Search

This brief focuses on the asynchronous bumpless transfer finite-time $H_{\infty }$ control problem for a class of Markovian jump systems (MJSs). The hidden Markov principle is introduced to represent the asynchronous jumping phenomenon between the MJSs and the designed bumpless transfer finite-time $H_{\infty }$ controller. The asynchronous jumping transient behavior of MJSs is described by using a bumpless transfer constraint index. First, an asynchronous bumpless transfer constraint is proposed to limit the control bumps caused by asynchronous jumping. Then, an asynchronous bumpless transfer finite-time $H_{\infty }$ controller is constructed, and sufficient conditions are given to ensure the finite-time $H_{\infty }$ performance for MJSs under the proposed asynchronous bumpless transfer method. Finally, an RLC circuit model is provided as an application example to show the effectiveness of the proposed control method. [ABSTRACT FROM AUTHOR]

In this communication, two compact co-polarized decoupled microstrip patch array antennas (MPAAs) based on TM02/TM03 modes cancellation are proposed. It is only necessary to introduce shorting pins and etch slots on the patch antenna to precisely adjust the frequency and impedance of TM02 and TM03 modes to make them consistent, which can significantly reduce the strong coupling between the tightly spaced patch antennas. In addition, the L-shaped open-stub coupling feeding is used to eliminate the parasitic inductance caused by the shorting pins to achieve better impedance matching. Finally, the proposed two-/eight-port MPAAs are fabricated and measured. The simulation and measurement results show that the coupling of adjacent ports is decreased lower than −30 dB within the impedance matching bandwidth (3.42–3.58 GHz), while the edge-to-edge spacing is only $0.023 \lambda _{0}$ between the antenna elements. With the advantages of simple design, significant decoupling effect, and good radiation performance, the proposed decoupling scheme has broad application prospects in the fifth-generation (5G) array antennas. [ABSTRACT FROM AUTHOR]

Although ship detection methods based on deep learning have achieved remarkable progress, the design of the universal ship detection (USD) system is rarely studied. The main challenge of USD lies in the notorious domain bias and shift problem across multiple domains. This article implements USD based on domain-invariant representations to alleviate this issue. Specifically, the proposed method integrates a multilevel domain classification network (MDCN) and a domain-centric cut-paste module (DCM). First, the backbone network is facilitated to learn domain-independent image features from multiple domains through MDCN, thereby reducing the disturbance of domain-specific features to universal detector. Furthermore, the proposed method combines the domain-related synthetic samples generated by DCM to provide MDCN with strong supervision information, which further motivates the network to be more attentive to the domain-invariant representations at the instance level. Finally, we conduct experiments on multiple ship datasets in the synthetic aperture radar (SAR) and optical domains to verify the effectiveness of the method. The results show that the proposed method outperforms baseline by around 2.95% average precision (AP50), which achieves an effective USD system by complementing the information between domain-invariant representations. [ABSTRACT FROM AUTHOR]

In this article, a compact differentially fed ±45° dual-polarized filtering magneto-electric (ME) dipole antenna with improved harmonic suppression is proposed and analyzed. First, the fusion of loop-slot dipole and half-wave vibrator is calculated and mentioned to furnish theoretical guidance. Then, by reasonably placing two pairs of second-order step impedance feedlines and a loop-slot patch, a corresponding two-mode loop-vibrator combination filtering antenna with an upper radiation null is naturally constructed. Moreover, simply by loading four open-circuit stubs at the center of octagonal slot patch edge, another extra resonance mode and radiation null in the upper band can be simultaneously obtained. Later, through hiring the reconstructed third-order stub-loaded-resonator (SLR) feedlines, wideband harmonic suppression along with a third radiation null is unaffectedly achieved. Finally, the optimized antenna is fabricated and tested. The measured results reveal it realizes a wideband impedance bandwidth of 42.0% (2.78–4.26 GHz) and a high port isolation of 35 dB. Also, the length of its harmonic suppression is measured from 4.56 to 7 GHz, while the depth reaches more than 20 dB. In addition, the measured gain and radiation efficiency curves in the operating band are relatively stable and both are at a high average value (8.2 dBi, 85%). [ABSTRACT FROM AUTHOR]

Convolutional neural networks (CNNs) have been widely used for synthetic aperture radar (SAR) target detection. Typical methods based on CNN have obtained favorable detection accuracy at the cost of high model complexity, and thus are difficult to be directly applied to real-time satellites on board as well as maritime rescue. To deal with this problem, this article proposes an efficient and lightweight target detection network incorporating soft quantization. First, to compensate for the lack of accuracy caused by lightweight networks, a feature fusion module called split bidirectional feature pyramid network is proposed to alleviate the interference of complex background on SAR images. Meanwhile, to adapt the lightweight network and the feature fusion module, a linear transformation module is presented to enhance the linear representation of the model via learnable parameters. Eventually, to make the model size smaller, a soft quantization algorithm is proposed to reduce the accuracy degradation caused by quantization errors. We validate the robustness of the model in several publicly available datasets. Experimental results show that our model achieves 97.0% detection accuracy on SAR ship detection dataset, with a 0.9% accuracy improvement compared to mainstream methods using less than $15\times $ the number of parameters and less than $6\times $ the number of flops. [ABSTRACT FROM AUTHOR]

For switched linear parameter-varying (LPV) systems with possible actuator failures, the parameter-dependent multiple piecewise Lyapunov function is constructed to handle the $H_\infty$ bumpless transfer fault-tolerant control problem. First, a generalized bumpless transfer concept is presented for switched LPV systems to describe the transient performance, for which only the local bumpless transfer condition is required. Second, an event-triggered switching law, depending on the system states, external parameters, and dwell time, is designed to ensure a time span among adjacent switchings. Third, a solvability condition of the $H_\infty$ bumpless transfer fault-tolerant control problem is developed. A family of time-driven switching controllers with bumpless transfer constraint and fault-tolerant requirement are also designed. Finally, an application example of an aero-engine is given to verify the effectiveness of the developed methods. [ABSTRACT FROM AUTHOR]

