Pressure ulcers are a particularly high incidence of chronic trauma. When a superficial wound is visible, the underlying wound is often serious, so early detection of pressure ulcers is critical. Currently there is still no technique or real-time monitoring system that can visualize progressive tissue damage of pressure ulcers. Electrical impedance tomography (EIT) is a functional imaging technique that can diagnose the health of the tissue by visualizing the distribution of bioelectrical impedance parameters. Based on the differences between the electrical properties of pressure ulcers and normal tissue, a non-invasive pressure ulcer depth detection method based on EIT is proposed. Given surface voltage data measured on an open rectangular electrode array, the conductivity distribution under the skin surface was reconstructed to obtain pressure ulcer depth information which is expected to visualize progressive tissue damage. Based on EIT combined with flexible sensor arrays, finite element simulation models and tissue-mimicking agar phantom models were established to verify the effectiveness of the EIT-based depth detection method of pressure ulcers. The effects of time-difference imaging and frequency-difference imaging on the depth detection of pressure ulcer were compared. Both simulation and physical phantom experiments demonstrated the feasibility of EIT for detection of early pressure ulcer depth.
This paper presents the application of a dielectric resonator sensor to characterize glycerin solutions. Air and nine different concentrations were measured within a relative permittivity range from 1 to 78.3. Principal Component Analysis (PCA) and Support Vector Machine (SVM) were used to perform automatic classification with an 100% accuracy and the regression of both concentration and permittivity with a RMSE of 0.34% and 0.287 respectively.
IoT condition monitoring is becoming increasingly widespread. With the improvement of the performance of embedded microcontrollers, there is the possibility to increase the quantity of processed data as well as to process signals having higher bandwidths and resolution. An important aspect of IoT condition monitoring is the possibility of embedding the monitoring system in the monitored object as well as low cost. In this work the effect of embedding commercial MEMS accelerometer in filled resin is studied in terms of frequency response as a function of the geometry of the case and of the mechanical property of the epoxy resin. Resin embedding of accelerometers can be exploited in mechanical devices such as motors, gearboxes, actuators, where ad hoc housings can be realized in the structure to accommodate vibration sensors. In particular, the response of a low cost, wideband triaxial accelerometer, realized with three monoaxial devices ADXL 1005, was studied with finite element simulations and validated with measurements, adopting different filled epoxy resins.
Artificial intelligence (AI) has been widely adopted in industrial applications, making machine learning (ML) on edge devices essential. However, the power consumption of edge devices often limits their computational capabilities. In this study, we aimed to address this challenge by utilizing feature extraction (FE) techniques, explicitly considering domain shift scenarios where the data characteristics may change in real-world applications. We introduced a simple but effective FE method, the time-frequency feature extractor (TFEx). By using AutoML and performing over 6000 cross-validations on multiple time series datasets, we compared TFEx with six other FE methods. Our results showed that TFEx was the most effective method in reducing the dimensions of raw data while maintaining accuracy. Additionally, the structure of TFEx makes it suitable for implementation on edge devices, and even a simplified version can be used in many cases without significant loss of accuracy.
A simple digital interface circuit suitable for a three-wire resistance thermometer is reported in this article. The digitizer provides a linear transfer characteristic without dependence on the connecting wire resistances. Besides, the digital interface produces the output using two conversion cycles. The methodology of the digital interface circuit is mathematically derived and analyzed. Later, the effect of dominant error sources is also detailed. The performance of the circuit is initially verified using simulation studies as well as experimental studies. A linear output characteristic is obtained during these tests, and maximum output-nonlinearity does not exceed 0.25 %. Finally, the experimental results of the proposed circuit are compared with the theoretical and simulation results, and the prior art.
