Virtual Sensor Research Articles

Sensor placement determination for a wearable device in dual-arm manipulation tasks

To enhance the quality of life for individuals in various sectors, such as industry, healthcare, and everyday activities, the development of a highly efficient and cost-effective wearable device is essential. This paper presents a feasibility study for a general estimation methodology for human activity assessment using a single sensor placed on the back. The methodology considers the minimal variability in human body movements when carrying different weights. Our approach involves optimizing the sensor's position through a simple pre-processing technique. This technique utilizes virtual sensor position and orientation data derived from a virtual mannequin based on real inertial measurement unit (IMU) measurements obtained during two-arm manipulation tasks in a warehouse setting. The primary contribution is the optimization of a single, non-invasive sensor placement on the back to accurately assess activities involving two-hand manipulation with low variability in movement. We identified 54 different activity classes with an accuracy of 91.77% using a hybrid two-dimensional Convolutional Neural Network (CNN) combined with a Bidirectional Long Short-Term Memory (BiLSTM) network (2D CNN-BiLSTM) model. This model does not require additional sensors that measure other physical quantities, such as electromyography activity. Identifying repetitive tasks involving significant bending, stretching, and twisting, which pose risks of musculoskeletal disorders and back pain, offers a solution for designing wearable devices. The portability and autonomy of these devices are crucial, and current sensor technology meets these needs with low size, cost, and consumption. Wearable medical devices can thus be effectively used for self-monitoring health, preventive medicine, and rehabilitation.

Recurrent neural networks as virtual cavity pressure and temperature sensors in high-pressure die casting

High-pressure die casting (HPDC) is a permanent mold-based production technology that facilitates the casting of near net shape components from nonferrous alloys. The pressure and temperature conditions within the cavity impact the cast product quality during and after the conclusion of the die filling process. Die surface cavity sensors can deliver information describing the conditions at the die-casting interface. They are associated with high costs and limited service lifetimes below the achievable total cycle count of the die inserts and therefore ill-suited for industrial use cases. In this work, the suitability of long short-term memory (LSTM) recurrent neural networks (RNN) for substituting physical cavity temperature and pressure sensors virtually after the production ramp-up or at the end of the sensor service life is investigated. Training LSTMs with data of 233 casting cycles with different process parameters provides networks which are then applied to 99 further cycles. The prediction accuracy is investigated for different time interval lengths in the solidification and cooling phase. For longer time intervals, the cavity pressure prediction deteriorates, potentially due to a highly individual and hardly ascertainable buildup of casting distortion and internal stresses. Overall, however, the accuracy of the developed LSTMs is excellent for the cavity temperatures and good for the cavity pressures.

PCR virtual temperature sensor design based on system modeling and identification

The reaction results of Polymerase chain reaction (PCR) are highly sensitive to temperature accuracy, but the reaction liquid temperature is hard to be measured directly. In this paper, a virtual temperature sensor is creatively designed to directly obtain the reaction liquid temperature using a model-switching approach. This study first establishes the mathematical models of the temperature control system and obtains precise system models for the heating and cooling stages. Based on this, a virtual sensor is constructed, and a closed-loop fuzzy PID control system is established. Furthermore, the accuracy of the designed virtual temperature sensor is verified through simulation and experimentation. Simulation results show that the fuzzy PID control method has higher temperature accuracy, with a deviation of 0.23 °C, making it more suitable for Peltier systems, as Peltier systems suffer from inconsistent temperature rise and fall models. The experimental results show that the control method using the virtual sensor achieves higher precision in liquid temperature control, with an accuracy of ±0.21 °C. Finally, standard PCR procedures are performed for designed chip, and the results indicate that the temperature sensor and control system designed in this study exhibit significantly higher amplification efficiency, approximately 16 times greater than traditional control methods.

