Search

This paper proposes a novel method that can achieve continuous control of the excitation power (equivalent to input voltage adjustment) of different reagents during acoustic droplet ejection (ADE). When a single ejector is employed, the positioning means should allow the transducer to move from one source fluid well to another quickly and in a controlled manner, thereby scanning of the source fluid wells for effective acoustic assessment and ejection therefrom. Current biotechnology screening techniques may involve many thousands of fluid transfer operations, time is wasted in the repeated power evaluation and determination processes, which is one of the critical problems in the application of acoustic droplet ejection. In this paper, we predict the required power value for producing any selected droplet initial velocity of other reagents by comparing with a standard product with known viscosity and surface tension, and realized continuous power control. Then, the power control method was verified by exciting the dimethyl sulfoxide (DMSO) droplets used in cell assays, which provided an improved acoustic liquid hanging tool for research based on cell assays and other biological analysis.

The Adaptive Fourier Analyzer has been developed for measuring the harmonic components of periodic signals with changing or unknown fundamental frequency. Vibration analysis of rotating machines, active noise control or harmonic component measurements of the lines are typical applications. Nowadays, radio or Internet-based communication is gaining popularity in various fields, e.g., in signal processing. Due to the not reliable real-time communication, some of the data are lost during the transmission. In this paper, an extension of the Adaptive Fourier Analyzer is presented which is able to handle data loss. The paper briefly presents the mathematical description of the data loss and the modification of the Resonator-Based Observer for data loss. The novelty of this paper is the modification of the Adaptive Fourier Analyzer for signals with missing samples. The performance of the algorithm is demonstrated with simulation and measurement examples.

At present, the electrical impedance tomography (EIT) technology of the lungs is mostly two-dimensional (2-D) imaging, and the three-dimensional (3-D) information of the lungs helps doctors make faster and more accurate diagnosis. However, the 3-D image reconstruction process of the lungs requires a lot of calculation and the imaging speed is slow. For this reason, this paper proposes an interpolation algorithm suitable for EIT irregular images. Firstly, the contour of the interpolated image is obtained. Secondly, the pixel values are interpolated through corresponding points. The effects of different interpolation algorithms are evaluated through experiments and simulations. The results show that the interpolation algorithm proposed in this paper is better than the traditional conventional image interpolation algorithm, and the lung image after interpolation is clearer.

The detection of electricity theft is a crucial concern for electric utilities. The combination of the electricity consumption data and machine learning models bring a new way for analyzing the customer energy consumption pattern. This paper compares three gradient boosting machines for electricity theft detection, i.e., extreme gradient boosting, light gradient boosting machine and cat boosting. By taking the advantages of gradient boosting machines in processing numerical data, this paper conducts experiments on a realistic dataset released by State Grid Corporation of China with true malicious samples. We investigate the impact of different parameters of gradient boosting machines on the electricity theft detection performance by simulations. We also compare the performances of gradient boosting machines with the wide and deep convolutional neural network. Experimental results show that gradient boosting machines outperform the wide and deep convolutional neural network for electricity theft detection using customers' electricity consumption data.

The present paper presents a low-cost solution that can be used to measure and stimulate non-nutritive suction (NNS) of premature babies. Different suction patterns, pneumatically applied to a common pacifier, can be used to promote NNS, being possible to configure the parameters of the NNS stimulation signal in a flexible way. The main parameters of the suction patterns that can be modified include: the suction signal waveform; the suction signal frequency and amplitude; the burst rate; the number of suctions per burst and the interburst pause. The proposed solution can also measure the pneumatic pressure signals applied to the pacifier, being possible to use these signals to monitor NNS activity and, if desired, to use them as a mimic reference to produce the stimulation signals. The paper also focuses some topics related with system calibration and signal processing. Several simulation and experimental results are included in the paper.

A recent and promising approach to minimize the power consumption of always-on battery-operated sensors is to perform “smart” detection of events to trigger processing. This approach effectively reduces the data bandwidth and power consumption at the system-level and increases the lifetime of sensor nodes. This paper presents an always-on, event-driven ultra-low-power camera platform for motion detection applications. The platform exploits an event-driven VGA imager that features a motion detection mode based on a tunable scene background subtraction algorithm and a grayscale imaging mode. To reduce the power consumption in the motion detection mode, the platform implements a configurable refresh rate which allows for adaption to sensing requirements by trading off between power consumption and detection sensitivity. With accurate experimental evaluation the paper demonstrates that the proposed approach reduces the system-level power consumption for always-on motion sensing applications by switching between an active 15 FPS imaging mode, consuming 5.5 mW and a low-power motion detection mode consuming 1.8 mW. We further estimate the power consumption for a single-chip solution and show that the system-level power budget can be reduced to 2.4 mW in imaging, and $400\ \mu\mathrm{W}$ in detection mode.

This paper provides an overview of recent highlights and new work from our development of robotically deployed, self-sensing alkali-activated materials: a cement technology which can be used to simultaneously repair and monitor concrete structures. The paper focuses on deployments using robotic spray coating and 3D printing, and on moisture, chloride, and damage monitoring, showing pilot data. The electrical impedance of self-sensing materials follows an exponential decrease with increasing levels of moisture, chloride contamination and temperature as expected. The work demonstrates the promise of using these novel self-sensing materials for measuring environmental measurands that are responsible for the majority of concrete structural health degradation in the field.