This article focuses on the bumpless transfer $H_{\infty }$ anti-disturbance control problem for switching Markovian LPV systems under a hybrid switching law. A parameter-dependent multiple piecewise disturbance observer-based bumpless transfer control strategy is put forward to reject multiple disturbances and reduce switching bumps. First, a hybrid switching law making full use of determinacy and randomness is proposed to improve the bumpless transfer anti-disturbance level by introducing a fixed dwell time in random switching. Second, a generalized bumpless transfer anti-disturbance specification is given to describe the switching quality at the switching points of switching Markovian LPV systems. Third, a solvability condition is established for the bumpless transfer $H_{\infty }$ anti-disturbance control problem, and a parameter-dependent multiple piecewise disturbance observer-based bumpless transfer controller is designed. Finally, an application example has been supplied to demonstrate the availability of the developed method. [ABSTRACT FROM AUTHOR]

Time-sensitive networking (TSN) can help standardize deterministic Ethernet across industrial automation. The deterministic guarantee of TSN is based on network resource scheduling in the unit of flow. However, the state-of-the-art TSN single flow scheduling scheme cannot meet the coordinated scheduling requirements of multiple data flows in advanced industrial applications (e.g., control and safety applications). In this article, we propose a TSN chain flow abstraction, TC-Flow, for a coordinated multiple-flow scheduling model in industrial control and safety applications. Based on the proposed TC-Flow model, we design an offline TC-Flow scheduling algorithm using integer linear programming and an online heuristic TC-Flow scheduling algorithm to handle network dynamics. To deploy the proposed TC-Flow model and scheduling algorithms in the TSN, we design a CF-TSN network architecture that is compatible with the existing TSN single-flow scheduling scheme. Finally, we implement the proposed CF-TSN architecture and TC-Flow scheduling algorithms in real-world network environments. Experimental results show that the proposed scheduling algorithms can increase the number of schedulable flows by 26 percent compared to the state-of-the-art TSN scheduling benchmark. [ABSTRACT FROM AUTHOR]

This article addresses the problem of adaptive finite-time neural tracking control for nonstrict-feedback nonlinear systems via dynamic surface (DS) technique. First, a new quasi-fast finite-time practical stability (QFPS) criterion is proposed for a class of general nonlinear systems. Then, the presented QFPS criterion is applied to design the desired adaptive finite-time neural tracking controller for a class of nonstrict-feedback nonlinear systems. The presented design scheme for the nonstrict-feedback nonlinear system has the following two features: 1) the “explosion of complexity” issue of the backstepping design is addressed by utilizing the DS technique and 2) the structural feature of Gaussian functions is applied to solve the design difficulties caused by the nonstrict-feedback form. It is proved that the designed controller for the nonstrict-feedback nonlinear system can make the resulting closed-loop system stabilizable in a quasi-fast finite time and the tracking error converges to a sufficiently small neighborhood of the origin. Finally, the simulation results are given to show the validity and practicability of the proposed design scheme. [ABSTRACT FROM AUTHOR]

A low-profile dual-circularly polarized (CP) substrate integrated waveguide (SIW)-cavity-backed slot patch antenna with wide axial ratio beamwidths (ARBWs) is proposed. First, two orthogonal slots are etched on a square patch, and a SIW cavity is backed under the patch. Evolution of the SIW-cavity-backed cross-slot antenna is analyzed, and its beamwidths are enhanced by using the E-plane omnidirectional radiation of the slots. In order to broaden the ARBWs, four different SIW-cavity-backed slot patch antennas are also compared. Then, two rat-race couplers and two Wilkinson power dividers are skillfully cascaded together, and a novel six-port feeding network for the dual-CP antenna with high isolations is designed. Finally, a compact dual-CP slot patch antenna with dimensions of 0.53 λ0×0.53 λ0×0.035 λ0 is fabricated. Its measured results reveal that the antenna achieves 180°/194° left-handed CP ARBWs and 203°/200° right-handed CP ARBWs in two orthogonal principal planes at the center frequency of 5.2 GHz. Its measured isolation between two ports is greater than 16 dB in the entire bandwidth. [ABSTRACT FROM AUTHOR]

In this paper, we aim to make the best joint decision of device selection and computing and spectrum resource allocation for optimizing federated learning (FL) performance in distributed industrial Internet of Things (IIoT) networks. To implement efficient FL over geographically dispersed data, we introduce a three-layer collaborative FL architecture to support deep neural network (DNN) training. Specifically, using the data dispersed in IIoT devices, the industrial gateways locally train the DNN model and the local models can be aggregated by their associated edge servers every FL epoch or by a cloud server every a few FL epochs for obtaining the global model. To optimally select participating devices and allocate computing and spectrum resources for training and transmitting the model parameters, we formulate a stochastic optimization problem with the objective of minimizing FL evaluating loss while satisfying delay and long-term energy consumption requirements. Since the objective function of the FL evaluating loss is implicit and the energy consumption is temporally correlated, it is difficult to solve the problem via traditional optimization methods. Thus, we propose a “Reinforcement on Federated” (RoF) scheme, based on deep multi-agent reinforcement learning, to solve the problem. Specifically, the RoF scheme is executed decentralizedly at edge servers, which can cooperatively make the optimal device selection and resource allocation decisions. Moreover, a device refinement subroutine is embedded into the RoF scheme to accelerate convergence while effectively saving the on-device energy. Simulation results demonstrate that the RoF scheme can facilitate efficient FL and achieve better performance compared with state-of-the-art benchmarks. [ABSTRACT FROM AUTHOR]

