range of 10 °C to 40 °C. In this temperature range, the results show an inductance variation of 0.05 µH at 50 kHz, while the uncertainty of inductance measurement at this frequency was 0.03 µH (k = 2).

Unveiling non-classical correlations, quantum coherence, and steering measurement uncertainty in two-qubit $$XY-\Gamma$$ spin model

Unveiling non-classical correlations, quantum coherence, and steering measurement uncertainty in two-qubit $$XY-\Gamma$$ spin model

Negation of permutation mass function in random permutation sets theory for uncertain information modeling

Negation provides a novel perspective for the representation of information. However, current research seldom addresses the issue of negation within the random permutation set theory. Based on the concept of belief reassignment, this paper proposes a method for obtaining the negation of permutation mass function in the of random set theory. The convergence of proposed negation is verified, the trends of uncertainty and dissimilarity after each negation operation are investigated. Furthermore, this paper introduces a negation-based uncertainty measure, and designs a multi-source information fusion approach based on the proposed measure. Numerical examples are used to verify the rationality of proposed method.

Calibrations, Validations, and Checks of a Dual 23 nm and 10 nm Diffusion Charger-Based Portable Emissions Measurement System (PEMS).

The upcoming Euro 7 vehicle exhaust emissions regulation includes particle number (PN) limits for all vehicles, not only those with direct fuel injection. It also sets the lower detection particle size of the PN methodology to 10 nm from 23 nm. Recently, a commercial diffusion charger-based PEMS added the possibility of switching the lower size between 23 nm and 10 nm. In this study, we assessed the dual PEMS in the calibration laboratory using diffusion flame soot or spark discharge graphite particles following the regulated procedures. Furthermore, we compared the dual PEMS with a laboratory grade system (LABS) using soot, graphite, and vehicle exhaust particles. To put the results into perspective, we added comparisons (validations) of two additional 23 nm PEMSs with LABSs over a three-year period. The results showed that the differences of the 23 nm PEMSs remained the same (around 35% underestimation) over the years and were similar to the dual PEMS. This difference is still well within the permissible tolerance from the regulation (50%). We argued that the reason is the calibration material used by the manufacturer (spark discharge graphite). We demonstrated that calibrating with combustion soot could reduce the differences. The 10 nm PEMS gave similar results but with much smaller differences, indicating that the calibration material is of less importance for the Euro 7 step. The results showed that the measurement uncertainty has not increased but rather decreased for the specific PEMS switching from 23 nm to 10 nm.

Combination of Standard Addition and Isotope Dilution Mass Spectrometry for the Accurate Determination of Melamine and Cyanuric Acid in Infant Formula.

In the melamine scandals of the early 2000s, different companies of the dairy industry cheated their products by applying chemical substances to feign a higher content of nitrogen. However, this had a severe toxic impact on the kidney health of consumers. As a result, tremendous effort was put into the prevention of further harm to the public. In the present study, a fast-screening method for the determination of melamine and cyanuric acid in infant formula was developed. While a 1D-LC approach is faster and easier to set up, a 2D-LC approach allows for a more accurate result with better selectivity and sensitivity. For both instrumental approaches, the signal ratio of the isotopologues was crucial and had a dominant effect on the results and the measurement uncertainty. For this reason, the different contributions to the measurement uncertainty were determined experimentally using Matched Standard Addition-IDMS and compared to the Exact Matching Double IDMS.

Open-set domain adaptive fault diagnosis based on supervised contrastive learning and a complementary weighted dual adversarial network

In an actual industrial environment, the complex working environment of mechanical equipment may lead to new faults in the target domain, called the open-set domain adaptation problem. Recently, open-set adaptive fault diagnosis has been extensively employed. However, most studies not only require pre-set fixed thresholds to identify unknown class features but also ignore the learning of discriminable features under specific tasks, which affects the diagnostic performance. Hence, this paper proposes a complementary weighted dual adversarial network combined with supervised contrastive learning (CWDAN-SCL) to address the open-set cross-different working fault diagnosis of bearings. Specifically, a novel complementary weighted adversarial learning strategy is designed using supervised classification and uncertainty measurement to effectively control the participation of target domain features in the domain adaptation process and achieve the alignment of shared class fault features between the source and target domains. Moreover, an adaptive unknown fault separation module is designed using an adversarial learning method to construct a hyperplane between shared and unknown class fault features in the target domain to identify unknown class faults accurately. Additionally, a supervised contrastive loss term is designed based on contrastive learning and label knowledge to improve the aggregation of fault features of the same class and enhance the model’s generalization ability in target domain diagnosis tasks. Subsequently, the efficacy and advancement of the proposed method are substantiated through experimentation on two datasets. The experimental results illustrate that the average diagnostic performance of the proposed method is 91.73 %. This study contributes a dependable diagnostic approach for ascertaining the health status of rotating machinery equipment.