Horizontal liquid-liquid two-phase flows have universally existed in important industrial production fields such as meteorological, chemical, and aerospace fields. Active flow pattern control is of great significance to reveal the mechanism of flow pattern transitions, so that flow parameters under different flow patterns can be measured more easily and the measurement accuracy can be improved. In this paper, we carried out an experiment of horizontal liquid-liquid two-phase flows. A cylindrical bluff body was inserted into the pipeline to control the flow patterns. Flow images of stratified flow (ST) and stratified wavy flow (SW), and stratified flow with mixing at interface flow (ST&MI) were obtained based on the planer laser-induced fluorescence (PLIF) and particle image velocimetry (PIV) measurement system. Interface height and aqueous phase velocity distribution were extracted and the vortex structure in the aqueous phase was accurately detected based on the third generation of vortex identification method. The morphological evolution characteristics of vortices under different flow conditions were studied, and the interaction regularity among vortices, droplets, and the liquid-liquid interface was investigated.
Many industrial, medical, automotive, and consumer applications increasingly demand highly accurate pressure sensors to operate reliably. However, a variety of real-world conditions, between which prolonged exposure to high temperatures, cause sensor accuracy to drift. This article proposes a very tiny neural network (NN)-based compensation system that correct the drift measurement in real time. The system is capable of compensating for drift accuracy from 1.68 to 0.1471 hPa, reducing the variance from 0.0115 to 3.5x10 -4 in worst case conditions. To evaluate the integration of the system into the sensor digital logic, a hardware accelerator has been designed for the processing element required to run the proposed NN. By using CMOS standard cells with 130 nm TSMC technology, synthesis reports an area occupation of 0.0373 mm 2 and a power consumption of 1.07 μW. The integration of the system into sensor circuitry will allow for a more robust and reliable intelligent pressure sensor.
This paper proposes a method to estimate the frequency when only two signal periods of a multi-frequency signal is acquired using the nonparametric estimation approach in the frequency domain. In estimation, the weights of two Rife-Vincent I windows are used on the same set of samples and after that a combination of two successive DFT amplitude coefficients are put in the quotient. For reduction of the harmonic and the interharmonic leakages a procedure can be added to determine the periodic part of the signal. The obtained results show a very good compromise between fast and accurate estimation.
In recent decades, polymer-based additively manufactured (AM) products have seen a significant increase in utility in medical, aerospace, and textile industries, to name a few. AM has the capability of rapidly producing low-cost and customized products. However, on-line inspection of the polymeric filaments used to print a part, for quality control purposes, remains a challenge. It has been shown that the presence of a small amounts of moisture in the polymeric feedstock has the potential to significantly degrade the quality and strength associated with the final printed product. This paper investigates and compares the efficacy of three near-field millimeter wave probes for detecting absorbed moisture by the polylactic acid (PLA) filaments used in polymeric AM. The probes are designed to concentrate their respective electric fields in such a way to spatially produce high resolution and detection sensitivities. These probes include a standard open-ended rectangular waveguide probe and two others using the same probe but with dielectric inserts that tend to concentrate the electric field distribution to a smaller region than the waveguide aperture. Numerical electromagnetic (EM) simulations were performed at Ka-band (26.5-40 GHz) using CST Studio Suite ® , and the results showed that the open-ended rectangular waveguide is sufficiently sensitive to detect ~0.52% moisture level (by weight) absorbed by the PLA filaments, and the modified probes did not help in improving the detection sensitivity, even though they produced a higher spatial resolution. The efficacy of the open-ended rectangular waveguide probe for detecting a slight amount of moisture present in PLA filaments was also corroborated by measurements.
Deterioration of knee joint function greatly affects quality of life. Therefore, technology is needed that accurately, inexpensively, and easily measures knee joint function. As the first step for knee deterioration assessment in daily walking, we developed estimation models for six parameters that capture the features of healthy knee movement by using an in-shoe motion sensor (IMS). We recruited 72 healthy participants to construct the estimation models. Using intra-class correlation coefficients (ICCs), we showed that five of the six models achieved " fair" or "good" agreement as characteristics of walking behavior by healthy people. We demonstrated the possibility of estimating knee behaviors as the specific flexion knee angles in only using foot-motion via single IMS for the first time. It will be helpful to make knee condition monitoring in daily living available easier.
