Search

The stability of dual active bridge converter (DAB) is threatened when feeding the constant power loads (CPLs). This article proposes a deep reinforcement learning-based backstepping control strategy to solve this problem. First, a nonlinear disturbance observer is adopted to estimate the large-signal nonlinear disturbance. Then, a backstepping controller is used to stabilize the voltage response of the DAB under the large-signal disturbance. Finally, a compensation method based on deep reinforcement learning is developed to intelligently minimize output voltage tracking error and improve the operating efficiency of the system. The proposed controller can guarantee system stability under the large-signal disturbance of the CPL and achieve a fast dynamic response with accurate voltage tracking; it is more adaptive by using the deep reinforcement learning technique through the learning of its neural networks. The effectiveness of the proposed controller is verified by experiments. [ABSTRACT FROM AUTHOR]

Text based Visual Question Answering (TextVQA) is a recently raised challenge requiring models to read text in images and answer natural language questions by jointly reasoning over the question, textual information and visual content. Introduction of this new modality - Optical Character Recognition (OCR) tokens ushers in demanding reasoning requirements. Most of the state-of-the-art (SoTA) VQA methods fail when answer these questions because of three reasons: (1) poor text reading ability; (2) lack of textual-visual reasoning capacity; and (3) choosing discriminative answering mechanism over generative couterpart (although this has been further addressed by M4C). In this paper, we propose an end-to-end structured multimodal attention (SMA) neural network to mainly solve the first two issues above. SMA first uses a structural graph representation to encode the object-object, object-text and text-text relationships appearing in the image, and then designs a multimodal graph attention network to reason over it. Finally, the outputs from the above modules are processed by a global-local attentional answering module to produce an answer splicing together tokens from both OCR and general vocabulary iteratively by following M4C. Our proposed model outperforms the SoTA models on TextVQA dataset and two tasks of ST-VQA dataset among all models except pre-training based TAP. Demonstrating strong reasoning ability, it also won first place in TextVQA Challenge 2020. We extensively test different OCR methods on several reasoning models and investigate the impact of gradually increased OCR performance on TextVQA benchmark. With better OCR results, different models share dramatic improvement over the VQA accuracy, but our model benefits most blessed by strong textual-visual reasoning ability. To grant our method an upper bound and make a fair testing base available for further works, we also provide human-annotated ground-truth OCR annotations for the TextVQA dataset, which were not given in the original release. The code and ground-truth OCR annotations for the TextVQA dataset are available at https://github.com/ChenyuGAO-CS/SMA. [ABSTRACT FROM AUTHOR]

Partial discharge (PD) measurements under repetitive impulse voltages are critical for the qualification of inverter-fed motor insulation systems. Severe electromagnetic interference due to high frequency switching from power electronic devices can cause the traditional PD detection techniques of sinusoidal voltage unfeasible. This article presents the design of an Archimedes spiral antenna that can work effectively for PD detection under fast rise time repetitive impulse voltages. The antenna structure is optimized by a media superstrate with a high dielectric constant over the radiant surface. Through the optimized design, both the gains of the antenna in the 0.5–1.5 GHz frequency range and the signal-to-noise ratio for PD detection are increased substantially. Modeling and experimental results prove that the gain of the antenna can reach 2.5 dB in the frequency range of 500–900 MHz and become higher than 7.0 dB in the frequency range of 900–2.0 GHz, with a voltage standing–wave ratio smaller than 1.4. This seems to be a significant achievement for PD detection under fast rise time impulse voltages. [ABSTRACT FROM AUTHOR]

In modulated wideband converter (MWC), all signals are aliased and sampled indiscriminately, which reduces the reconstruction accuracy of target signal. The focused compressed sampling uses optimized mixing sequence to increase the mixing gain of the target signal. To establish the sequence library with selective characteristics, this brief analyzes the signal components aliased into baseband, then proposes a sequence optimization method based on the greedy algorithm. Simulation and experiments show that the optimized sequence possesses a high gain for a desired frequency band signal. In the three-carrier signal test, as the approximate desired signal power ratio (DSR) is raised to 0.99 in baseband, the normalized mean square error (NMSE) of the reconstructed desired signal is improved by nearly 10 dB compared to using pseudorandom sequences. [ABSTRACT FROM AUTHOR]