A modified capacitance tomography image reconstruction approach based on iterative shrinkage-thresholding algorithm combined with deep networks

Abstract Due to the ‘soft-field’ effect and the challenges posed by ill-posed and ill-conditioned inverse problems, it is difficult to obtain high quality images from an electrical capacitance tomography (ECT) system. To achieve both high-quality images and fast imaging speed with limited measurement data, an image reconstruction algorithm, which was initially proposed for compressive sensing, is adapted for ECT image reconstruction to optimize the ill-posed nature of its inverse problem. The proposed algorithm leverages deep learning networks inspired by the iterative shrinkage-thresholding algorithm (ISTA), thereby creating a model that is both mathematically interpretable and endowed with trainable parameters. Building upon this foundation, the conventional Landweber iteration is integrated with the ISTA-Net to refine the optimization process for ECT image reconstruction. In order to propose an effective model adapting to the actual multiphase flow characteristics and complex flow pattern changes, the training and test process is driven by a comprehensive dataset generated from dynamic simulations, rather than artificial samples of multiphase distributions. This numerical methodology simulates the dynamic measurement process of a virtual ECT sensor by coupling the gas–liquid two-phase flow field and the ECT electrostatic field. The results of the testing phase indicate that the proposed algorithm outperforms traditional ECT image reconstruction methods. Compared with the linear back projection algorithm, the average image error and gas fraction error have been reduced by 20.44% and 16.74%, respectively, while maintaining a computational speed comparable to that of the Landweber iteration. The accuracy of the new algorithm in reconstructing the two-phase interface and estimating the gas fraction has been validated by static experimental tests, showing its potential for practical application in online gas–liquid two-phase flow measurement scenarios.

Study on Strain Field Reconstruction Method of Long-Span Hull Box Girder Based on iFEM

The box girder’s condition significantly impacts the safety and overall performance of the entire ship because it is the primary stress component of the hull construction. This work used experimental research on the long-span hull box girder based on IFEM (Inverse Finite Element Method) technology to ensure the structural safety of the hull box girder. Due to the limitations of conventional experiments in this technical field, such as their reliance on finite element data and lack of input from physical tests, numerous research methods combining the strain sensing data from physical tests with the strain data from virtual sensors were conducted. The strain fields of the top plate, side plate, and bottom plate were each reconstructed in turn, and the verifier measuring points in the physical model test were used to assess the accuracy of the reconstruction results. The findings demonstrate that the top plate, side plate, and bottom plate reconstructions had relative errors of 0.24–7.86%, 0.75–8.13%, and 3.31–2.52%, respectively. This enables the reconstruction of the strain field of the long-span hull box girder using physical test data and promotes the use of iFEM technology in the field of structural health monitoring of large marine structures.

Train Trajectory-Following Control Method Using Virtual Sensors.

Trajectory-following control is the basis for the practical application of an articulated virtual rail train transportation system. In this paper, a planar nonlinear dynamics model of an articulated vehicle is derived using the Euler-Lagrange method. A trajectory-following control strategy based on the first following point is proposed, and a feedback linearization control algorithm is designed based on the vehicle dynamics model to achieve the trajectory following of the rear vehicle. Based on the target trajectory formed by the first following point and measured by virtual sensors, a vector analysis method grounded in geometric relationships is proposed to solve in real time for the desired position, velocity, and acceleration of the vehicle. Finally, a MATLAB/SIMPACK dynamics virtual prototype is established to test the vehicle's trajectory-following effectiveness and dynamics performance under lane change and circular curve routes. The results indicate that the control algorithm can achieve trajectory following while maintaining good vehicle dynamics performance. It is robust to variations in vehicle mass, vehicle speed, tire cornering stiffness, and road friction coefficient.