A wide axial-ratio beamwidth (ARBW) and half-power beamwidth (HPBW) circularly polarized (CP) antenna is realized by arranging two curved patch dipole pairs orthogonally in a cruciform contour that conforms with a hemispheric shell. First, two orthogonal curved dipole pairs are designed via step-by-step investigation of the evolution of traditional crossed dipoles. Then, an antenna with a wide CP working bandwidth and wide ARBWs and HPBWs is developed by enlarging the diameter of the curved arms, replacing the narrow arms with the wide patches and making these patches conformal with the shell surface; this improves the impedance bandwidth (IBW) and further reduces the antenna’s size. Finally, a prototype with reduced dimensions of $0.52\times 0.52\times0.15$ ($\lambda _{0}^{3}$) is designed, fabricated, and measured. The resulting antenna exhibits a −10 dB IBW of 55.8% (1.06–1.89 GHz) and a 3 dB AR bandwidth of 70.9% (1–2.1 GHz), within which the ARBW and HPBW reach maximum values of 202°/196° and 113°/111° in the two orthogonal principal planes at 1.602 and 1.268 GHz, respectively. Furthermore, a 3 dB ARBW exceeding 100°, an HPBW exceeding 96°, and efficiency exceeding 60% are acquired within the 1.252–1.89 GHz (40.6%) target service coverage band. [ABSTRACT FROM AUTHOR]

Active learning is a unique abstraction of machine learning techniques where the model/algorithm could guide users for annotation of a set of data points that would be beneficial to the model, unlike passive machine learning. The primary advantage being that active learning frameworks select data points that can accelerate the learning process of a model and can reduce the amount of data needed to achieve full accuracy as compared to a model trained on a randomly acquired data set. Multiple frameworks for active learning combined with deep learning have been proposed, and the majority of them are dedicated to classification tasks. Herein, we explore active learning for the task of segmentation of medical imaging data sets. We investigate our proposed framework using two datasets: 1.) MRI scans of the hippocampus, 2.) CT scans of pancreas and tumors. This work presents a query-by-committee approach for active learning where a joint optimizer is used for the committee. At the same time, we propose three new strategies for active learning: 1.) increasing frequency of uncertain data to bias the training data set; 2.) Using mutual information among the input images as a regularizer for acquisition to ensure diversity in the training dataset; 3.) adaptation of Dice log-likelihood for Stein variational gradient descent (SVGD). The results indicate an improvement in terms of data reduction by achieving full accuracy while only using 22.69% and 48.85% of the available data for each dataset, respectively. [ABSTRACT FROM AUTHOR]

Performing deep neural network (DNN) inference in real time requires excessive network resources, which poses a big challenge to the resource-limited industrial Internet of things (IIoT) networks. To address the challenge, in this paper, we introduce an end-edge-cloud orchestration architecture, in which the inference task assignment and DNN model placement are flexibly coordinated. Specifically, the DNN models, trained and pre-stored in the cloud, are properly placed at the end and edge to perform DNN inference. To achieve efficient DNN inference, a multi-dimensional resource management problem is formulated to maximize the average inference accuracy while satisfying the strict delay requirements of inference tasks. Due to the mix-integer decision variables, it is difficult to solve the formulated problem directly. Thus, we transform the formulated problem into a Markov decision process which can be solved efficiently. Furthermore, a deep reinforcement learning based resource management scheme is proposed to make real-time optimal resource allocation decisions. Simulation results are provided to demonstrate that the proposed scheme can efficiently allocate the available spectrum, caching, and computing resources, and improve average inference accuracy by 31.4 % compared with the deep deterministic policy gradient benchmark. [ABSTRACT FROM AUTHOR]

With the rapid development of artificial intelligence and industrial Internet of Things (IIoT) technologies, intelligent predictive maintenance (IPdM) has received considerable attention from researchers and practitioners. To efficiently predict impending failures and mitigate unexpected downtime, while satisfying the instant maintenance demands of industrial facilities is very important for improving the production efficiency. In this article, a self-attention based “Perception and Prediction” framework, called DeepHealth, is proposed for the instant IPdM. Specifically, the framework is composed of two submodels (i.e., DH-1 and DH-2), which are respectively utilized to perform the health perception and sequence prediction. By operating the framework, the proposed models can predict the health conditions via predicting the future signal samples, thereby completing the instant IPdM. Considering the potential temporal correlation in time series, we deploy an enhanced attention mechanism to capture global dependencies from the vibration signals, and leverage the long- and short-term sequence prediction of sensor signals to support instant maintenance decision-making. On this basis, we conduct a destructive experiment based on the IIoT-enabled rotating machinery and construct a balanced industrial dataset for model evaluations. Extensive experiment results show that the proposed solution achieves good prediction accuracy for instant IPdM on the automatic washing equipment and Case Western Reserve University datasets. [ABSTRACT FROM AUTHOR]

In this article, an adaptive approximation-based output-feedback tracking control scheme is presented for a class of stochastic switched lower-triangular nonlinear systems with input saturation and unmeasurable state variables. First, to overcome the design obstacle caused by the nondifferential saturation nonlinearity, a carefully selected nonlinear function of the control input signal is applied to estimate the saturation function. Then, a reduced-order state observer is designed to model the unmeasured system states, which also means the error system can be established. Furthermore, the fuzzy-logic systems are utilized to approximate the unknown system nonlinearities in the adaptive backstepping-based controller design procedure. It is ensured that all the closed-loop system variables are bounded in probability and the error signal belongs to a compact set in the mean square sense. Finally, the effectiveness and the practicability of the proposed control scheme are shown by two examples. [ABSTRACT FROM AUTHOR]

This paper explores how to promote business model innovation through intraindustry search and extraindustry search with the assistance of fit dynamic capabilities (internal coordination capability and stakeholder engagement capability). Based on evolutional learning and business model literatures, six hypotheses are proposed and examined with data from 204 firms in China. The results show that intraindustry search has an inverted U-shaped effect on business model innovation, whereas extraindustry search has a positive effect. Internal coordination capability strengthens the effect of intraindustry search but weakens that of extraindustry search. Stakeholder engagement capability strengthens the effect of extraindustry search but weakens that of intraindustry search. Our findings enrich the business model literature by identifying different roles of intraindustry search and extraindustry search as the antecedents of business model innovation and their fitness with different types of dynamic capabilities. [ABSTRACT FROM AUTHOR]