In this work, we reported a theoretical model for a novel piezoelectric micromachined ultrasonic transducer (PMUT). The sensing cells with hexagonal transduction diaphragms in the PMUT array are designed and arranged in a bioinspired honeycomb structure. Based on the classical piece-wise approach commonly used for PMUT with circular diaphragm, an enhanced theoretical model is developed. In this model, the deflection of the honeycomb diaphragm is analytically determined by the volume deflection, rather than the path deflection commonly used in the previous models. The hexagonal diaphragm is converted as a circular one, whose radius is derived by correcting the characteristic size of the hexagonal diaphragm multiplied by a correction factor. The calculated parameters by the reported theoretical model of a transduction cell in the PMUT array, such as resonant frequency and static deflection, matched well with the data obtained by finite element analysis and experiment. Special investigations are made on the effects of top electrode coverage and transduction layer thickness towards PMUT’s performance. The reported theoretical model in this work is suitable not only for PMUTs with circular and hexagonal transduction cells, but also for square and other arbitrary shape diaphragms. This enhanced theoretical model provides an interesting insight on design optimizations of PMUT with various transduction cell geometries. [2022-0093] [ABSTRACT FROM AUTHOR]

Electricity price forecasting (EPF) is essential to the bidding strategy formulation and market operation. Since EPF is important in the electricity market, lots of forecasting approaches are proposed. However, the new scene caused by the volatility of renewable power generation and other volatility factors has made previous methods inaccurate and inapplicable. To address this problem, we propose an online self-adaptive forecasting method based on random forest, which is different from the traditional batch learning. Our approach takes possible fluctuations of the market into consideration, and adapts to them by maintaining training sets of different sizes. A case study using actual electricity market data has shown that our proposed approach obtains higher accuracy than ordinary approaches, as well as sheds light on possible concept drift in the market. [ABSTRACT FROM AUTHOR]

Semantic scene completion (SSC) is a computer vision task aiming to simultaneously infer the occupancy and semantic labels for each voxel in a scene from partial information consisting of a depth image and/or a RGB image. As a voxel-wise labeling task, the key for SSC is how to effectively model the visual and geometrical variations to complete the scene. To this end, we propose the Anisotropic Network (AIC-Net), with novel convolutional modules that can model varying anisotropic receptive fields voxel-wisely in a computationally efficient manner. The basic idea to achieve such anisotropy is to decompose 3D convolution into three consecutive dimensional convolutions, and determine the dimension-wise kernels on the fly. One module, termed kernel-selection anisotropic (KSA) convolution, adaptively selects the optimal kernel sizes for each dimensional convolution from a set of candidate kernels, and the other module, termed kernel-modulation anisotropic (KMA) convolution, directly modulates a single convolutional kernel for each dimension to derive more flexible receptive field. By stacking multiple such anisotropic modules, the 3D context modeling capability and flexibility can be further enhanced. Moreover, we present a new end-to-end trainable framework to approach the SSC task avoiding the expensive TSDF pre-processing as in many existing methods. Extensive experiments on SSC benchmarks show the advantage of the proposed methods. [ABSTRACT FROM AUTHOR]

In this paper, we investigate security and energy efficiency of multicast communication in cognitive satellite terrestrial networks. To achieve high spectrum efficiency and low interference power transmissions in the cognitive terrestrial network, the rate splitting multiple access scheme (RSMA) is employed to prompt the massive access of the secondary users. In particular, we derive analytical and asymptotic closed-form expressions for secrecy outage probability of the satellite terrestrial network with eavesdropping, interfering, and imperfect channel state information. Hence, a beamforming (BF) scheme based on RSMA for secondary network is proposed to suppress eavesdropping, mitigate interference, and enhance energy efficiency of the proposed system model. Then, we transform non-convex optimization constraints into resoluble problems utilizing Taylor series expansion and successive convex approximation. Finally, simulation results corroborate the above theoretical analysis and highlight the superiority of the designed BF scheme on security and energy efficiency. [ABSTRACT FROM AUTHOR]

Although smart home has received wide attention in recent years, numerous scholars focus more on energy optimization strategy than energy dispatch hardware device (named energy router). Meanwhile, this energy router should have several features, i.e., high renewable energy utilization, energy multi-port and low volume. Thus, this paper designs a nine-port energy router regarding smart home and proposes a multimode hierarchical management strategy for this energy router. First, for the multi-port demand of wind, solar, storage and utilization, this paper presents a nine-port energy router to improve the renewable energy consumption and power supply flexibility. In addition, to reduce the volume of the energy router, a non-isolated AC/DC hybrid topology is constructed through embedding the integrated power electronic converters, which achieves the miniaturization of the energy router. In order to improve the renewable energy utilization rate, the decentralized module control is proposed for the components of energy router to provide the voltage and frequency support for system, and realizes the power sharing of distributed generations (DGs). Furthermore, the power exchange control with three-mode switching is proposed to guarantee the global energy flow balance under complex conditions. Eventually, the feasibility of the energy router is verified by the simulation and experiments. [ABSTRACT FROM AUTHOR]

