Search

Rain-like layer removal from hot-rolled steel strip surface has been proven to be a workable measure for suppressing the false alarms frequently triggered in automated visual inspection (AVI) instruments. This article extends the scope of the “rain-like layer” from dispersed waterdrops to splashing water streaks and tiny white droplets. And a targeted method with both channel-wise and spatial-wise attention, namely attentive dual residual generative adversarial network (ADRGAN), is proposed. Meanwhile, a newly updated steel surface image dataset with typical natures of a “rain-like layer” gathered from an actual hot-rolling line, Steel_Rain, is opened for the first time. The comparison of experimental results between our proposed network and 11 prestigious networks shows that our ADRGAN-restored images are the closest to the ground-truth images on six public datasets, especially on the newly opened industrial dataset Steel_Rain; it yields the best scores of 56.8627 peak signal to noise ratio (PSNR), 0.9980 structural similarity index (SSIM), 0.134 mean-square error (MSE) and 0.006 learned perceptual image patch similarity (LPIPS). In the final verification test, the concept of rain-like layer removal has been proved to perform best in defect inspection, where four traditional defect detection algorithms are involved. And as expected, defect detection methods assisted by ADRGAN yield the minimum false alarms.

The missile-target assignment (MTA) is a typical weapon-target assignment problem in Command and Control of modern warfare. Despite the significance of the problem, traditional algorithms still lack efficiency, solution quality, and practicability in the adversarial environment. In this article, we propose a data-driven policy optimization with deep reinforcement learning (PODRL) for the adversarial MTA. We design a comprehensive reward function to motivate the optimization of assignment policy. As such, the learned policy can implicitly model the penetration of missiles under an adversarial environment in a data-driven way. We also present a fair sample strategy to improve the sample efficiency and accelerate the policy optimization. Experimental results show that PODRL can adaptively generate satisfactory solutions in both small-scale and large-scale instances. Furthermore, we evaluate the effectiveness of PODRL in a multiobjective scenario. The result demonstrates that a well-optimized policy can achieve high-quality allocation and demand forecast of the missile resources simultaneously.

MOV is often used to limit the overvoltage generated by the current breaking of DC breaker to protect power electronic device from breakdown. However, while suppressing overvoltage, the MOV also causes the degradation of continuous operating voltage capability. This paper proposes a novel solid-state switch scheme with high voltage utilization efficiency by using modular gapped-MOV (MG-MOV) for DC breakers. The breakable gap is first applied to the protection of power electronic devices. By using the gap to share the voltage in the static state and be broken down in the dynamic state, the overvoltage and continuous operating voltage of the solid-state switch can be decoupled. Moreover, based on the modular scheme, the instability of the gap breakdown and the voltage imbalance of the series devices can be solved well. The working mechanism, voltage-balance and stability issue of the MG-MOV are discussed in detail in this paper, and a 15kA prototype is developed to verify the feasibility of the novel solid-state switch scheme. The experiment results show that voltage utilization efficiency can be increased from 44.4% to 72.0% by using MG-MOV.

The diode bridge submodule (SM) is one of the basic units of a hybrid high-voltage dc circuit breaker. This article works on the low inductance design of a single SM and series SMs. For the layout design of a single SM, this article proposes a simple and clear +-shaped SM layout, the low inductance design is simplified to the optimization of two laminated busbars with a simple structure. Based on the +-shaped SM layout, this article further proposes an SM structure with holes in diodes region, the stray inductance is further reduced, and only one integrated laminated busbar is needed in a single SM. For the layout design of series SMs, this article makes full use of the layout of a single SM and the mutual inductances between the SMs, series SMs are alternately connected in reverse. The single SM layout proposed in this article is effective in reducing the stray inductance. The loop stray inductance in stage II reduced from 148.4 to 72.3 nH, a 51.3% reduction. When the two parallel IGBTs turn off a 6 kA fault current in 2 μ s, the voltage spike is reduced from 786.1 to 423.4 V, a 46.1% reduction. And the stray inductance difference of the bidirectional current path is controlled to about 10 nH. The layout of series SMs proposed in this article is effective in reducing the length of the series SMs, the length is reduced by 16.7%. The experimental results show that the SMs layout proposed in this article is effective.

This paper investigates the formation control problem for a group of nonholonomic mobile robots (NMRs) with unknown parameters and deception attacks. The information transmitted among different robots is represented by a directed graph and only a subset of the robots can obtain the full information of the desired trajectory directly. For those robots which cannot access to the reference trajectory directly, distributed estimators are designed to estimate the unknown trajectory information by using only locally available information. Besides, the sensor-to-controller transmitting channels and the communication among connected robots are suffering from deception attacks. Adaptive laws and compensation terms are designed to handle the issue of attack-induced uncertainties. Then, a novel distributed resilient formation control scheme is designed, based on which a sufficient condition is developed to guarantee that the formation errors converge to a compact set and all the closed-loop signals are bounded. Experimental results are given to validate the theoretical studies.