Development of an Adaptive Force Control Strategy for Soft Robotic Gripping

Using soft materials in robotic mechanisms has become a common solution to overcome many challenges associated with the rigid bodies frequently used in robotics. Compliant mechanisms allow the robot to adapt to objects and perform a broader range of tasks, unlike rigid bodies that are generally designed for specific applications. However, soft robotics presents its own set of challenges in both design and implementation, particularly in sensing and control. These challenges are abundant when dealing with the force control problem of a compliant gripping mechanism. The ability to effectively regulate the applied force of a gripper is a critical task in many control operations, as it allows the precise manipulation of objects, which drives the need for enhanced force control strategies for soft or flexible grippers. Standard sensing techniques, such as motor current monitoring and strain-based sensors, add complexities and uncertainties when establishing mathematical models of soft grippers to the required gripping forces. In addition, the soft gripper creates a complex non-linear system, compounded by adding an adhesive-type sensor. This work develops a unique visual force sensor trained on synthetic data generated using finite element analysis (FEA) and implemented by integrating a non-linear model reference adaptive controller (MRAC) to control gripping force on a fixed 6-DOF robot. The robot can be placed on a mobile platform to perform various tasks. The virtual FEA sensor and controller, combined, are termed virtual reference adaptive control (VRAC). The VRAC was compared to other methods and achieved comparable control sensing and control performance while reducing the complexity of the sensor requirements and its integration. The VRAC strategy effectively controlled the gripping force by driving the dynamics to match the desired performance after a limited amount of training cycles. The controller proposed in this work was designed to be generally applicable to most objects that the gripper will interact with and easily adaptable to a wide variety of soft grippers.

Light-Activated Virtual Sensor Array with Machine Learning for Non-Invasive Diagnosis of Coronary Heart Disease

HighlightsPhotoresponsive black phosphorus (BP)/Ti3C2Tx composites were synthesized by a self-assembly strategy.Enhanced gas sensitive property was achieved by visible light modulation.Light activated virtual sensor array was fabricated based on BP/Ti3C2Tx composite.Diagnosis of coronary heart disease was achieved with the help of machine learning.

Utilization of genetic algorithm in tuning the hyper-parameters of hybrid NN-based side-slip angle estimators

This paper proposes a solution to enhance and compare different neural network (NN)-based side-slip angle estimators. The feed-forward neural networks (FFNNs), recurrent neural networks, long short-term memory units (LSTMs), and gated recurrent units are investigated. However, there is a lack in the selection criteria of the architectures’ hyper-parameters. Therefore, the genetic algorithm is integrated with the NN-based estimators to find the optimal hyper-parameters for the studied architectures. The tuned hyper-parameters in this work include the number of neurons, number of layers, activation function, optimizer type, and learning rate. The objective function of the optimization problem is minimizing the root-mean-square error (RMSE) on multiple testing data. The optimal models are further included in the design of a hybrid NN estimator with Kalman filter. In the hybrid estimators, the optimal NN estimators are used as virtual sensors to correct the prediction of the side-slip angle resulting from the mathematical lateral vehicle model. Eventually, the performance of the best selected model is evaluated in terms of different metrics; mean RMSE, mean error variance, mean training time, and mean estimation time. LSTMs are found to achieve the lowest mean RMSE while being tested on highly generalized data yielding the highest training and estimation time. However, FFNNs achieve the lowest RMSE while being tested on low generalized data and the lowest training and estimation time. Meanwhile, it is observed that the hybrid estimators achieved lower RMSE with great enhancement compared to the non-hybridized ones proving the effectiveness of the proposed approach and increasing the side-slip estimation generalization ability in unknown environments with high uncertainties, which are not covered by the training dataset for the NNs estimators.

Chronic cannabis use alters the spontaneous and oscillatory gamma dynamics serving cognitive control.

Regular cannabis use is associated with cortex-wide changes in spontaneous and oscillatory activity, although the functional significance of such changes remains unclear. We hypothesized that regular cannabis use would suppress spontaneous gamma activity in regions serving cognitive control and scale with task performance. Participants (34 cannabis users, 33 nonusers) underwent an interview regarding their substance use history and completed the Eriksen flanker task during magnetoencephalography (MEG). MEG data were imaged in the time-frequency domain and virtual sensors were extracted from the peak voxels of the grand-averaged oscillatory interference maps to quantify spontaneous gamma activity during the pre-stimulus baseline period. We then assessed group-level differences in spontaneous and oscillatory gamma activity, and their relationship with task performance and cannabis use metrics. Both groups exhibited a significant behavioral flanker interference effect, with slower responses during incongruent relative to congruent trials. Mixed-model ANOVAs indicated significant gamma-frequency neural interference effects in the left frontal eye fields (FEF) and left temporoparietal junction (TPJ). Further, a group-by-condition interaction was detected in the left FEF, with nonusers exhibiting stronger gamma oscillations during incongruent relative to congruent trials and cannabis users showing no difference. In addition, spontaneous gamma activity was sharply suppressed in cannabis users relative to nonusers in the left FEF and TPJ. Finally, spontaneous gamma activity in the left FEF and TPJ was associated with task performance across all participants, and greater cannabis use was associated with weaker spontaneous gamma activity in the left TPJ of the cannabis users. Regular cannabis use was associated with weaker spontaneous gamma in the TPJ and FEF. Further, the degree of use may be proportionally related to the degree of suppression in spontaneous activity in the left TPJ.