Long Range (LoRa) has rapidly become one of the most widespread transmission technologies for the development of distributed monitoring systems for a wide range of applications within the Internet of Things (IoT) framework. Indeed, thanks to its robustness and high performances in terms of low power consumption and long transmission range, it has been employed in a wide range of critical application scenarios. In this context, this paper analyses the usability of LoRa as solution for data transmission in icy environments. While the technical feasibility of radio transmission channel in ice has been widely demonstrated, this paper aims at demonstrating the effectiveness of a low cost transmitter developed using off-the-shelf components, when completely immersed in ice. Fields tests were performed by embedding a Low Power Wide Area Network (LoRaWAN) compliant node inside a 10 cm block of dry ice, positioned then in an environmental chamber, at a −10°C temperature. Preliminary tests demonstrate the operation of the transmitter in these operating conditions, while the achieved link margin suggests the usability of this technology even in case of notably thicker layers of ice.

The oil fume effect of the range hood is one of the important technical indicators, which provides quantitative data support for the improvement of the equipment and standardization evaluation. However, traditional performance test experiments have high environmental requirements and measurement is difficult. We propose a new comprehensive evaluation method for oil fume exhausting performance by quantifying the diffusion area of oil fume and using auxiliary measurement data of volatile compounds (VOC) in the oil fume, and integrating it into a whole-process testing system. First, the quantified result of oil fume is obtained by using the oil fume diffusion area extraction algorithm based on background difference method proposed in this paper; Then use the sensor array to obtain VOC data; Finally, comprehensive quantitative indicators are obtained through multi-source data fusion. The experimental results demonstrate that the evaluation method proposed in this paper has the advantages of low test environment requirements, high accuracy and high test repeatability.

Leakage failure risks of oil and natural gas pipelines caused by internal and external defects are different. In the magnetic flux leakage (MFL) detection, different quantification algorithms for internal and external defects are required to obtain more accurate defect quantification. However, MFL detection can not distinguish internal and external defects well. To address this problem, this paper proposed an eccentric magnetic circuit branch structure suitable for MFL detection. Based on the magnetic circuit branch structure, this paper developed a magnetic concentrating testing (MCT) method to identify internal defects. Moreover, this paper put forward a magnetic circuit isomagnetic potential point method to pinpoint the optimal MFL sensor point. Combining MFL and MCT detection, this paper presented a composite MFL detection method to discriminate between internal and external defects in the pipelines. The results of the simulation demonstrated the effectiveness of the composite MFL detection method.

Nowadays, the Internet of Things is spreading in several different research fields, such as factory automation, instrumentation and measurement, and process control, where it is referred to as Industrial Internet of Things. In these scenarios, wireless communication represents a key aspect to guarantee the required pervasive connectivity required. In particular, Wi-Fi networks are revealing ever more attractive also in time- and mission-critical applications, such as distributed measurement systems. Also, the multi-rate support feature of Wi-Fi, which is implemented by rate adaptation (RA) algorithms, demonstrated its effectiveness to improve reliability and timeliness. In this paper, we propose an enhancement of RSIN, which is a RA algorithm specifically conceived for industrial real-time applications. The new algorithm starts from the assumption that an SNR measure has been demonstrated to be effective to perform RA, and bases on Reinforcement Learning techniques. In detail, we start from the design of the algorithm and its implementation on the OmNet++ simulator. Then, the simulation model is adequately calibrated exploiting the results of a measurement campaign, to reflect the channel behavior typical of industrial environments. Finally, we present the results of an extensive performance assessment that demonstrate the effectiveness of the proposed technique.

In the dredging field, the concentration of the solid-liquid two-phase flow in the transmission pipeline is one of the most important parameters. It is measured by a radioactive source densitometer currently. However, the densitometer has limited precision, low intelligence level, and high risk of nuclear pollution. Electrical Resistance Tomography (ERT) is a noninvasive visual detection technology with the advantages of high time resolution, no radiation and low cost. So far, three ERT methods for calculating concentration have been proposed, whereas these methods have individual applicable range and complementary characteristics. Hence, in this paper, a Choquet integral-based method is proposed to fuse the calculation results of these three methods towards higher calculation accuracy. Compared with other fusion methods, the proposed method can overcome these problems of inconsistency, uncertainty, and interaction among various concentration calculation methods. Based on actual data from a working dredger, key parameters in the new method are determined and the concentration calculations of the three existing methods are integrated. Compared with the induvial concentration calculation methods, tested results show that the integrated concentration results are more accurate and consistent with the actual situations under typical dredging conditions.

As a powerful tool to get a recursive solution of least squares estimation, the Kalman filter has been used for image reconstruction in Electrical Capacitance Tomography (ECT). In the Kalman-filter-based image reconstruction method, some key parameters, e.g., initial guess, observation noise covariance and initial estimate error covariance, greatly influence the performance of the method. Inappropriate values of these parameters may cause a series of problems, such as lower convergence rate, artifacts, or filter divergence. This paper aims to analyze the influence of the parameters on the image quality for ECT and guide the selection of the parameters. Numerical simulation and experiment were carried out and the results show that with an initial guess obtained by linear back projection (LBP) method and a good match of observation noise covariance and initial estimate error covariance, the performance of the Kalman-filter-based method can be improved.