The mechanism for the radiation efficiency local minimum in antennas, especially for open-slot antennas and T-shaped antennas, has been investigated. It is found that the local minimum in those antennas always corresponds to an ideal differential mode. Furthermore, the frequency position of the ideal differential mode is decided by the parasitic parts of the antennas. Simple methods for estimating the local minimum position are proposed based on the mechanism we have found. In addition, the duality relation between those two types of antennas is also discussed. Our findings provide a guideline for designing mobile phone antennas with improved efficiency.

In this article, we investigate the output consensus tracking problem for a class of high-order nonlinear systems with unknown parameters, uncertain external disturbances, and intermittent actuator faults. Under the directed topology conditions, a novel distributed adaptive controller is proposed. The common time-varying trajectory is allowed to be totally unknown by part of subsystems. Therefore, the assumption on the linearly parameterized trajectory signal in most literature is no longer needed. To achieve the relaxation, extra distributed parameter estimators are introduced in all subsystems. Besides, to handle the actuator faults occurring at possibly infinite times, a new adaptive compensation technique is adopted. It is shown that with the proposed scheme, all closed-loop signals are globally uniformly bounded and asymptotically output consensus tracking can be achieved.

Metal oxide varistors (MOVs) are usually applied to clamp the voltage across a dc circuit breaker (DCCB) after fault current elimination. This letter focuses on the transient process of a DCCB turn- off and has three contributions. First, it reveals that snubber circuit inductance, the internal and external stray inductances of an MOV could cause a significant overvoltage across a DCCB, which exceeds 47.5% of its nominal clamping voltage. The overvoltage mechanism is clearly analyzed based on the working principle of a DCCB, providing theoretical equations to estimate the overvoltage magnitude. Second, a nonlinear model of an MOV is proposed to simplify the calculation and provide an accurate estimation result. Third, a new arrangement of series submodules with discrete MOVs is proposed to suppress the overvoltage. A hybrid 500 kV/25 kA DCCB prototype with series insulated gate bipolar transistors (IGBTs) is implemented to validate the proposed analysis. The series-IGBT rating voltage is 1440 kV, and the MOV with a nominal clamping voltage of 800 kV is used to protect IGBTs. However, a 25 kA turn- off experiment clearly shows that the overvoltage increased to 1203 kV by an external stray inductance of 39.5 μ H and d i /d t of 1480 A/ μ s. The relative error between measurements and simulations is within 3%. Besides, simulations on the new arrangement show the overvoltage is decreased by 23.4% than before.

Over the course of the past two decades, a great deal of researchers has investigated synchronization of networked systems with determined node dynamics. However, in many realistic situations, noise is inevitable in man-made and naturally occurring networked systems. Therefore, dynamical networked systems with stochastic perturbations have gained substantial attention and have been extensively studied both in theoretical research and in practical applications. The primary goal of this paper is focused on the novel synchronization protocol design problem of nonlinear stochastic dynamical networked systems. Sufficient conditions are given to select protocol parameters. By employing stochastic analysis techniques and selecting appropriate Lyapunov functions, we proved that global synchronization of nonlinear stochastic dynamical networked systems can be reached in mean square. Numerical examples are depicted to demonstrate the efficiency of the theoretical results.

Automated visual inspection (AVI) instrument of surface defects for hot-rolled steel strips is conventionally installed closely before the terminal crimping machine, where the adjacent upstream process is laminar spray cooling. Waterdrops, spreading more or less over the steel strip surface, often trigger false alarms, which is a quite common problem in AVI. Stimulated by the idea of image rain removal in visual enhancement field, this article considers the surface waterdrops, pseudodefects in essence, as a conceptual “rain-like layer.” A targeted method, namely progressive recurrent generative adversarial network (PReGAN), is designed for active waterdrop tracking and fine-grained image inpainting. Meanwhile, a steel surface database (2400 raw images with the resolution of $1000\times1000$ ) captured from actual hot-rolling line is constructed for the first time for open evaluation of waterdrop removal. The experimental results indicate that images enhanced by the PReGAN perform the most informative and spotless, with 52.2073 peak signal-to-noise ratio (PSNR) and 0.9502 structural similarity (SSIM) index, when compared with four prestigious networks. Assisted by the PReGAN, the false alarms are proved to be reduced at least a half during the application tests using four traditional simple detection methods.