Enhanced Porous Electrode Theory Based Electrochemical Model for Higher Fidelity Modelling and Deciphering of the EIS Spectra

Electrochemical impedance spectroscopy (EIS) is essential for non-invasive battery characterization. This paper addresses the challenge of adequate interpretation of EIS spectra, which are often complicated by overlapping internal phenomena occurring on similar time scales. We present, for the first time, a high-fidelity numerical time-domain electrochemical model that can virtually replicate experimental EIS spectra with three superimposed high-frequency semicircles, a transition to the diffusion tail at elevated imaginary values, and a tilted diffusion tail at low frequencies. These advanced features were made possible by extending state-of-the-art porous electrode model with innovative sub-models for the double layer phenomenon at the carbon black/electrolyte and metal Li-anode/electrolyte interfaces, and transport phenomena of charged species through the solid electrolyte interphase at the Li-anode interface. Additionally, we modelled the diffusion tail inclination by introducing representative active particles of varying sizes. Results from custom-made half-cells confirm the model’s ability to decipher EIS spectra more accurately compared to existing models. Moreover, innovative physics-based battery model that is capable of accurately modelling intra-cell phenomena can reveal internal states and physical parameters of batteries using measured EIS spectra. The model, therefore, also enables functionality of an advanced virtual sensor, which is an important diagnostics feature in next-generation battery management systems.

Quantitative Analysis of the Hsu-Nielsen Source through Advanced Measurement and Simulation Techniques

Abstract This paper presents the results from conducting a series of experiments with a Hsu-Nielsen Source, accompanied by corresponding numerical simulations on a solid block. The aim being to illustrate a Finite Element Analysis (FEA) approach for simulating Acoustic Emission (AE) wave propagation in a Hsu-Nielsen Source, by employing virtual sensors to enhance existing AE research methodologies. The objective was to examine and establish the actual unload rate derived from Pencil Lead Breaks (PLBs) by comparing results from simulations and experimental trials. These experiments and simulations were conducted using a solid cylindrical steel block, capturing the propagating Acoustic AE waves from both sources over a two-second span. When comparing the experimental data with the simulation results, it is evident that replicating the structure of an impulsive AE source is feasible for brief durations. Furthermore, both the experimental and simulated signals on the steel cylinder displayed comparable patterns in the initial 25-30 µs. The methodology presented in this study demonstrates the effectiveness of Finite Element Analysis (FEA) in precisely identifying the specific modes present in AE wave propagation, including the actual unload rates affecting the AE signals recorded.

Novel optimal sensor placement method towards the high-precision digital twin for complex curved structures

The complex shape of the structure and the new needs for high-precision in digital twin modeling pose challenges for sensor placement optimization. A novel optimal sensor placement towards the high-precision digital twin (OSP-HDT) method is proposed for complex curved structures. It comprises three key aspects. Firstly, leveraging the spatial dimensionality reduction method, the complex curved surface is simplified into a planar representation. Subsequently, candidate sensor placement points can be easily identified by dividing the background mesh in the plane and screening them within the curved surface. These candidate points are then binary encoded to facilitate the subsequent optimization. Secondly, the method collects result data from the finite element model, treating it as virtual sensor data. Using this data, a surrogate model is constructed and then the objective function is formulated based on both the global and local critical areas precision of the surrogate model. Thirdly, the sensor placement optimization model is constructed, followed by optimization design using the efficient multi-objective covariance matrix adaptive evolutionary strategy. Through the steps above, the optimal sensor placement can be identified. To validate the proposed OSP-HDT method, an experiment is conducted on an S-shaped variable cross-section stiffened shell, with the construction of the corresponding digital twin. Compared to the uniform placement with an equivalent number of sensors, the OSP-HDT method demonstrated a significant 9.0% improvement in global precision and a remarkable 62.1% enhancement in local precision of critical areas. Furthermore, when compared to the random sensor placement strategies, the OSP-HDT method exhibited a 20.5% increase in global precision, together with a 44.2% increase in the local precision. Notably, even when compared to the full sensor placement, the OSP-HDT method can maintain comparable local precision, while significantly reducing the number of sensors by 77.6%. The above comparison indicates that the proposed OSP-HDT method can build a digital twin model with higher global and local precision for complex structures.