The main shaft is an essential component of fan, of which defect detection and estimation is of importance. This paper proposes a rotating magnetic field detection method for high precision identification of fan main shaft defects. This method has ability of detect the defect edges in different directions by controlling the magnetization direction flexibly, which can improve the identification effect of defect edge greatly. The edge recognition algorithm is also used here to deal with the detected magnetic signals. The presented procedure of defect detection and recognition is tested using arbitrary based on finite element method, and the result shows a good agreement between the target and recognized defect edge, which demonstrates the validity and efficiency of the proposed method.

In this paper, we developed a contactless method to estimate human subjects' heart rates using a magnification of video sequences. Existing methods based on the Eulerian video magnification (EVM) technique are severely affected by motion (i.e., object movements). The proposed method can consider both scenarios of still and moving subjects to magnify the skin color changes over time. In the proposed method, we employed a wavelet decomposition process along with a denoising layer based on principal component analysis (PCA). PCA smoothed the frames and reduced the noise due to high-frequency fluctuations caused by motion (i.e., unwanted source of color changes in the skin). Finally, the video data was magnified to extract the heart rate information. As a result, subtle changes caused by blood flow were made clearly visible to the naked eye. We compared the heart rate estimation results obtained through our proposed method with the results produced by linear EVM. The experimental results indicate an improvement in the heart rate estimation accuracy in the presence of motion using the proposed method.

This paper determines the scale factor for an electromagnetic groundwater flowmeter using an air-cored square pancake coil as the electromagnet. It also proposes the use of an array of electrodes to provide a better estimate of the average water speed between the electrodes.

As a non-intrusive, fast-response and highly sensitive and diagnostic tool, Wavelength Modulation Spectroscopy (WMS) has been extensively applied in accurate retrieval of gas properties, e.g. species concentration and temperature. Using the calibration-free WMS (CF-WMS) strategy, the first harmonic normalised second harmonic signal, e.g. $2f/1f$ of the modulated laser transmission is extracted, and then fitted to calculate the path-integrated absorbance. However, the fitting process mainly suffers from (a) noise in the fitting results introduced by the shift of the centre wavelength of the laser, and (b) a relatively high computational cost due to the least square optimisation. To improve the measurement precision and efficiency, this paper proposes a machine learning regression algorithm to calculate the gas properties. The proposed method employs artificial neural networks (ANN) to compute the path-integrated absorbance rapidly with a high signal-to-noise ratio, which was experimentally validated by calculating the absorption of water vapour at the wavelength of 1391.2 nm. In comparison with the traditional fitting method, the proposed machine learning based WMS is two times more noise-resistant with high capability to compute 100 sets of $2fl1f$ signals in approximately $0.4s$ , denoting its potential applicability in real-time and rapid trace gas sensing.

Traditional regularization algorithms for electromagnetic tomography (EMT) reconstruction merely obtains the approximate estimation value of a single conductivity, and the model information provided is limited. A large number of reasonable model parameter estimates are obtained by statistical methods. In this paper, we propose an EMT image reconstruction method based on Bayesian statistical model. According to the sparsity of defect distribution, the solved conductivity can be divided into a series of block structures. With the aid of sparse Bayesian learning (SBL) framework, statistical information, including the prior probability of the conductivity sparse distribution and the noise information in the measurement data, is taken into account. Hence the full statistical description of the conductivity distribution can be obtained. Both simulation and experimental results show that this method effectively improves the quality and accuracy of the defect image.

To improve the reliability and safety of complex electrical systems, an end-to-end fault diagnosis method for analog circuits is proposed in this paper. First, by combining the convolutional neural networks (CNN) and the gated recurrent unit (GRU) networks, a feature extraction model based on CNN-GRU is developed to obtain information that characterizes the essential states of the circuit under test (CUT) from the its signals. Compared with traditional feature extraction methods, the CNN-GRU model can obtain the spatial features of signals while retaining the time sequence features. Then, a dual-input structure of the time domain and frequency domain is designed for the CNN-GRU model, and the time-frequency domain fusion features of the signals are obtained by using the dual-input fault diagnosis model based on CNN-GRU, thereby fully reflecting the circuit states. The Sallen-Key bandpass filter circuit in ISCAS'97 circuit set is adopted to comprehensively evaluate the proposed method. Experimental results prove that the proposed fault diagnosis method can implement the accurate identification for incipient single fault classes and double fault classes.

The focus of this study is to characterize a soft elastomeric capacitor (SEC) capable of monitoring large levels of deformation towards smart bandage applications. The SEC is a highly compliant, low cost, and scalable strain gauge, designed to transduce strain into a measurable change in capacitance. Early work on the SEC was conducted on untextured versions, where the dielectric was sandwiched between two flat electrodes. It has been shown that the use of currugated surfaces significantly improves the sensor's sensitivity, linearity, and resolution, and that the sensor could be used to effectively monitor elongation in intact (i.e., unwounded) canine skin. In this paper, work on the SEC technology is extended to evaluate the use of corrugated SEC as a smart bandage capable of monitoring strain and reducing stresses on soft tissue. A series of experimental tests is conducted on two corrugated patterns, namely a reentrant hexagonal honeycomb pattern (auxetic pattern) and a symmetric diagonal grid pattern, to further characterize the signal of the corrugated sensor under various levels of localized strain resembling an opening wound. After, prior numerical investigations on canine skin are extended to study opportunities in leveraging corrugation to improve stress distribution around a wound. Results show that corrugation significantly improves both the signal and stress distribution onto soft tissue, with the auxetic pattern exhibiting best performance, therefore demonstrating the promise of the corrugated SEC as a smart bandage.