Extraction of the current commutation loop (CCL) inductance is necessary for the overvoltage analysis of power devices. In general, each transient process of a converter corresponds to a specific CCL. However, there is no such single CCL in the full-bridge module in hybrid dc circuit breakers during the turn- off transient. First, this article proposes a coupled inductance model of full-bridge modules. The model has less number of inductance compared to the traditional partial inductance model. Second, this article proposes the extraction method of inductances that can be easily obtained by both simulation and measurement. Moreover, the measured overvoltage waveforms of insulated gate bipolar translators (IGBTs) in the full-bridge module agree with the simulation results. Third, we propose an equivalent circuit of the series full-bridge modules in a 500 kV hybrid circuit breaker based on the proposed coupled inductance model. The mutual inductances among different modules are considered due to the compact structure of the series modules. The measured results show that the overvoltage of the IGBT in the series module could be well-predicted by the proposed equivalent circuit.

In the paper, synchronization of nonlinear dynamical complex networks with stochastic coupling under adaptive PI control is considered. In light of the relative states information of the network nodes, node- and edge-based adaptive PI synchronization protocols are proposed and several sufficient conditions are derived. By employing stochastic analysis techniques combined with a version of the LaSalle’s invariant principles applied to stochastic process and Lyapunov function approaches, synchronization of complex networks with nonlinear dynamics and stochastic coupling reaches in mean square via these two types of adaptive PI control protocols are proved. The proposed adaptive PI control protocols for synchronization of complex dynamical networks are useful for future realistic engineering systems design. Numerical examples are finally simulated to validate the theoretical analysis.

This article addresses the leader-following consensus of a class of multi-agent systems (MASs) subjected to saturation. Unlike previous literature, the followers are with heterogeneous dynamics. To solve this problem, we employ the low-gain feedback technique and the parameterized algebraic Riccati equations to design the controllers. For the fixed and switching network topologies, sufficient conditions are put in place to guarantee the semiglobal stability of the consensus error system. Numerical results are also provided to validate the effectiveness of the control design.

This paper investigates an event-based tracking problem of networked second-order systems with unknown inertias and switching networks, where the leader’s trajectory is dynamic and is available to only a subset of followers. Besides, the followers have only local interactions. A new class of distributed event-triggered control protocols with and without velocity measurements are designed to guarantee the tracking performance. Furthermore, it is strictly proved that the proposed algorithms guarantee a fully Zeno-free triggering for all the agents. Finally, simulation results are provided to demonstrate the effectiveness of the proposed theoretical results.

In memory computing and high-end distributed storage demand low latency, high throughput, and zero data loss simultaneously from datacenter networks. Existing reactive congestion control approaches cannot both minimize queuing latency and ensure zero data loss. A token-oriented proactive approach can achieve them together by controlling congestion even before sending data packets. However, state-of-the-art token-oriented approaches only strive to optimize network-level metrics: maximizing throughput while achieving flow-level fairness. This article answers the question of how to support objective-aware traffic scheduling in token-oriented approaches. The novelty of Token-Oriented in-network Prioritization (TOP) is that it prioritizes tokens instead of data packets. We make three contributions. Via simulations over a hypothetical TOP system, our first contribution is demonstrating the potential performance gain that can be brought by TOP. Second, we investigate the applicability of TOP. Although the overhead of enabling necessary TOP features in switches is trivial, we find that mainstream commodity datacenter switches do not support them. We hence propose a readily-deployable remedy to achieve in-network prioritization by pushing both switch and end-host hardware capacity to an extreme end. Lastly, we implement a running TOP system with Linux hosts and commodity switches, and evaluate TOP in testbeds and with large-scale simulations for various scenarios.

Nowadays, deep neural architectures have acquired great achievements in many domains, such as image processing and natural language processing. In this paper, we hope to provide new perspectives for the future exploration of novel artificial neural architectures via reviewing the proposal and development of existing architectures. We first roughly divide the influence domain of intrinsic motivations on some common deep neural architectures into three categories: information processing, information transmission and learning strategy. Furthermore, to illustrate how deep neural architectures are motivated and developed, motivation and architecture details of three deep neural networks, namely convolutional neural network (CNN), recurrent neural network (RNN) and generative adversarial network (GAN), are introduced respectively. Moreover, the evolution of these neural architectures are also elaborated in this paper. At last, this review is concluded and several promising research topics about deep neural architectures in the future are discussed.

Multifunctional convergence is one of the key physical features in future generation networks and Internet of things architectures. In this paper, we propose and experimentally demonstrate a unified terahertz (THz) system operating in the 300 GHz band, with a potential of simultaneously enabling high-speed communication and high-resolution ranging over a common optical infrastructure. Both THz communication and THz sensing signals are generated based on THz photonics and cutting-edge terahertz transceiver technologies. In the experiment, 16-quadrature amplitude modulation modulated THz signal is generated by photo-mixing two free running lasers for the communication, and linear frequency modulated (LFM) THz pulses are generated based on optical interferometer-based frequency-to-time mapping (FTM) for sensing. The experimental results show that up to 56 Gbit/s net rate is successfully transmitted over a 2 m free-space line-of-sight link, and the THz LFM pulses with a time-bandwidth product of up to 207 are successfully generated, which is potentially able to enable a cm-scale range resolution. We also investigate the frequency multiplexing schemes for two signals by changing the channel gap at the transmitter side. To the best of our knowledge, such a system represents the first demonstration of integrated generation system in the THz region above 300 GHz, which has great potential in prospective applications of future converged networks.