Investigation of sorptive interactions between volatile organic compounds and supramolecules at dynamic oscillation using bulk acoustic wave resonator virtual sensor arrays

Supramolecules are considered as promising materials for volatile organic compounds (VOCs) sensing applications. The proper understanding of the sorption process taking place in host-guest interactions is critical in improving the pattern recognition of supramolecules-based sensing arrays. Here, we report a novel approach to investigate the dynamic host-guest recognition process by employing a bulk acoustic wave (BAW) resonator capable of producing multiple oscillation amplitudes and simultaneously recording multiple responses to VOCs. Self-assembled monolayers (SAMs) of β-cyclodextrin (β-CD) were modified on four BAW sensors to demonstrate the gas-surface interactions regarding oscillation amplitude and SAM length. Based on the method, a virtual sensor array (VSA) type electronic nose (e-nose) can be realized by pattern recognition of multiple responses at different oscillation amplitudes of a single sensor. VOCs analysis was realized respectively by using principal component analysis (PCA) for individual VOC identification and linear discriminant analysis (LDA) for VOCs mixtures classification.

Virtual sensor-based proxy for black carbon estimation in IoT platforms

Black carbon (BC) has been under the spotlight of research during the last few years due to its non-regulation, its role in air pollution, and its hazardous effects. Given the high cost of the instrumentation needed to measure BC concentrations, data-driven techniques have been adopted to implement proxies that provide BC measurements from other sensor measurements. These sensors may present data quality issues due to maintenance actions, loss of data, or relocation, among others. In this paper, we propose a data-driven proxy model for BC estimation that is powered by a hybrid sensor array, including physical and virtual sensors created from machine learning techniques and governmental air quality monitoring networks. Therefore, the proposed method provides an accurate alternative to traditional data-driven BC proxies in scenarios where some physical sensors are unavailable. The results show how a BC proxy can be partially implemented using virtual sensors, obtaining only an increase in the estimation error of around 4%, allowing the estimation of BC levels even when some physical sensors are absent.

Mitigating Ill-posedness in Parameter Estimation under Sparse measurement for Linear time-varying systems employing Virtual Sensor Responses

Monitoring linear time-varying (LTV) systems using model-based approaches typically requires dense instrumentation and high-fidelity support models, resulting in significant computational and financial costs. Bayesian filtering-based methods adopt a joint state-parameter estimation approach for LTV system monitoring wherein states/parameters are observed as well as inferred from the measurements. Eventually, sparsity in measurement while dealing with high-fidelity models can aggravate the ill-posedness in the estimation diverging the estimation to impractical or no solutions. To enhance estimation resolution and precision, an alternative approach can be supporting estimation with virtual sensor measurements sampled from future time steps. Virtual measurements are in fact measurements sampled from a future time step relative to the time at which estimation is sought. To leverage virtual measurements, the estimation needs to be time-delayed, enabling the observation of states/parameters through measurements taken at both current and subsequent time steps, thereby alleviating the ill-posedness. This method has been explored for an LTV spring-mass-dashpot system, where the numerical investigation utilizes an interacting filtering environment to estimate states and parameters in the presence of sparse measurements. The study has shown how incorporating additional information can stabilize the estimation process, leading to improved estimates for both states and parameters.