Robust detection and localisation is a crucial step in developing an autonomous rock melon harvester. Rock melons have a distinct texture, shape and colour that computer vision techniques can capitalise on. Any applied method must also consider performance that goes beyond computer vision such as hardware, speed, and suitability against the robotic arm that harvests the melon. This paper proposes a hybrid 2D with depth method, and compares its performance against traditional 2D image processing techniques. In addition, two types of texture descriptors were examined: (1) scaled Haralick Features (HF) and (2) uniform Multi-scale Local Binary Patterns (MSLBP). These methods were tested on a robotic system and compared using the following performance metrics (a) recall, (b) precision, (c) computational cost and (d) harvesting speed. Testing was performed for detection and localisation of rock melons under 4 levels of leaf occlusion: 0%, 25%, 50% and 75%. The proposed hybrid-MSLBP method is the best performing method for the robotic system used, achieving an average harvest time of 19.71s while a traditional image processing method lags behind at 102.32s.

The measurements of electrocardiographic and plethysmographic signals are two of the standard measurement techniques which are performed as noninvasive spot check methods of human health condition in hospitals, home-care or other health care facilities. Therefore, specialised electronic systems exist, which perform either of the two measurements. For the separate analysis of one signal at a time this is a sufficient approach, but if the correlation of the two signals is of particular interest it might be better to combine the measurement systems into one system as a signal delay between the different measurements is avoided. This paper aims to describe a specialised hardware frontend, which is capable of high speed, high precision, concurrent and time synchronised recording of the electrocardiographic and plethysmographic signals. It makes it possible to correlate both signals and to generate new data such as the pulse transit time. This parameter is presumed to be a marker of blood pressure or stress and is therefore of high interest of being investigated.

MAXimum Feasible Subsystem (MAX FS) algorithms have recently been shown to outperform traditional sparse recovery algorithms for compressive sensing of noiseless signals. In this paper, we extend an existing MAX FS algorithm to enable the recovery of compressed noisy signals by introducing a regularization parameter (the absolute error tolerance) and provide a recovery pipeline for the regularized MAX FS algorithm. The new technique provides much better results than the state-of-the-art Smoothed L0 (SL0) algorithm tested on noisy speech signals. For 50% compression and with additive White Gaussian Noise (WGN) at −5 dB, the new pipeline outperforms SL0 by about 6 dB on average over 50 male and 50 female speech signals from the TIMIT dataset.

Ultrasound method is widely used in liquid-solid two-phase system for flow parameter measurement. Ultrasound attenuation characteristic of the two-phase mixed system has a strong relationship with the solid phase fraction, mixture density, and particle size distribution in the mixing process. This paper presents a multi-frequency ultrasound excitation method combined the multiple single frequency and chirp signals to investigate the ultrasound attenuation under different solid phase distribution, phase fraction, particle size distribution. The solid particles are polymethyl methacrylate with diameter of 1.5, 2, 2.5, 3 mm and density of 1.18 g/cm3, and the liquid phase is tap water with density of 1g/cm3. The ultrasound frequency ranges from 1 to 5 MHz and the transducers work at the Pitch-catch mode. The experimental results show that the ultrasound attenuation coefficient increases with the solid phase fraction under both homogeneous and heterogeneous liquid-solid mixtures. The attenuation coefficient of the heterogeneous mixtures is higher than that of the homogeneous at the same solid fraction in the present study.

Uncertainty of a flowmeter normally changes after long-term service, which brings about a difference both in steady-state and dynamic measurement result by the meter. In this paper, a piezoelectrically-actuated single nozzle flapper valve generates a short-rise time flow change as an excitation for dynamic flowmeter calibration. A commercially available piezoelectric stack is adopted for its fast response. The dynamic performance of the flowmeter is obtained by comparing the time between the pressure difference across the reference restrictor and the indicated flow rate. In order to widen the excitation flow and improve the linearity with the voltage applied, design of the return restrictor is studied. Due to ferromagnetic hysteresis, nonlinear compensation to the piezoelectric stack is implemented based on modeling the hysteresis effect. Simulation and experiments show that proper flow area of the return restrictor allows wider excitation flow rate range. The compensation on the stack can effectively improve the linearity of the excitation flow rate and the input required.

We present a distance measurement technique in a high-interference environment using ultrasonic array sensors and a direct sequence spread spectrum. In this paper, we first explain our time-of-flight calculation method for enhancing the reflected waves from an object, followed by the results obtained using this method. Object detection in an environment with significant interference could be achieved using machine learning. The evaluation results show that the proposed method can measure the distance between the sensors and static target in the range of 4-12.5 m ± 5 cm (with 99% accuracy) and a target moving at a typical human walking speed in the range of 2–7 m ± 30 cm (with 77% accuracy). We also conducted 3D measurements that were able to detect an object at 11 m outdoors from the sensors.