Research about online failure prediction of rail vehicle core components (such as wheels, bearings, and bogies) based on big data and artificial intelligence (AI) has become popular in view of its role of improving rail vehicle operation safety. The recent vibration energy harvesting sensor network relieves sensor nodes' dependence on wired power, which provides a green and low-cost way of collecting data from rail vehicle core components. However, the integration of an energy harvesting sensor network and AI to provide online failure prediction for rail vehicle components still faces several challenges, such as weak energy harvesting power and unstable vehicle-ground communication data rate. In this article, EH-Edge, an energy harvesting-driven cloud-edge-end device collaborative Internet of Things (IoT) platform, is proposed to efficiently integrate energy harvesting and AI to solve these challenges. A two-level collaborative AI failure prediction is proposed and deployed in the EH-Edge platform to reduce energy consumption in terms of sensor node, amount of data upload, and time delay of failure prediction. Detailed software and hardware designs and real-world data sets are also published. [ABSTRACT FROM AUTHOR]

In this letter, a broadband 1 bit reconfigurable transmitarray antenna (RTA) based on polarization rotation element is proposed. The transmitarray element is composed of two split circular rings connected by two narrow strips and sandwiched by two orthogonal polarizers. As phase shift and polarization conversion structure, the two patches are separated by the air layer, and the p-i-n diodes are soldered in the center of the narrow strips. The polarizer grids are used as dc bias lines to conduct p-i-n diodes. By controlling the p-i-n diodes in the on and off states, the polarization orientation of the wave emitted by the RTA element is changed. Then, the outgoing waves in the two states physically form a broadband 180° phase difference, which broadens the operating bandwidth of the RTA effectively. By utilizing the novel unit cell, a 16 × 16 elements RTA is designed, fabricated, and measured. The measured results show that the 3 dB and 1 dB gain bandwidths are 45% and 17%, respectively. The peak gain is 16.8 dBi with aperture efficiency of 18.4%. Furthermore, the proposed RTA has good beam scanning performance with 2-D scanning range of ±40°. [ABSTRACT FROM AUTHOR]

Malfunction of critical components in the power system can lead to large scale cascading failures and blackouts. Identifying critical components and taking protection on them can prevent the above problem effectively. In this article, investigating from the sight of network dynamics, a critical node identification method of complex power systems is proposed to find the nodes that have a great influence on the controllability and observability of the power system. First, based on electrical betweenness, a directed network model of power systems is proposed, which solves the problem that assessment results are not accurate because the traditional network model only represents the structural characteristics of power systems at a certain time. Then, combining structure controllability theory and the electrical characters of power systems, the electric cactus structure is established to provide a theoretical basis for assessment indexes. The node number of electric cactus is defined as the electric control capability index (ECCI), which measures the node's influence on the controllability of power systems. Besides, based on the duality principle and ECCI, the electric observe capability index is defined to measure the node's influence on the observability of power systems. Combining the two capability indexes, the electric dynamic characteristic index is defined. Finally, the method has been demonstrated on the IEEE 118-bus system. The results show that the identified nodes have a strong impact on the controllability and observability of the IEEE 118-bus system. [ABSTRACT FROM AUTHOR]

A compact dual-polarized filtering base-station antenna is presented in this letter. The entire size (include the ground plane) of the element is only 0.53λ × 0.72λ × 0.14λ (λ is the wavelength of the central working frequency). Four vertical slots etched on the coupling cross slot are for improving the high-frequency suppression of the antenna. The C-slots on the Gnd1 reduce the cross-polarization ratio at the axial direction. The antenna shows a wide working bandwidth of 25.6% and a low cross-polarization ratio about 22 dB. In order to validate the design, a 1 × 3 linear array is fabricated and tested. The antenna array works at 3.3–4.0 GHz with the average gain about 11 dBi. Besides, the array remains a good out of band rejection more than −20 dB up to 9 GHz which exhibits a good harmonic suppression performance over a wideband. [ABSTRACT FROM AUTHOR]

Hyperspectral images (HSIs), featured by high spectral resolution over a wide range of electromagnetic spectra, have been widely used to characterize materials with subtle differences in the spectral domain. However, a large number of bands and an insufficient number of sample pixels for each class are challenging for traditional machine learning-based classifiers. As alternative tools for feature extraction, neural networks have received extensive attention. This letter proposes to combine t-distributed stochastic neighbor embedding (t-SNE) with a convolutional neural network (CNN) for HSI classification. Our framework is designed to automatically capture the potential assembly features, which are extracted from both the dimension-reduced CNN (DR-CNN) and the multiscale-CNN. Experimental results show that the proposed classification framework outperforms several state-of-the-art techniques for three real data sets. [ABSTRACT FROM AUTHOR]

Recent advances in deep learning for medical image segmentation demonstrate expert-level accuracy. However, application of these models in clinically realistic environments can result in poor generalization and decreased accuracy, mainly due to the domain shift across different hospitals, scanner vendors, imaging protocols, and patient populations etc. Common transfer learning and domain adaptation techniques are proposed to address this bottleneck. However, these solutions require data (and annotations) from the target domain to retrain the model, and is therefore restrictive in practice for widespread model deployment. Ideally, we wish to have a trained (locked) model that can work uniformly well across unseen domains without further training. In this paper, we propose a deep stacked transformation approach for domain generalization. Specifically, a series of ${n}$ stacked transformations are applied to each image during network training. The underlying assumption is that the “expected” domain shift for a specific medical imaging modality could be simulated by applying extensive data augmentation on a single source domain, and consequently, a deep model trained on the augmented “big” data (BigAug) could generalize well on unseen domains. We exploit four surprisingly effective, but previously understudied, image-based characteristics for data augmentation to overcome the domain generalization problem. We train and evaluate the BigAug model (with ${n}={9}$ transformations) on three different 3D segmentation tasks (prostate gland, left atrial, left ventricle) covering two medical imaging modalities (MRI and ultrasound) involving eight publicly available challenge datasets. The results show that when training on relatively small dataset (n = 10~32 volumes, depending on the size of the available datasets) from a single source domain: (i) BigAug models degrade an average of 11%(Dice score change) from source to unseen domain, substantially better than conventional augmentation (degrading 39%) and CycleGAN-based domain adaptation method (degrading 25%), (ii) BigAug is better than “shallower” stacked transforms (i.e. those with fewer transforms) on unseen domains and demonstrates modest improvement to conventional augmentation on the source domain, (iii) after training with BigAug on one source domain, performance on an unseen domain is similar to training a model from scratch on that domain when using the same number of training samples. When training on large datasets (n = 465 volumes) with BigAug, (iv) application to unseen domains reaches the performance of state-of-the-art fully supervised models that are trained and tested on their source domains. These findings establish a strong benchmark for the study of domain generalization in medical imaging, and can be generalized to the design of highly robust deep segmentation models for clinical deployment. [ABSTRACT FROM AUTHOR]