An Application of Uncertainty Quantification to Efficiency Measurements and Validating Requirements through Correlating Simulation and Physical Testing Results.

Validation is a critical aspect of product development for meeting design goals and mitigating risk in the face of considerable cost and time commitments. In this research article, uncertainty quantification (UQ) for efficiency testing of an Electric Drive Unit (EDU) is demonstrated, considering confidence in simulations with respect to the validation campaign. The methodology used for UQ is consistent with the framework mentioned in the guide to the expression of uncertainty in measurement (GUM). An analytical evaluation of the measurement chain involved in EDU efficiency testing was performed and elemental uncertainties were derived, later to be propagated to the derived quantity of efficiency. When uncertainties were associated with measurements, the erroneous measurements made through sensors in the measurement chain were highlighted. These results were used for the assessment of requirement coverage and the validation of test results.

Investigating the influence of measurement uncertainty on chlorophyll-a predictions as an indicator of harmful algal blooms in machine learning models

Advancements in data availability, including high frequency, near real-time multiparameter sensors, laboratory analysis, and in-situ and remote observations, have driven the development of machine learning (ML) models for applications such as toxic Harmful Algal Bloom (HABs) monitoring. However, the performance of ML predictions is influenced by both model uncertainties due to inherent model structures and errors associated with input dataset measurements. For example, measurement uncertainty arises from sample collection, sensor drift and laboratory analysis and sample handling errors. While impacts of model uncertainty are commonly addressed using probabilistic approaches, the effect of measurement uncertainty is less studied due to the limited availability of detailed measurement information. This study focuses on assessing the impact of measurement uncertainty on the ML prediction of chlorophyll-a concentration as an index of HABs in a mesotrophic lake. Using randomized subsets of input measured datasets that mimic possible chlorophyll-a concentration distributions, the study built 1000 Random Forest (RF) and Support Vector Regression (SVR) models. An independent measured dataset was used to validate the ensemble models, allowing for model performance evaluation and the creation of prediction intervals to measure the propagated uncertainty. Our findings showed that the model predictions have MAE that ranged between 0.16 μg/l and 5.19 μg/l, and RMSE ranging between 0.20 μg/l and 7.39 μg/l. The highest uncertainty coverage of 0.71 was observed in the RF model without chlorophyll-a sensor values as a predictor. The study found that the training dataset sizes due to the high frequency and manually sampled nature influence how much measurement uncertainty is covered. The results of this study demonstrate how well ML models can capture various HABs patterns when given diverse measurement variables. Our findings will give researchers insightful information on how to lessen the impact of measurement uncertainty when using ML models as decision-support tools for HABs management.

Information-Theoretic Bayesian Inference for Multi-Agent Localization and Tracking of an Radio Frequency Target With Unknown Waveform

Abstract Information-theoretic motion planning and machine learning through Bayesian inference are exploited to localize and track a dynamic radio frequency (RF) emitter with unknown waveform (uncooperative target). A target-state estimator handles non-Gaussian distributions, while mutual information is utilized to coordinate the motion control of a network of mobile sensors (agents) to minimize measurement uncertainty. The mutual information is computed for pairs of sensors through a four-permutation-with-replacement process. The information surfaces are combined to create a composite map, which is then used by agents to plan their motion for more efficient and effective target estimation and tracking. Simulations and physical experiments involving micro-aerial vehicles with time difference of arrival (TDOA) measurements are performed to evaluate the performance of the algorithm. Results show that when two or three agents are used, the algorithm outperforms state-of-the-art methods. Results also show that for four or more agents, the performance is as competitive as an idealized static sensor network.