This paper focuses on the construction of distributed observers in the presence of arbitrarily large communication time delays. In contrast with the traditional centralized observer with the ability to acquire full output of the plant, we design a set of distributed observers, each having access to partial output of the plant through a distributed sensor network. More specifically, each observer obtains partial plant output and communicates with its neighboring observers through consensus protocols. The communication among the network is subject to arbitrarily large time delays. We consider three representative network topologies and for each topology establish conditions to guarantee the observation error systems be exponentially stable. We also consider the design of a pinning synchronization problem as a dual problem of the design of distributed observers. Numerical simulation is carried out to verify the effectiveness of our theoretical analysis.

In this letter, we experimentally demonstrate terahertz (THz) photonic wireless transmission of high-speed quadrature-phase-shift-keying (QPSK) and 16 quadrature amplitude modulation (16-QAM) signals in the 350-GHz band, with an effort to extend the wireless reach of ultrafast 100 Gbit/s. In the experiment, a narrowband unitraveling-carrier photodiode (UTC-PD) is used for heterodyne generation of THz communication signals, which is transparent to optical modulation formats. Experimental results show that 30 GBd/s QPSK (60 Gbit/s) and 25 GBd/s 16-QAM signals (100 Gbit/s) are successfully transmitted over a 350-GHz wireless link with an extended distance of 2 m. This enhancement is enabled by combining the techniques of employing cutting-edge narrowband UTC-PD as the photomixing THz emitter, spectrally efficient modulation format and advanced digital signal processing algorithms. The achievement makes 100 Gbit/s THz wireless communication systems promising for future smart wireless services.

Over the last decades, the cooperative design of complex networked systems has received an increasing attention in real-world engineering practices. Traditionally, each node in the network is assumed to obtain the same signal. However, each agent often possesses different measurement due to the observability or configuration of the systems. To solve the stabilization problem in this case, we aim to establish a unified framework for the cooperative control of complex network with distributed observers. In detail, different from the traditional centralized design, this paper initiates a cooperative approach by only using the local information of the networked systems. In allusion to the three kinds of representative networks, this paper establishes some sufficient or necessary conditions for the existence of network parameters that guarantee the stabilization of the LTI plants. Moreover, some corresponding algorithms are developed to find the suitable parameters of networked observers. Last but not least, numerical simulations are also provided to verify the above theoretical results.

A B-dot was used for measuring cathode emission current in the injector of a linear induction electron accelerator. The measured current waveforms in commissioning were different from our initial expectation. This could well be explained by that the measured current waveform was the sum of actual emission current and displacement current. The displacement current could be derived by the measured diode voltage and the calculated diode capacitance. The actual emission current waveform was then obtained by subtracting the displacement current from the measured waveform. The pure displacement current waveform was also measured by the B-dot, while there was no emission from a thermionic cathode. The calculated and measured displacement current waveforms were well consistent with each other. The B-dot’s sensitivity factor calibrated by comparing the aforementioned two waveforms was equal to our theoretical estimate within an error range. The obvious displacement current may also exist in the similar pulsed power facilities, which contain high-voltage and fast pulses. It can be processed by the same methods discussed in this paper.

It is well known that multi-agent systems (MASs) are ubiquitous in natural and artificial systems. This paper aims at bridging the gap between transmission noise and sampled data for robust consensus of MASs. In detail, we have developed a theoretical framework for analyzing the robust consensus of MASs with sampled-data controllers and transmission noises. Using the delay-input and discretization approaches, we obtain two sufficient conditions on the existence of sampling periods and controller parameters for robust consensus of MASs, respectively. In particular, we deduce the estimates of the convergence speeds of consensus errors for the above two methods. Finally, numerical simulations are also given to validate our theoretical results.

The interaction between two (parallel) one-way waveguides formed by photonic crystals is investigated theoretically. It is shown that when the two waveguides support modes propagating in opposite directions, they can effectively interact with each other only within a narrow guiding region where their propagation constants are nearly zero. In this coupling window, the waveguides are contra-directionally coupled and the efficiency grows monotonously with the coupling length, reaching 100% as the coupling length is large enough. When the one-way waveguides support the modes propagating in the same direction, they may be efficiently coupled through the whole guiding regime, and their coupling exhibits the same behavior as the conventional uniform waveguides. The coupling between the one-way waveguides is first analyzed with the coupled-mode approximation and then verified by rigorous numerical simulation.