A compact single-layer filtering microstrip antenna with diverse characteristics of low profile, high gain, wide band, and high selectivity is proposed in this letter. Simple structure as it is, the presented antenna mainly consists of a rectangle driven patch, four parasitic strips, a pair of symbiotic strips, and a set of shorting pins. By attaching four parasitic strips to the original rectangular patch antenna, an additional resonance can be excited while the high-band edge selectivity is greatly improved due to the formation of a radiation null outside. Then, two symbiotic strips are embedded on both sides of the driven patch, which in turn creates an extra resonance mode and a lower-band radiation null. The set of shorting pins are applied to further improve operating and filtering response. Finally, the use of two parasitic patches can effectively improve the level of impedance bandwidth (IBW) and upper-band suppression. To verify this design, the corresponding antenna is fabricated and tested. The measured and simulated results are highly consistent, which reveals that it owns a wide IBW of 20.1% (2.19–2.68 GHz) with three resonance points, a higher average gain larger than 9.5 dBi, and a flat radiation efficiency above 88% in-band. The suppression level of the out-of-band gain on both sides can reach 14.5 dB. [ABSTRACT FROM AUTHOR]

In this paper, a parameter-dependent multiple discontinuous Lyapunov function (PMDLF) approach is proposed to study the $H_\infty$ refined antidisturbance control problem of switched linear parameter-varying systems. The $H_\infty$ refined antidisturbance means the disturbance appearing in the control channel can be accurately compensated by means of the estimation of the disturbance and the energy bounded external disturbance can be restrained. A key point is to set up a PMDLF framework that provides an effective tool for attenuating the energy bounded disturbances and rejecting the disturbances generated by the exosystem accurately. A parameter-driven and dwell time-dependent switching law is designed, and a solvability condition ensuring the $H_\infty$ refined antidisturbance performance is developed. Then, the $H_\infty$ refined antidisturbance switched parameter-dependent disturbance observers and the disturbance observer-based refined controllers are established to achieve required disturbance attenuation and rejection. Finally, an example of an aero-engine control system is given to verify the availability of the acquired approaches. [ABSTRACT FROM AUTHOR]

This paper investigates dissipativity and related properties for switched linear parameter-varying (LPV) systems by using parameter and dwell time-dependent multiple storage functions, which are different from traditional multiple storage functions because of the introduction of the external parameters into the storage functions construction. First, a novel concept relating to external parameters is proposed to describe the overall dissipativity property of switched parameter-varying systems in the absence of the conventional dissipativity property of the subsystems. Second, a state and parameter-triggered and time-driven switching laws is proposed, which avoids Zeno behavior by guaranteeing a dwell time among the adjacent switching. Third, several classic forms of dissipativity are addressed. For passivity and $ {L_{2}}$ -gain, under certain negative output feedback asymptotic stability is guaranteed with the aid of asymptotic zero state detectability. Finally, the effectiveness of the proposed result is illustrated by its application to a speed adjustment problem of an aero-engine. [ABSTRACT FROM AUTHOR]

In this article, a nonvacuum-cooling compact (NVCC) scientific charge-coupled device (CCD) camera is presented that includes low noise clocks, a bias driver circuit, a data acquisition circuit, and a temperature control circuit. Experiments are conducted to test the readout performance of the proposed imaging system. The scheme for generating the CCD clocks and the bias driver circuit through ultralow noise low-dropout regulators is designed. The entire design can operate from −40 °C to 40 °C. The test results of the camera show that the system can run at a maximum readout rate of 5 Mpixels/s. The readout noise of the camera is as low as $9.29\,\, {e}^{-}$ when the readout speed is 500 Kpixels/s. The designed camera is currently used in a 1.2-m telescope system at Delingha Observatory. [ABSTRACT FROM AUTHOR]

With the tremendous revival of artificial intelligence, predictive maintenance (PdM) based on data-driven methods has become the most effective solution to address smart manufacturing and industrial big data, especially for performing health perception (e.g., fault diagnosis and remaining life assessment). Moreover, because the existing PdM research is still in primary experimental stage, most works are conducted utilizing several open-datasets, and the combination with specific applications such as rotating machinery is especially rare. Hence, in this paper, we focus on data-driven methods for PdM, present a comprehensive survey on its applications, and attempt to provide graduate students, companies, and institutions with the preliminary understanding of the existing works recently published. Specifically, we first briefly introduce the PdM approach, illustrate our PdM scheme for automatic washing equipment, and demonstrate the challenges encountered when we conduct a PdM research. Second, we classify the specific industrial applications based on six algorithms of machine learning and deep learning (DL), and compare five performance metrics for each classification. Furthermore, the accuracy (a metric to evaluate the algorithm performance) of these PdM applications is analyzed in detail. There are some important conclusions: 1) the data used in the summarized literature are mostly from public datasets, such as case western reserve university (CWRU)/intelligent maintenance systems (IMS); and 2) in recent years, researchers seem to focus more on DL algorithms for PdM research. Finally, we summarize the common features regarding our surveyed PdM applications and discuss several potential directions. [ABSTRACT FROM AUTHOR]

As a key component of rotating machinery, the health perception of bearings is essential to ensure the safe and reliable operation of industrial equipment. In recent years, research on equipment health perception based on data-driven methods has received extensive attention. Overall, most studies focus on several public datasets to verify the effectiveness of their algorithms. However, the scale of these datasets cannot completely satisfy the representation learning of deep models. Therefore, this paper proposes a novel method, called CarNet, to obtain a more robust model and ensure that the model is sufficiently trained on a limited dataset. Specifically, it is composed of a data augmentation method named equitable sliding stride segmentation (ESSS) and a hybrid-stacked deep model (HSDM). The ESSS not only amplifies the scale of the original dataset but also enables newly generated data with both spatial and temporal correlations. The HSDM can, therefore, extract shallow spatial features and deep temporal information from the strongly correlated 2-dimensional (2-D) sensor array using a CNN and a bi-GRU, respectively. Moreover, the integrated attention mechanism contributes to focusing limited resources on informative areas. The effectiveness of CarNet is evaluated on the CWRU dataset, and an optimal diagnostic accuracy of 99.92% is achieved. [ABSTRACT FROM AUTHOR]