Metrology for sensor networks: metrological traceability and measurement uncertainties for air quality monitoring

Abstract In situ calibration of sensors delivering SI traceable measurement results still provides an open question to the design and operation of sensor networks. Particularly when addressing low-cost sensors, currently, the use of sensor networks for air quality monitoring is limited by the low or unknown accuracy of measurements that they can achieve, while the data quality of individual sensor networks is mainly derived by algorithms. Standardization bodies like DIN and CEN therefore announced the need for investigations of validation methods on gas phase species and particulate matter on the one hand side, and for the development of fully digitized quality assurance/quality control and calibration techniques for sensor networks on the other (CEN/CENELEC, Opportunity for Standardisation to Contribute to the European Partnership on Metrology EPM under Horizon Europe). This contribution concentrates on the metrological traceability of sensor networks for air quality monitoring to the international system of units (SI) based on FAIRified intra-network communications (M. Wilkinson, et al., “The FAIR guiding principles for scientific data management and stewardship,” Sci. Data, vol. 3, 2016, Art. no. 160018) and including delocalized Optical Gas Standards operated according to the digital TILSAM method (O. Werhahn, et al., The TILSAM Method Adapted into Optical Gas Standards – Complementing Gaseous Reference Materials, PTB Open Access Repository, 2021). Informed by related activities in EURAMET (Partnership project FunSNM, EMNs COO & POLMO, TC-IM 1551) (European Metrology Network Climate and Ocean Observation (COO), European Metrology Network Pollution Monitoring (POLMO), EURAMET Project TC-IM 1551, Project Database) this contribution discusses the importance of measurement uncertainties in the context of sensor networks, comprising different sensor principles and promoting an efficient uptake of state-of-the-art methods. We discuss how the sensor network case can be addressed with sensors individually using the GUM principles (Joint Committee for Guides in Metrology, Guide to the Expression of Uncertainty in Measurement (GUM), JCGM 100: 2008 (E)). For sensor network measurements becoming metrologically traceable to the SI, documented and unbroken chains of calibrations need to be implemented each contributing to the measurement uncertainty. This applies to each individual sensor of the network including the potential gold standard among them, but also to the network’s output viewed as a single entity. The contribution provides first approaches to be tested and validated that are underpinned by fundamental design strategies for sensor networks. It follows on practical applications in real world scenarios aside from model uncertainties discussed in artificial intelligence prospects.

Empirical derivation of the metallicity evolution with time and radius using TNG50 Milky Way and Andromeda analogues

Recent works use a linear birth metallicity gradient to estimate the evolution of the Fe/H profile in the Galactic disk over time, and infer stellar birth radii (R$_ birth $) from Fe/H and age measurements. These estimates rely on the evolution of Fe/H at the Galactic center ( Fe/H (0, tau )) and the birth metallicity gradient (nabla Fe/H over time --- quantities that are unknown and inferred under key assumptions. In this work, we use the sample of Milky Way and Andromeda analogues from the TNG50 simulation to investigate the ability to recover Fe/H (R, tau ) in a variety of galaxies. Using stellar disk particles, we tested the assumptions required in estimating R$_ birth Fe/H (0, tau ), and nabla Fe/H using recently proposed methods to understand when they are valid. We show that nabla Fe/H can be recovered in most galaxies to within 26<!PCT!> from the range in Fe/H across age, with better accuracy for more massive and stronger barred galaxies. We also find that the true central metallicity is unrepresentative of the genuine disk Fe/H profile; thus we propose to use a projected central metallicity instead. About half of the galaxies in our sample do not have a continuously enriching projected central metallicity, with a dilution in Fe/H correlating with mergers. Most importantly, galaxy-specific Fe/H (R, tau ) can be constrained and confirmed by requiring the R$_ birth $ distributions of mono-age, solar neighborhood populations to follow inside-out formation. We conclude that examining trends with R$_ birth $ is valid for the Milky Way disk and similarly structured galaxies, where we expect R$_ birth $ can be recovered to within 20<!PCT!> assuming today's measurement uncertainties in TNG50.

Tracking trend output using expectations data

Abstract This article proposes a new approach to measuring trend output that exploits survey data on expectations to distinguish the effects of permanent and transitory shocks and to track the time-variation in the processes underlying the determination of output. The approach is illustrated using measures of output expectations and output uncertainties based on a business survey conducted for UK manufacturing. The measures are employed in a time-varying vector autoregression (VAR) to track trend output and to provide a compelling characterization of the output fluctuations in UK manufacturing over the last 20 years.