Leak localization in pipelines is a subject frequent in the literature, with most studies focusing either on the errors in the acquisition process, e.g. what types of sensors to use, or on the weak points of the signal processing methods utilized for this matter, how to deal with these vulnerabilities and improve the accuracy of those methods etc. However, another important factor that needs to be considered is the presence of several physical phenomena that take place in a pipeline and affect the creation and propagation of acoustic waves. Such phenomena like the acoustic dispersion, the frequency dependent attenuation and the resonances are studied in this paper, based on synthesized signals, and their influence on the leak localization accuracy is discussed. Furthermore, a leak localization method combining the cross-correlation technique with temporal and spectral segmentation of the acoustic signals is proposed. This method is applied to measured (stochastic) leak signals from a pipeline in a laboratory setup. Experimental results show that the proposed method can efficiently determine the position of a leak in a pipeline, since the localization error in most cases is below 3%.

This paper presents the software design, hardware implementation, and performance characterization of a reference instrument for the metrological certification of commercial flickermenters. The proposed architecture is enriched by two novel functionalities: a spectral analysis approach for identifying the modulation frequency and type, and a programmable trigger for a precise time alignment of the considered observation interval. The estimation accuracy of the proposed instrument is characterized in laboratory tests as provided by the IEC Std 61000-4-15 in terms of instantaneous and short-term flicker.

Aerospace and automotive industries make more and more use of carbon fiber laminates. They offer high advantages in terms of mechanical performance but are difficult to work on and need precise and expensive tools. Drilling is the most common operation; drill bits must be accurately verified before use to avoid damages during the process. This paper proposes an optical inspection system for drill bits wearing estimation based on a custom-designed illumination device and a processing algorithm based on computational imaging. The elaboration of the computational image is used to measure the drill bit cutting profile. Drill bits have been worn on carbon fiber laminates, and a tool wear model has been extracted using the proposed system. The proposed method has been tested on both used and new drill bits, used in the aerospace industry.

Tunable diode laser absorption spectroscopy (TDLAS) tomography is a well-established technique for combustion diagnostics, which is capable of imaging the 2-D distribution of critical flow-field parameters over cross section of the flame. Reconstruction quality and time resolution are two key aspects that need to be compromised in dynamic monitoring. In this paper, we develop a quality-hierarchical TDLAS tomographic algorithm based on Long Short Term Memory (LSTM) network. From limited amount of line-of-sight projections measured on current tomographic field, this algorithm outputs a coarse-quality temperature image and a fine-quality temperature image with different computational costs. Simulation results validated the efficiency-effectiveness trade-off achieved by this quality-hierarchical temperature imaging network.

This paper develops a non-invasive method to roughly estimate the human muscle oxygen saturation. A reflective optical system was developed to capture red and infrared images of the medial forearm for oxygen saturation calculation and two-dimensional visualization. The measured oxygen saturation is divided into muscle oxygen saturation and microvascular oxygen saturation through a developed algorithm. The subject's rest and three types of exercise are performed and his oxygen saturation was measured accordingly. The experimental results show the numerical and 2D visualization relationship between exercise level and muscle oxygen saturation and blood oxygen saturation. The purpose of this study is to propose a non-invasive measurement of physiological information to reduce the risk of infection and increase the biocompatibility of the instrument.

Electrical capacitance tomography (ECT) is a technique that uses image reconstruction algorithms to reconstruct permittivity distributions in the region of interest (ROI) with capacitances measured at the boundary of ROI. In this paper, a closed-loop iterative reconstruction algorithm is proposed, and it aims to eliminate the error between the measured capacitances and recuperative capacitances, and in this way, the quality of the reconstructed images is improved. The proposed algorithm includes three parts, namely, Fractional-order Proportional Integral Derivative (FPID) controller, Calderon's method, and a fast-solving forward method for numerical simulated capacitances in ECT. The error to be eliminated is the difference between measured capacitances and the simulated capacitances and is input to the FPID controller. When the coefficients in the FPID controller are determined, the output of controller is used to reconstruct permittivity distribution in the Calderon's method. Reconstructed image provides simulated capacitances as an updating reference for measured capacitances for the iterative reconstructions. Both noise-free and noisy data are used and verified the feasibility and effectiveness of the proposed algorithm for ECT.

Bioderived mechano-electrical transducers are a family of composites relevant to the implementation of a circular economy. In the paper, we investigate and compare two different configurations for bacterial cellulose-based generating motion sensors. Both open circuit voltage generation and short circuit current generation are experimentally investigated and compared. Obtained results can give meaningful insight into choosing the adequate conditioning solution for this class of green transducers.