The power supply discrepancy between advanced CMOS technology and legacy optical module leads to poor power efficiency or extra DC–DC component. A new stacked multi-channel receiver architecture and its synergetic power supply scheme for power-efficient optical links are proposed to solve the problem. Consequently, power reduction can be achieved thanks to the full utilization of available power supply headroom and the reuse of supply current. As a proof of concept, a stacked $2 \times 10$ Gb/s optical receiver with a synergetic-integrated linear multi-output regulator is implemented in 0.18- $\mu \text{m}$ CMOS technology, demonstrating the feasibility of the proposed architecture. State-of-the-art receiver performance and power efficiency are achieved. [ABSTRACT FROM AUTHOR]

An approach for the energy measurement of high-repetition rate narrow laser pulses based on a high-performance field-programmable gate array (FPGA) chip is presented in this paper. An incident narrow laser pulse is converted to an electrical signal of an appropriate width and of an amplitude linear to the energy of the laser pulse. The electrical signal is then digitalized by a high-speed analog-to-digital converter and fed to an FPGA chip. The amplitude of the electrical signal is acquired by real-time calculations in the FPGA. The test results show that our solution is well designed for the laser pulse with a full-width at half-maximum (FWHM) of 200 ps and a frequency of 20 MHz. [ABSTRACT FROM AUTHOR]

Industrial wireless sensor networks (IWSNs) have mainly been used to monitor applications, but recently an interest in control and safety applications has emerged. Functional safety and communication in open transmission systems have been laid down in the IEC 61784-3-3 standard. The standard is based on a cyclic polling mechanism, which consumes a considerable amount of bandwidth; since existing IWSNs are very resource-constrained, this becomes a major challenge. To overcome this problem, this paper proposes a novel framework that uses an event-triggered failsafe mechanism based on synchronous wired polling and wireless time-slotted time division multiple access. We analytically derive the minimum and maximum bound for the most important metric for safety-critical applications, safety function response time (SFRT). A new metric, normal state interrupt time (NSIT), is proposed in this paper. Furthermore, we also implement the proposed framework by using the WirelessHART standard. The results are compared to the classical time-triggered approach used in the IEC 61784-3-3 standard. The obtained results show that the proposed framework can reduce the bandwidth usage by 90% and support safety-critical applications that require a SFRT less or equal to 150 ms. [ABSTRACT FROM AUTHOR]

To investigate the difference between the sulfur hexafluoride (SF6) decomposition characteristics initiated by two sorts of insulation defects — free metal particles and metal protrusion defects — under negative DC partial discharge (PD), SF6 decomposition data was obtained under different DC voltages on the basis of constructing a SF6 decomposition experimental platform. The differences between the characteristic components and the characteristic ratios of the two defects at the same discharge quantity and voltage were systematically analyzed. Results show that the two defects have significant differences in the type of decomposition components, the yield sizes of the components, and the changes in the component ratios. This paper proposed c(CF4 + CS2)/c(CO2) and c(CF4)/c(SO2F2) as the characteristic ratios to characterize the fault severity of the two defects, and their physical nature were revealed. The combination of the above ratios with the existing c(SO2F2)/c(SOF2 + SO2) can construct a feature quantity for identifying the two defects. [ABSTRACT FROM AUTHOR]

To study the correlation characteristics between SF6 decomposition components and discharge strength under negative DC partial discharge (PD) initiated by free metal particle defect, SF6 decomposition data were obtained under different voltages on the basis of constructing SF6 decomposition experimental platform. The variation law of SF6 decomposition components and characteristic ratios were systematically analyzed. The study shows that SF6 decomposes into CO2, SO2F2, SOF2, SO2, CF4, and a very small amount of CS2 under PD. The SOF2 concentration and its formation rate have the highest correlation with the PD discharge quantity among all components. The yield relationship of decomposition components is different at different discharge quantities. In addition, the fluctuation of each characteristic ratio is small after 72 h and can achieve a certain degree of steady state. c(CF4+CS2)/c(CO2) and the proposed c(SO2)/c(SO2F2) can be used as the fault severity characteristic ratio in free metal particle defect. [ABSTRACT FROM PUBLISHER]

To study the relationship between the short-wavelength infrared reflectance and the moisture content in the beach, we conducted a rigorous analysis using regression models under different parameter settings defined by band selection, average grain size, and sand mineralogy. The spectral reflectance data were collected by a hand-held spectrometer with 3–6 nm spectral resolution. The grain sizes measured from 0.15 to 0.92 mm, with an average of 0.75 mm. The mineral components were Quartz–Alkali Feldspar–Plagioclase and Calcite-Quartz-Alkali Feldspar-Plagioclase. We found the best spectral bands were located at 1400, 1900, and 2200 nm. The regression model on these selected bands yielded an R-squared over 0.74 and RMSE lower than 0.05. We expect this study to contribute toward understanding the spectral response of beach sand with regard to its moisture content more clearly as we also consider its mineral composition and grain size. [ABSTRACT FROM PUBLISHER]

This paper investigated quench initiation and propagation characteristics of grease insulation coils by conducting thermal-quench tests for two GdBCO single-pancake coils, namely a coil co-wound with silicon grease (termed as the SiG coil) and the other coil co-wound with Kapton tape (termed as the INS coil), as turn-to-turn insulation. The test results confirmed that the SiG coil exhibited better thermal and electrical stabilities when compared with the INS coil because the operating current could be bypassed in the radial direction through the turn-to-turn contacts when a local hot spot was generated. The SiG coil had superior thermal and electrical stabilities. However, a nonrecovering resistive zone could be generated because excessive Joule heat energy could be induced by radial current flow due to the existence of radial resistance that was mainly generated by silicon grease and nonsuperconducting materials including the substrate, stabilizer, and buffer layers within the high-temperature superconductor (HTS). Therefore, it is essential to consider critical Joule heat energy that is influenced by operating current and stored magnetic energy as well as the radial resistance to achieve self-protective 2G HTS coils. [ABSTRACT FROM PUBLISHER]