ANALYSIS OF UNCERTAINTY IN THE DETERMINATION OF THE CRITICAL YAW RADIUS USING MONTE CARLO SIMULATION

Background: The speed of a vehicle in a curve can be estimated by analysing yaw marks, which indicate its orientation around the vertical axis. When it exceeds the "critical speed," the vehicle slides laterally, evidenced by the oscillation of the rear tires relative to the front ones. Aim: To present a methodology for reconstructing yaw accidents, with an analysis of uncertainties. Methods: This article uses Monte Carlo simulation software to address uncertainties in determining the critical yaw speed of vehicles. The radius of curvature is calculated considering the chord, average arrow, and vehicle gauge, while the road is characterized by negative superelevation and specific friction coefficient. The simulation models input variables with probabilistic distributions, analysing uncertainties related to road characteristics, vehicle properties, environmental conditions, and measurement errors. Results: The simulation results provide a detailed assessment of the uncertainties involved in determining the critical yaw speed, highlighting the importance of a probabilistic approach to ensure a more realistic and reliable assessment of vehicular safety. Discussion: When evaluating measurement uncertainties, it is essential to consider the sensitivity coefficient to ensure a precise and comprehensive analysis of uncertainty sources. Conclusions: This study highlights the importance of analysing and controlling critical variables, such as the radius of curvature and friction coefficient, to accurately determine the critical yaw speed and ensure reliable vehicular safety measures..

Optimal control technique applied to the minimization of uncertainty measurements in surveying instruments

The objective of this study was to develop an optimal control approach by numerical calculus to predict how to reduce the overall uncertainty of survey instruments unable to directly measure inaccessible points. To reach our goal, two approaches were used to attain the objective. The first was inspired by mathematical models related to three methods appropriately selected and contained in Zhuo’s work proposed in 2012. These were Remote Elevation Measurement (REM), Remote Elevation Dual Measurement (REDM), and Front-to-Back Measurement (FBM) methods whose uncertainties on the measurements of points were deduced using error propagation equations. Optimal control technique helps us to show that for the REM, the height h of the prism contributed more than 70% compared to the global uncertainty for ranges [Formula: see text] from the prism. For the REDM, when the distance between two consecutive stations increases, the weight of the contribution of the two zenith angles [Formula: see text] and [Formula: see text] tends to 50% each for [Formula: see text] close to [Formula: see text], which is to be avoided. For the FBM, the weight of the contribution during the front measurement process before is negligible. The second approach used the Swedish regulation of SIS-TS 21143:2009 which classified total stations according to types of uncertainty to compare the results given by the total station of class T3 unable to directly measure inaccessible points with the more sophisticated class T1 station with direct measurements. Thus, for small spans at the rear measurements [Formula: see text], the height [Formula: see text] of the front prism has the greatest relative contribution more than 90% for zenithal differences [Formula: see text]. This results of our analysis were convincing and provided designers with the data to minimize the overall uncertainties essential in the conception of total stations.

Validation and measurement uncertainty of analytical methods for various azo dye adulterants in Curcuma longa L.

Food adulteration involving the illegal addition of dyes to foodstuffs has become an alarming issue in recent years. This study developed and validated a high-performance liquid chromatography (HPLC)-DAD (diode array detector) method for the simultaneous determination of nine azo dyes (Butter Yellow, Sudan Orange G, Para Red, Sudan I, Sudan II, Sudan III, Sudan IV, Sudan Red 7B, and Scarlet 808). Moreover, a qualitative analysis method using liquid chromatography-tandem mass spectrometry was developed to more accurately identify peaks detected in HPLC-DAD. The calibration curve represented good linearity (r2≥0.9998) over the measured concentration range of 0.5-25mg/kg. limit of detection and limit of quantification were 0.01-0.04 and 0.04-0.12mg/kg, respectively. Accuracy and precision were 96.0-102.6 and 0.16-2.01 (relative standard deviation%), respectively. Additionally, the measurement uncertainty and HorRat value were estimated. Several Curcuma longa L. distributed in Korea were collected and monitored for azo dye contaminants. PRACTICAL APPLICATION: The proposed HPLC-DAD method represents a significant advancement in the field, offering a reliable means of quantifying azo dyes and identifying their presence even at trace levels in adulterated turmeric. This not only contributes to ensuring the safety and integrity of turmeric products but also establishes precedent for robust analytical techniques in addressing food safety challenges.