The increasing installation of renewable energy sources threatens the correct functioning of power systems, affecting especially power quality and power system inertia. To tackle these issues, a possible solution consists of the placement of Phasor Measurement Units (PMUs) and Power Quality (PQ) meters which could enhance the system observability and real-time control. Given the different characteristics and working principles of these devices, in this paper, we focus on the interoperability of PMUs and PQ meters. To achieve this, a distribution network including an extensive penetration of renewable energy sources is simulated, jointly with PMU and PQ meters. The correlation of their measurements is studied, arising issues in the aggregation and comparison of different measurements not only in high harmonic distortion but also in low-inertia conditions.
Railway turnout is critical equipment for changing trains' directions, which directly impacts the safety and efficiency of operation. This study presents a novel abnormal detection method for railway turnouts with convex hull-based one-class tensor machine (CH-OCSTM) and monitoring signal images. As opposed to existing methods, it fully preserves the spatial structure and profile information in both data processing and model-building processes. Besides, a novel tensor-form classifier called CH-OCSTM is developed to improve the one-class support tensor machine (OCSTM)'s limitations in high computing complexity. First, the one-dimensional original time-series signals are converted into two-dimensional images by data preprocessing and 2D representation. Next, the feature tensor is calculated by the CANDECOMP/PARAFAC decomposition method with the curve image data. Then, the CH-OCSTM model is built with the extracted feature tensor to implement the abnormal detection. The performance of the proposed method is evaluated and tested on two real-world operational current and power datasets. Experimental results show that the proposed method performs better than other existing approaches in accuracy and recall.
Wireless Power Transfer (WPT) is a promising technique of extending the battery lifetime of battery-powered enddevices (EDs) without getting physical contact between the power transmitter and the ED. At the same time, many of these devices use a long range radio technology to report data of their measurements. Both WPT transmitters and long range radio technologies for Internet of Things (IoT) devices, such as LoRa and Mioty, operate at the same Industrial, Scientific, and Medical (ISM) bands in the sub-GHz spectrum. This paper presents an interference analysis between the two technologies and provides evidences of substantial levels of interference around the US915 Central Frequency (CF) through a number of lab experiments. The results reveal that a number of conditions exist which allow collision-free transmissions (or a very low collision probability) when an ED is transmitting data and receiving energy at the same time.
Measuring the human gaze is an important area of research due to this measurement's ability to give insight into what or where a person is focused and/or paying attention to. However, gaze has been very challenging to measure effectively and then convert into a metric. The problem of measuring human gaze is challenging in the context of dynamic environments with motion of the subject or the environment itself. One domain of research that has sought these gaze-related attention metrics has been the area of automotive driver assessment. Being able to understand if a driver is looking at relevant areas as well as scanning the road for hazards is a valuable metric to evaluate if an individual is fit to drive. Eye-tracking glasses measure where a person is looking relative to their head position but do not map this information against important regions within the visual field. This paper provides a computationally scalable method to identify relevant regions within a dynamic visual field and allow for the measurement of what a driver is focused on, reducing the need for extensive manual segmentation. The paper provides a method of identifying the windshield and other key regions within a motor vehicle typical for a driver's field of view. The identification of key regions was accomplished through the application of convolutional neural networks (CNNs) with a Dice score of 0.9404 The model is then shown to allow for the assessment of visual focus for drivers.
Modern industrial systems require some important features such as clock synchronization, deterministic behavior, low latency, and well as high scalability, flexibility, and reliability. Time-sensitive networking (TSN) is the perfect candidate to meet these requirements since it is meant to seamlessly provide real-time performance both on wired and wireless networks. To achieve this, TSN deeply relies on the capability of synchronizing to a primary clock across the network. In this paper, we provide a measurement approach for end-to-end time synchronization in a wired and wireless heterogeneous TSN-capable network. The paper provides insight into a set of features available on commercial–grade TSN-ready devices, useful to measure time synchronization error with high accuracy and repeatability. Moreover, the paper also describes preliminary experimental results that validate the proposed measurement methodology.