Although there is much active research on software-defined radios (SDRs) with receive (RX) or transmit (TX) functionality, little work has been done on SDR transceivers supporting frequency division duplex (FDD). In this paper, we present a new circuit concept in which a distributed TX circuit cancels the transmitted signal at a reverse RX port through destructive interference while adding signal constructively at a forward TX port. We pair the distributed transmitter with a receiver-tracking PA degeneration technique to suppress the injected noise from TX circuits in the RX band. The system does not require off-chip filters or circulators, but still achieves both SDR flexibility and both FDD and time division duplex function. Measurements from the transceiver implemented in 65-nm CMOS show a frequency tuning range of 0.3–1.6 GHz with TX–RX isolation >23 dB and transmitted power up to 19 dBm. [ABSTRACT FROM AUTHOR]

During the operation of gas insulated switchgear (GIS) and gas insulated line (GIL), the conducting bar generates joule heat, and a non-uniform distribution of temperature forms in the bulk of the insulators. As the load varies, the temperature distribution changes, and it has an influence on the bulk conductivity of the insulators. Moreover, the surface charge accumulation can be affected. In order to clarify this, a surface charge measurement system which could achieve temperature control was constructed, and a model insulator with truncated cone type was employed, as well as two types of high-voltage electrodes, i.e. plate and needle electrodes. It is found that both polarities of charges existed on the insulator surface when a dc high voltage was applied to the plate or needle electrode, and homo-charge density was much higher than that of hetero-one. As for the plate electrode, homo-charges resulted from micro-discharges, while generated by corona in the case with needle electrode. As temperature increased, homo-charge density decreased, kept unchanged and increased for three cases, respectively, i.e. when dc voltage was applied to the plate electrode, and when voltage of 3 and 20 kV was applied to the needle electrode. Moreover, a simulation model involving multi-physics was established, which included heat conduction in solid and transient field changing from the initial capacitive to stationary resistive field distribution. It is proved that the electric conduction of the insulator bulk contributed to the accumulation of hetero-charges, and temperature could enhance this. Besides, when temperature increased, the corona inception voltage reduced, and hence the homo-charge density increased. Due to the effect of neutralization and the differences in the sensitivity of corona with different intensity to temperature, the tendency of homo-charge density changing with temperature for the three cases showed distinct in the experiments. [ABSTRACT FROM PUBLISHER]

In high-precision detector systems, the power supply usually need to be ultra-low noise and stabilized. Design and implementation of an ultra-low noise power supply is described in this paper. We implemented and tested the power system actually in a scientific CCD (Charge-coupled device) detector system, but the design structure could be used in many high-precision detector systems as infrared detector or high-energy particles detectors. The power system uses DC-DC switching regulators and LDO regulators for power generating, uses multi-stage filters for noise reducing. Multi-output power supply with noise about 40~\mu \text V\mathbf {rms} is achieved, the overall energy efficiency is 68.5%. Besides, the electronic system takes the low temperature as an important consideration, and the system has been tested in an extreme environment as low as 193 K. [ABSTRACT FROM AUTHOR]

Collaborative representation (CR), known as an effective way to address the signal representation (regression) problem, has achieved remarkable success in visual classification. According to our theoretical analysis, the subspace learning issue can also be deemed as a signal representation problem. Therefore, we extend the graph embedding (GE) framework as a CR model to improve the discriminating power of the subspace learning algorithm. The new GE framework, which is named collaborative GE (CGE) framework, enjoys many desirable properties of CR. From theoretical analysis, CGE is robust to the noise and has the same computational complexity as GE. From experimental analysis, CGE can generally enhance the subspace learning algorithms and a reasonable regularization parameter can be inferred from its intrinsic graph. Several state-of-the-art subspace learning algorithms are plugged into our framework to produce their collaborative versions. Meanwhile, by exploring the intrinsic relation among GE methods, we present a new collaborative method named collaborative class-scattering locality preserving projections (CCSLPPs). The results of extensive experiments on ORL, AR, Scene15, Caltech256, LFW-A, and OU-ISIR-A databases demonstrate that the collaborative versions consistently outperform their original algorithms with a remarkable improvement and CCSLPP gets the best performance compared with all used methods. [ABSTRACT FROM AUTHOR]

Moving target detection is an important technique in visual surveillance systems. If a camera is freely moving, it becomes more difficult to detect a moving target, especially in the environment of a wide-range background. To compensate for the global motion of a wide-range background, a disparity-based adaptive multi-homography method is proposed. The proposed method comprises four steps: 1) feature point extraction; 2) generation of adaptive multi-homography matrices using motion grouping; 3) generation of background model and detection of background; and 4) target detection. Experimental results show that the proposed method can robustly detect moving targets in sequences taken by a freely moving camera. [ABSTRACT FROM PUBLISHER]

Helical coil electromagnetic launchers (HEMLs) using brush-commutation strategy, solving the problem of synchronization control perfectly, have a promising prospect in accelerating big mass to high speed. One of its current technical bottlenecks is the arc erosion caused by the brush commutation, which damages the current-fed rails, besides reduces the energy conversion efficiency of HEMLs. During the commutation process, the leading, trailing edges of the commutating brush act as the closing, opening switches, respectively. When the brush-controlled circuit turns ON and OFF, the commutation-induced voltage (CIV) is a key to research the arc erosion of brushes. It can be concluded that the CIV is proportional to the current, the commutation inductance gradient (CIG), and the projectile velocity by a theoretical analysis and numerical simulation. The CIG is a critical structural parameter of HEMLs, and the formulas of the CIG for single and multiple turn commutation are deduced. Compared with single turn commutation, the current in the commutating turns changes moderately, and the CIV is lower in the case of multiple turn commutation. Two armatures of 35 turns and 135 turns were tested in an HEML of coil-unit barrel; the experimental results are in good agreement with the theoretical calculations. [ABSTRACT FROM AUTHOR]