Understanding the performance of nuclear reactor plate fuels is essential to the development of new fuel systems. The condition of pre- and post-irradiation plates must be characterized effectively and efficiently before and after irradiation within a high rad environment. Characterization work after irradiation must be performed remotely, in an environment hostile to instrumentation. Laser-based characterization methods provide a capability to be robust within a hot-cell environment. Laser-based characterization can provide high spatial resolution appropriate for scanning and imaging large areas. Idaho National Laboratory is developing a laser-shock system for use in the post-irradiation examination station of the Hot Fuel Examination Facility. This laser-shock system is used to characterize the strength of bonds between fuel and cladding and between cladding and cladding. The technique induces large-amplitude shockwaves to mechanically characterize interfaces between the materials. As the shockwave propagates as a compressional wave through the fuel plate to the free (unconfined) backside surface, it is reflected back through the test plate as a rarefaction (tensile) wave. This wave is the physical mechanism that induces internal delamination failure. However, accurate measurement of surface velocity to quantify the energy traveling through the interfaces to the back surface of the fuel plate was complicated by the original fiber-based velocimeter design, which was biased by parasitic back reflections. This paper discusses the discovery of the back-reflection issue and the design changes to resolve the issue. Supporting data will be presented.

The SVOM (Space-based multi-band astronomical Variable Objects Monitor) is an international cooperation project lead by CNSA (Chinese National Space Agency) and CNES(Centre National d'Etudes Spatiales, France), which focuses on the detection of GRB (Gamma-Ray Burst). X-band system is the data transmission channel for SVOM science data. In corporation with both the Chinese and French x-band ground stations, on board x-band system sends the science data of different payloads to the scientists according to the arrangement of downloading plan. In this paper, on board x-band system design and board to ground interface is elaborated. The validation and measurement of the board to ground interface proves the effectiveness of on board system design.

Ultrasonic attenuation model is a very important theoretical basis for the measurement of droplet radius and volume fraction by ultrasonic attenuation method. The accuracy of measurement results depends on the rationality of model selection. In this paper, the measurement of droplet in the mid-wavelength region is studied. The BLBL (Bouguer-Lambert-Beer-Law) model, the superposition model of McClements model and BLBL model and the superposition model of H&T (Harker & Temple) coupled phase model and BLBL model are taken as theoretical models of ultrasonic attenuation method. Based on COMSOL simulation, the measurement effects of the three models are compared. By comparing the errors between simulation settings and inversion results obtained by these models, the superposition model of the McClements model and BLBL model is the best of the three models. The error of inverse particle radius is within 10.1%, and the error of inverse volume fraction is within 6.54%. Therefore, the superposition model of the McClements model and BLBL model is the most suitable of the three models for the measurement of large droplets in the mid-wavelength region.

RTV (Room Temperature Vulcanizing silicone rubber) is widely applied to prevent serious pollution flashover accidents on the porcelain and glass insulators. The substandard coating thickness will seriously affect the insulation performance of the insulator and bring hidden trouble to the safe and reliable operation of the equipment. This paper proposes a nondestructive detection technique for RTV coating quality based on photothermal radiation measurement. Firstly, using COMSOL simulation methods, RTV coating models of insulators of different thicknesses are created to solve temperature curves under pulse excitation. Fast Fourier Transform (FFT) transformation is performed on the descending part of the temperature curve. In the spectrum diagram, the coating thickness is characterized by two different thermal properties in the frequency domain, the coating thickness is fitted to the phase at a specific frequency and the reciprocal of the square root of the minimum phase frequency. Both methods can characterize the coating thickness and the characterization effects are validated through experiments. The research shows that using the thermal characteristics of this frequency domain can achieve quantitative detection of insulator RTV coating thickness, which will become a portable, safe, large-scale, non-destructive, on-site RTV coating thickness detection method.

This paper shows the opportunities of data preprocessing and how it influences the time required to record a sufficient amount of valid calibration data samples. Specifically, we approach the minimum needed time for calibration from two sides: on the one hand, repetitions are omitted for training one by another to define the lowest number of valid data that is needed for a model to achieve a reasonable accuracy. On the other hand we add samples, that are labeled as valid data points by steady-state detection to the dataset compared to a time-consuming manual annotation. The results will be demonstrated on a dataset of a metal oxide semiconductor gas sensor in temperature cycled operation measuring mixtures of artificial room air containing several volatile organic compounds and quantifying formaldehyde which is carcinogenic and therefore of high concern in indoor environments. The dataset is generated with an automated gas mixing system and then optimized with the help of data pre-processing methods based on steady-state detection, outlier detection and ResNet neural networks. The dataset can be reduced to only 50 % of the original data and is still able to train an artificial neural network with a root mean square error smaller than 25 % compared to the guideline value for formaldehyde concentration defined by the WHO.

Lung cancer is one of the deadliest diseases in which adenocarcinoma account for nearly 40%. To make an effective treatment and diagnosis, it is vital to accurately discriminate invasive adenocarcinoma (IA) from non-IA by analyzing ground glass nodules (GGNs) from patient's CT images. Compared with solid nodules and normal lung parenchyma, the contours of GGN are blurred and the gray scale is little changed. So far, the problem to accurately discriminate IA and non-IA remains unsolved due to insufficient labeled GGN images. In this paper, considering the generalization of convolutional neural network (CNN) and various flexible transfer strategies, we proposed a lung adenocarcinoma classification method after combining transfer learning and CNN, where the use of transfer learning strategies aims at overcoming the problem of insufficient GGN samples. Firstly, the CT image on IA and non-IA patients were collected which were labeled by surgical pathology. Secondly, two transfer learning strategies that consist of CNN feature extractor and fine-tuning network were applied to classify IA and non-IA. Finally, in the fine-tuning network process, a Progressive Fine-Tuning (PFT) strategy was combined to determine the effective depth of fine-tuning to avoid inaccurate induction of GGNs. In the CNN feature extractor experiment, four comparable models were used including linear discrimination, Support Vector Machines, K-nearest neighbor, and subspace discrimination. The indicators of sensitivity, specificity, accuracy, and AUC (area under curve) were used to quantitatively assess the performance of the two transfer strategies. Experiments show that the strategy of CNN feature extractor based on transfer learning had the highest accuracy, which was significantly higher than fine-tuning network strategy with PFT. In the experiment of CNN feature extractor, the model of linear discrimination to predict the invasiveness of GGNs has 94% accuracy whereas the other three models have 92.9%, 93.1% and 92.9%, respectively.