A dynamic image reconstruction method considering the spatiotemporal evolution characteristics of time-varying distribution is proposed for electrical resistance tomography (ERT). The dynamic inversion problem of ERT is constructed by state-space modeling method with state evolution and observation update equations, and is solved by Kalman filter. To accurately describe the state evolution process of time-vary parameters, the latent variable based statistical modeling method is proposed to construct the state evolution equation. The potential characteristics of the state parameters in the dynamic change process are fully explored and characterized from the data with multivariate regression methods. Numerical and experimental results show that the proposed dynamic image reconstruction method can improve the imaging quality of ERT for time-varying distribution.
Field programmable gate arrays (FPGAs) based time-to-digital converters (TDCs) have been broadly investigated and widely used in recent years. However, voltage variations on the power distribution network (PDN) of FPGA are significantly affecting and sometimes devastating to the TDC performance. In this paper, we experimentally measure the performance changes of the tapped delay line (TDL) based TDC under conditions that the PDN has static and transient voltage variations and analyze the voltage influences theoretically. The test results show 0.2%/mV degradation of the TDC resolution with voltage drop and 12%1mV deterioration of the average RMS precision with voltage deviation from typical. Besides, performance degradation due to transient voltage fluctuations caused by switching activities of adjacent digital circuits is also experimentally demonstrated. In particular, the RMS precision exhibits periodic deterioration with the measured time interval, which can be reasonably explained in theory. For the influence coming from static voltage variations, we find the performance change can be greatly suppressed by the dynamic calibration method. However, for the influence coming from voltage transients, the performance degradation is inevitable, which raises a warning for FPGA management, especially for multi-tenant FPGAs where multiple untrusting cloud users simultaneously reside in a single FPGA. The test results and theoretical analysis in this paper can guide system optimization and real-time correction to reduce the impact of voltage transients.
Electrical capacitance tomography (ECT) may be used to visualize a cross-section distribution of solids concentration in a gas-solids flow. However, it is difficult to eliminate the effect of particle charging on capacitance measurements because it is difficult to separate the measurement noise from capacitance measurements. In this research, a method based on robust principal component analysis (RPCA) is introduced and applied to recover the corrupted data induced by electrostatic sparks. In this approach, recovering the corrupted data is recast as recovering a low-rank matrix from the noisy matrix constructed by time-series data. An alternating direction method of multipliers is applied to obtain the low-rank data matrix and the sparse error matrix. The performance of RPCA is evaluated using simulation and experimental data. It has been shown that by applying the proposed technique in the measurement space, significant improvement in image quality can be achieved.
The guided wave testing method cannot accurately obtain the defect's normal shift at the rail web in existing research. Therefore, we investigate the normal energy distribution characteristics of hybrid high-order shear horizontal guided waves and propose a sensitive normal location method for buried defects based on guided wave relative order coefficient $\boldsymbol{G}_{\mathbf{roc}}$ . As a result, we can achieve rapid and accurate location and quantification of buried defects by obtaining reflected and transmitted wave energies of different guided wave orders. The simulation and experimental results and the proposed quantification theory can be mutually verified. Results show $\boldsymbol{G}_{\mathbf{roc}}-\mathbf{based}$ defect normal shift quantification is better than commercial scanning ultrasonic body wave and eddy current detectors.
Coriolis flowmeters have been proven to be effective while measuring single phase flows, however, the measurement accuracy degrades in case of multiphase flows. In this paper a Gaussian Process Regression (GPR) based soft-computing correction model is proposed for two-phase (sand-water) slurry mass flow measurement using Coriolis flowmeters. Experimental tests were conducted on a purpose-built slurry flow test rig for two different orientations of Coriolis measuring tubes i.e. upward and downward. Five different mass flowrates, 8200, 12000, 14300, 17000 and 20000 kg/h, were tested with Solid Volume Fraction (SVF) ranging between 0 – 1.6%. A number of features, including apparent mass flowrate, density, SVF, and solid weight concentration are used as inputs to GPR models. Two GPR models are trained and tested to estimate the measurement errors of slurry mass flow measurement for the upward and downward orientations of Coriolis flowmeters, respectively. The performances of the GPR models are assessed in comparison with the reference readings. The experimental results suggest that the proposed correction models have successfully limited the relative errors within ±0.2 % for all the five mass flowrates and SVFs from 0-1.6% for both upward and downward orientations of Coriolis flowmeters.