НЕЙРОСЕТЕВЫЕ ТЕХНОЛОГИИ В СОВРЕМЕННЫХ АВТОМАТИЗИРОВАННЫХ СИСТЕМАХ УПРАВЛЕНИЯ

Modern production needs to increase efficiency and reduce costs, thereby introducing new technolo-gies, including complex nonlinear and ill-formalized processes to be controlled. Therefore, the de-mand for new high solutions. "Demand creates supply," so many of the major suppliers of measuring and control systems offer a wide range of solutions based on artificial intelligence. The paper reveals the concept of one of the core modules of modern automated control systems – virtual sensor, and also describes the solutions developed on this technology. The analysis of the terminology associat-ed with the virtual sensor. The conclusions about the relevance of research and applications in to-day's production decisions based neural network technology

Virtual Sensor for Estimating the Strain-Hardening Rate of Austenitic Stainless Steels Using a Machine Learning Approach

This study introduces a Multiple Linear Regression (MLR) model that functions as a virtual sensor for estimating the strain-hardening rate of austenitic stainless steels, represented by the Hardening Rate of Hot rolled and annealed Stainless steel sheet (HRHS) parameter. The model correlates tensile strength (Rm) with cold thickness reduction and chemical composition, evidencing a robust linear relationship with an R-coefficient above 0.9800 for most samples. Key variables influencing the HRHS value include Cr, Mo, Si, Ni, and Nb, with the MLR model achieving a correlation coefficient of 0.9983. The Leave-One-Out Cross-Validation confirms the model’s generalization for test examples, consistently yielding high R-values and low mean squared errors. Additionally, a simplified HRHS version is proposed for instances where complete chemical analyses are not feasible, offering a practical alternative with minimal error increase. The research demonstrates the potential of linear regression as a virtual sensor linking cold strain hardening to chemical composition, providing a cost-effective tool for assessing strain hardening behaviour across various austenitic grades. The HRHS parameter significantly aids in the understanding and optimization of steel behaviour during cold forming, offering valuable insights for the design of new steel grades and processing conditions.

Harnessing NOx emission management: A virtual sensor model for natural gas power generation engines with active pre-chamber

This paper presents a NOx Sensor machine learning model for a highly efficient 2MW power generation lean burn gas engine with an active pre-chamber. As gas power generation continues to evolve, engine efficiency and the accurate measurement of NOx emissions has taken on greater importance. However, the working life of existing NOx sensors pales in comparison to the lifecycle of an engine. This paper studies a series of machine learning NOx sensor models and carries out a comparative analysis. We introduce a predictive analytics model used for monitoring the emissions of active pre-chamber gas engines. This real deployment aims to detect the deviations in emissions across an entire fleet of gas engines, even when no physical sensors are installed in the engines. Finally, a model has been implemented that allows the prediction of NOx values, in engines that did not have them initially, with a 5–10 % precision. For example, with NOx at 120 ppm, the virtual sensor provides values close to +-10 ppm. It has been observed that such features as the setpoint value of the active pre-chamber gas pressure, in a lean burn gas engine with active pre-chamber, are among those parameters that have the greatest impact on the model estimation.

Development of a semi-empirical physical model for transient NOx emissions prediction from a high-speed diesel engine.

With emissions regulations becoming increasingly restrictive and the advent of real driving emissions limits, control of engine-out NOx emissions remains an important research topic for diesel engines. Progress in experimental engine development and computational modelling has led to the generation of a large amount of high-fidelity emissions and in-cylinder data, making it attractive to use data-driven emissions prediction and control models. While pure data-driven methods have shown robustness in NOx prediction during steady-state engine operation, deficiencies are found under transient operation and at engine conditions far outside the training range. Therefore, physics-based, mean value models that capture cyclic-level changes in in-cylinder thermo-chemical properties appear as an attractive option for transient NOx emissions modelling. Previous experimental studies have highlighted the existence of a very strong correlation between peak cylinder pressure and cyclic NOx emissions. In this study, a cyclic peak pressure-based semi-empirical NOx prediction model is developed. The model is calibrated using high-speed NO and NO2 emissions measurements during transient engine operation and then tested under different transient operating conditions. The transient performance of the physical model is compared to that of a previously developed data-driven (artificial neural network) model, and is found to be superior, with a better dynamic response and low (<10%) errors. The results shown in this study are encouraging for the use of such models as virtual sensors for real-time emissions monitoring and as complimentary models for future physics-guided neural network development.