Transmission lines play a primary role in power distribution and communication. Manufacturing or assembly mistakes, external constraints or normal functioning can create different types of defects inside cables, which can lead to catastrophic consequences. In this paper, we propose a novel method for the diagnosis of a complex wire network, including multiple subnetworks, to detect and identify hard faults and connection inversion based on pulse excitation signals. The method can detect multiple errors with a maximum of one error per subnetwork and is suitable even for very short cables. Furthermore, we introduce a method to optimize the used excitation signals so that the longest distance can be considered, and the signal can be realized at low frequencies by low-cost embedded systems using standard circuit elements. The proposed method does not need analog-to-digital converters since it requires only two-level signals nor high-frequency components since it does not rely on the signal's propagation time like classical methods. The efficiency is proved by simulation results on a complex wire network, including seven subnetworks with 20 ports and four defects.

The accurate detection of weak current signal plays an important role in academic, industrial, daily life, and other fields. However, the weak current signals of nano-ampere levels are easily affected by sensor noise, electronic device noise, and others, which makes the precise detection of weak current signal become critical issues. Based on these challenges, to improve the detection ability of weak current signal at a specific frequency band, this paper proposes a current-voltage conversion circuit based on resonant amplifiers and a crystal filter. The resonant amplifiers convert and amplify the weak current signal, firstly. Moreover, a crystal filter is used to further compress and filter the noise and harmonics in a weak current signal. Practical test results show that this proposed weak current conversion circuit has good performance that the maximum spurious-free dynamic range improvement can be 73.9dB.

The application of TMR (Tunneling Magneto Resistance) sensor in Electromagnetic Tomography (EMT) system provides a novel idea for the improvement of system performance. In order to increase the number of independent measurement, the number of TMRs needs to be increased accordingly, and the coil location is an important parameter to be optimized. In this paper, an 8TMRs-8coils EMT sensor model and a 12TMRs-12coils EMT sensor model are built using finite element software. The image correlation coefficient is used as a quantitative evaluation index. The effect of the normalization method of measurement data, the step size of the Landweber iterative algorithm, and the coil location on the reconstructed image quality is investigated systematically. Based on the simulation results, the optimal step of the Landweber iteration algorithm and the normalization method are determined. The influence of the location of coils are qualitatively compared based on the reconstructed images and quantitatively evaluated according to the correlation coefficient between the reconstructed image and the ground truth. Based on the quantitative analysis, the optimal location of the coils are suggested.

Electrical Impedance Tomography (EIT) is a kind of non-invasive technique for reconstructing the conductivity distribution in the observation domain. The main reason that restricts the development of EIT technology is the low spatial resolution of reconstruction caused by the ill-posed inverse problem. Use information from high resolution or high precision measurement modes to guide the EIT reconstruction process is helpful to improve the reconstructed accuracy. In this paper, a B-mode ultrasound images guided electrical impedance tomography image reconstruction method is proposed. B-mode ultrasound image is used as prior information constraint in the EIT image reconstruction process via cross gradient function. A non-iterative solution of the B-mode ultrasound image guided Tikhonov regularization is derived. In view of the influence of speckle and artifact on ultrasound image, a series of image processing methods are adopted to obtain accurate image prior information. A detachable sensor which can realize EIT and B-mode ultrasound time-sharing measurement is designed. The effectiveness of the proposed algorithm is verified by phantom experiments. The experimental test results show that the proposed algorithm has the capacity of improving the reconstructed accuracy of the inclusions.

Linear frequency modulation (LFM) signals exist widely in many engineering applications, such as radar, sonar, wireless communications, and electronic reconnaissance, where stretch processing is commonly adopted to reduce the processing bandwidth. In this paper, a method of sub-Nyquist sampling and parameter estimation of LFM signals is proposed based on the stretch processing and FRI sampling theory. The method first mixes the received LFM signal with its delayed version. Then, the mixed signal is filtered by the FRI sampling kernel and sampled at a low sampling rate. Finally, the parameters of the LFM signal are estimated from the low rate samples. Theoretical analysis proves that at least 13 low rate samples are needed to estimate the amplitude, time delay, time width, chirp rate, and initial frequency parameters of the LFM signal. Theoretical analysis and simulation prove the effectiveness of the proposed sampling structure and method.