We initiate the study of passive environment-assisted communication via a quantum channel, modeled as a unitary interaction between the information carrying system and an environment. In this model, the environment is controlled by a benevolent helper, who can set its initial state such as to assist sender and receiver of the communication link (the case of a malicious environment, also known as jammer, or arbitrarily varying channel, is essentially well-understood and comprehensively reviewed). Here, after setting out precise definitions, focusing on the problem of quantum communication, we show that entanglement plays a crucial role in this problem: indeed, the environment-assisted capacity where the helper is restricted to product states between the channel uses is different from the one with unrestricted helper. Furthermore, prior shared entanglement between the helper and the receiver makes a difference, too. [ABSTRACT FROM AUTHOR]

We introduce potential capacities of quantum channels in an operational way and provide upper bounds for these quantities, which quantify the ultimate limit of usefulness of a channel for a given task in the best possible context. Unfortunately, except for a few isolated cases, potential capacities seem to be as hard to compute as their plain analogues. We thus study upper bounds on some potential capacities. For the classical capacity, we give an upper bound in terms of the entanglement of formation. To establish a bound for the quantum and private capacity, we first lift the channel to a Hadamard channel and then prove that the quantum and private capacity of a Hadamard channel is strongly additive, implying that for these channels, potential and plain capacity are equal. Employing these upper bounds, we show that if a channel is noisy, however close it is to the noiseless channel, then it cannot be activated into the noiseless channel by any other contextual channel; this conclusion holds for all the three capacities. We also discuss the so-called environment-assisted quantum capacity, because we are able to characterize its potential version. [ABSTRACT FROM AUTHOR]

We have discovered that there exist dynamic scaling behaviors for the nucleation and the saturation field in hysteresis loop of Co/Pt multilayer thin films. The dynamic hysteresis measurement was carried out with variation of sweeping rate of cycling fields, which reveals that not only the coercivity and the loop area, well known as the Steinmetz law, but also the nucleation and the saturation field exhibit similar scaling behaviors. We have systematically analyzed the scaling factor of nucleation, coercivity, and saturation phase with respect to the Co sublayer thickness. Local structural configuration depending on the Co sublayer thickness were examined by extended X-ray absorption fine structure, implying that damping parameter and interface effect modifies the coercivity and nucleation scaling factor but does not significantly induce the change in saturation scaling factors. [ABSTRACT FROM AUTHOR]

As a kind of future weapon, an electric launcher uses high current to drive projectiles reaching velocities on the order of several kilometers per second. As is known, high and rapidly changing currents through electric launcher can lead to strong radiation. A PSPICE model is used to analyze the frequency content of currents in the helical electromagnetic launcher (HEML) by nonuniform fast Fourier transform. The frequency spectrum is mainly determined by conducting duration and intensity of entering current. The main frequency of HEML is low, but high-frequency components can produce more radiation. The total radiation can be considerable. [ABSTRACT FROM PUBLISHER]

As L-shaped array can provide good angle estimation performance and is easy to implement, many two-dimensional (2-D) direction-of-arrival (DOA) estimation algorithms have been developed. In this letter, we present a low-computation-complexity DOA estimation method for L-shaped array that exploits the conjugate symmetry property of the array manifold matrix to increase the effective array aperture and improve the angle estimation performance. Hence, the DOA estimation is obtained via combining propagator method (PM) with ESPRIT algorithm, which does not need spectrum peak search or additional angle pair matching procedure. Simulation results demonstrate the effectiveness of the proposed method. [ABSTRACT FROM PUBLISHER]

A reconfigurable system with high-speed data transmission and high-precision time measurement capability based on optical fiber and FPGA technology is presented in this paper. The design is generic for applications where data transmission and time measurement are required. This could be adapted to particle physic experiment, large imaging system, quantum bit commitment, etc. The system is extendable and reconfigurable, allowing data transmission in a bandwidth of 7.5 Gbps. TDC-GPX for time measurement, signal conditioning, data storage, real-time data processing with an FPGA with standard interface such as USB 2.0 and an expansion interface are included. As an example of an application, the design for bit commitment experiment is fully introduced. [ABSTRACT FROM PUBLISHER]

Industrial wireless sensor networks (IWSNs) have the potential to contribute significantly in areas such as cable replacement, mobility, flexibility, and cost reduction. Nevertheless, the industrial environment that the IWSNs operate in is very challenging because of dust, heat, water, electromagnetic interference, and interference from other wireless devices, which make it difficult for current IWSNs to guarantee reliable real-time communication. In this paper, we present a novel method based on the segmented slot assignment, fast slot competition, and free node concept that will improve the reliability and real-time communication significantly so that more advanced applications can be enabled. The main purpose of the algorithms is to improve the retransmission efficiency for time-division-multiple-access-based multihop IWSNs by using limited shared slot resources more efficiently. More importantly, the proposed algorithms support efficient slot rescheduling caused by link or node failure. We evaluate the proposed methods by using simulations and a real implementation targeting monitoring of welder machines. Our obtained results show that the proposed method outperforms the first published and most widely used IWSN standard called WirelessHART. [ABSTRACT FROM PUBLISHER]

The main type of obstacles of practical applications of quantum key distribution (QKD) network are various attacks on detection. Measurement-device-independent QKD (MDIQKD) protocol is immune to all these attacks, and thus, a strong candidate for network security. Recently, several proof-of-principle demonstrations of MDIQKD have been performed. Although novel, those experiments are implemented in the laboratory with secure key rates less than 0.1 b/s. Besides, they need manual calibration frequently to maintain the system performance. These aspects render these demonstrations far from practicability. Thus, justification is extremely crucial for practical deployment into the field environment. Here, by developing an automatic feedback MDIQKD system operated at a high clock rate, we perform a field test via deployed fiber network of 30 km total length achieving a 16.9 b/s secure key rate. The result lays the foundation for a global quantum network, which can shield from all the detection-side attacks. [ABSTRACT FROM PUBLISHER]