Recent advancements in energy harvesting technologies have made them potential candidates for development of self-powered biomedical systems. Integration of energy harvesters with wearable and implantable biomedical sensors could open new horizons of medical diagnosis and treatment of chronic diseases such as sleep apnea. Continuous monitoring is crucial in order to measure and analyze the breathing signal to detect sleep apnea. In this paper, an RF energy harvester powered wearable respiratory monitoring system is presented. The proposed system is implemented using low-power circuit design techniques and thus possesses the capability to be powered by RF energy harvesters. The fabricated energy harvester device realized in 130 nm CMOS process is very compact and therefore can be used in a wearable respiratory monitoring system for neonatal infants suffering from sleep apnea.

Structural health monitoring systems of wind turbines with autonomous energy harvesting can extend lifetime, avoid blade corruption and improve energy efficiency. Their sensor measurement unit and wireless transmission are supplied from an energy harvester which transforms mechanical rotation into electrical energy. The wide dynamic range of the rotational speed of the rotor and the blade pitch angle result into a variable power demand of the sensor and transceiver, which must be adapted to the power generation. The electrical energy is generated twice in a complete rotation and has a pulse shaped waveform. It is stored in a capacitor followed by a DC/DC converter. An on board measurement system observes the charge status and can predict the remaining operation time. The wireless sensor node is reconfigured to prevent an approaching energy shortage in case of a shrinking energy budget. This paper introduces an adaptive power management by means of low latency wireless reconfiguration. It enables high flexibility under all operating conditions of a variable speed wind turbine.

For preventive medicine, regular glucose monitoring is the potential production to detect the possibility of diseases at an early stage. The noninvasive and optical only glucose sensor is more user-friendly for customers because of the advantages of small size, low cost, and easy operation. The near-infrared (NIR) glucose sensor is the most popular analytic method based on the glucose properties of overtones and combinational bands. For hardware setup, the 3D printing machine was used to fabricate the assembly components including the sample cuvette structure, the sample holder, and the laser diode stage, etc. The sample cuvette was composed of the structure with two square glasses with optical path length of 1, 1.5, 2 mm, and the optical component include the 1456 nm laser diode and the PD (photodiode). This paper investigates the glucose precipitation experiment (0–90 min), the response voltages with the different optical path lengths (1, 1.5, and 2 mm), and the response voltages with the laser power was 574, 798, 950, and 1045 mW/cm2 according to Beer-Lambert law, respectively. In the results, the different response voltages had not different when the precipitation time less than 90 min. In addition, the concentration of the glucose solution with the same optical path length and the same laser power was proportional to the response voltage. The optical path length with the same concentration of the glucose solution and the same laser power was inversely proportional to the response voltage. The laser power with the same optical path length and the same concentration of the glucose solution was proportional to the response voltage. The 1 mm optical path length and the 1045 mW/cm2 laser power had the best discrimination than the other optical path lengths and the other laser powers. The experiment results indicate that the shortest optical path length and the largest laser power are required for the best discrimination of the noninvasive and optical only glucose monitoring production.

The accurate measurement of the magnetizing inrush current in transformers is of the greatest interest for the correct operation and protection of power systems. This paper presents a method for analysis and visualization of time-varying harmonics using wavelets and its application to inrush currents in transformers. The Wavelet Packet Transform and the Inverse Wavelet Packet Transform are used as a filter bank for decomposition and reconstruction in the time domain of the harmonics in the inrush current. The peak and the r.m.s. magnitude and their decays, and the time evolution of harmonics in time and frequency domains as a function of the instant of connection of the input voltage to the primary winding are studied in detail. The results obtained can be used for the adequate design of the transformer protection and the selection of the instant of connection of voltage supply to the primary winding.

Foreign object damage (FOD) is a frequent but dangerous failure of the compressor. Blade tip timing (BTT) is regarded as a promising on-line monitoring method for blades, but it is difficult to obtain the real FOD data in a laboratory environment. This paper proposes a virtual BTT measuremnt method for FOD. Firstly, we establish the dynamic response model of the multi-degree of freedom damped system, and use the modal superposition method to infer its response displacement under the impact of a foreign object. Secondly, we use the Johnson-Cook model to calculate the stress limit of the blade model. Thirdly, finite element analysis is employed to calculate the influence of different impact positions on the blade tip displacement and the blade failure. Finally, we use the Simulink module to simulate the blade vibration displacement signal, and the measurement process according to the principle of tip timing. By this means, the laboratory simulation of BTT measurement under FOD is realized.

Diabetes Mellitus (DM) is a chronic disease characterized by the reduced metabolic action of glucose in the human body that triggers the failure of proper functions of various organs. Diagnosis of diabetes is challenging due to the complex correlation with other diseases and initial asymptomatic behavior. Efficient hardware design that incorporates the evaluation of risk factors of the disease can improve early diagnosis, regular monitoring, and clinical decision-making. In this paper, we present an implementation of a deep learning (DL) inference in Field Programmable Gate Array (FPGA) to predict DM. The model hyperparameters have been tuned to obtain the DL model that ensures acceptable performance in low-power and miniaturized silicon area with reduced storage requirements. We trained a four-layer Fully Connected Neural Network (FCNN) with RMSProp optimizer and binary cross-entropy loss function. The best-learned model among 200 epochs is used to extract the weights and biases for hardware implementation of the inference module. We achieved an accuracy of 91.15%. The outcome of this research can be integrated with the system-on-chip platform to develop smart diabetic monitoring and management tools.