High renewable energy penetration increases the electricity seasonal imbalance in the long-term timescale. Power system planning needs to consider the optimal configuration of various flexibility resources and electricity balance in different timescales. The coupling of multiple timescales largely increases the computation complexity of the power system planning problem. Thus, this paper presents an efficient source-grid-storage co-planning model which incorporates a year-round hourly operation simulation. To improve the computation efficiency of the planning model, from the temporal scale, a self-adaptive compact panorama time series (CPTS) model is applied, which greatly reduces the number of variables related to short-term decisions. From the spatial scale, a network-constrained relaxed clustered unit commitment (NC-RCUC) model is introduced, which significantly reduces the number of variables related to unit commitment decisions. Case studies on the modified Garver’s 6-node system and HRP-38 system prove the validation and efficiency of the proposed model (“HRP” stands for high renewable penetration). The studies on the China power grid in 2035 demonstrate the future planning results of generation, transmission and storage in China power systems based on the proposed model. [ABSTRACT FROM AUTHOR]

Text spotting in natural scene images is of great importance for many image understanding tasks. It includes two sub-tasks: text detection and recognition. In this work, we propose a unified network that simultaneously localizes and recognizes text with a single forward pass, avoiding intermediate processes such as image cropping and feature re-calculation, word separation, and character grouping. The overall framework is trained end-to-end and is able to spot text of arbitrary shapes. The convolutional features are calculated only once and shared by both the detection and recognition modules. Through multi-task training, the learned features become more discriminative and improve the overall performance. By employing a 2D attention model in word recognition, the issue of text irregularity is robustly addressed. The attention model provides the spatial location for each character, which not only helps local feature extraction in word recognition, but also indicates an orientation angle to refine text localization. Experiments demonstrate that our proposed method can achieve state-of-the-art performance on several commonly used text spotting benchmarks, including both regular and irregular datasets. Extensive ablation experiments are performed to verify the effectiveness of each module design. [ABSTRACT FROM AUTHOR]

Battery fault diagnosis is crucial for stable, reliable, and safe operation of electric vehicles, especially the thermal runaway early warning. Developing methods for early failure detection and reducing safety risks from failing high energy lithium-ion batteries has become a major challenge for industry. In this article, a real-time early fault diagnosis scheme for lithium-ion batteries is proposed. By applying both the discrete Fréchet distance and local outlier factor to the voltage and temperature data of the battery cell/module that measured in real time, the battery cell that will have thermal runaway is detected before thermal runaway happens. Compared with the widely used single parameter based diagnosis approach, the proposed one considerably improve the reliability of the fault diagnosis and reduce the false diagnosis rate. The effectiveness of the proposed method is validated with the operational data from electric vehicles with/without thermal runaway in daily use. [ABSTRACT FROM AUTHOR]

State-of-charge (SoC) imbalance and bus voltage deviation are two of the main problems in autonomous dc microgrids. Based on this concern, this paper presents an improved dual-quadrant SoC weighted control strategy and a distributed optimization control method to achieve SoC balance, ensuring accurate power-sharing and bus voltage recovery. Firstly, this paper couples the injected/released power with the current SoC and observed average SoC value to weight the droop coefficient, which is based on the charge/discharge mode for the energy storage system. Then a secondary controller is designed based on distributed optimal control to eliminate the bus voltage deviation caused by the line impedance difference. The proposed optimal control method optimizes the average bus voltage to the nominal value and achieve accurate power-sharing by constructing the correlated variables and voltage independent intermediate variables exchanged among bulk energy storage units (ESUs). Since the voltage observer cannot accurately observe the true average bus voltage under the communication delay, the proposed distributed optimal control method without the voltage observer can ensure that the average bus voltage is optimized to the nominal value, thus improving the robustness of the control system. Finally, the correctness and effectiveness of the proposed method are verified in Simulink/MATLAB. [ABSTRACT FROM AUTHOR]

Distributed renewable energy source is an advisable solution for dc microgrids to reduce fuel consumption and CO2 emission. In such microgrids, the installation of two or more battery energy storage (BES) units is utilized to compensate the power imbalance between the sources and loads. Nevertheless, energy management with numerous BES units does not simultaneously consider the impacts of distributed generators (DGs) and constant power loads (CPLs). Since the inaccurate current sharing will shorten the lifetime of the batteries and cause instability problem, this article proposes a distributed secondary $H_{\infty }$ consensus approach based on the dynamic event-triggered communication method to realize accurate current sharing and efficient operation in the presence of numerous DGs and CPLs. First, the whole state-space function model of the dc microgrid consisting of DGs, batteries, resistive loads, and CPLs, is first built in detail. This model is further transformed into standard linear heterogeneous multiagent systems, which provides an indispensable preprocessing for advanced control strategy application. Then, the distributed secondary $H_{\infty }$ consensus approach based on the foresaid systems is designed to achieve accurate current sharing. For reducing the communication among batteries and the controller updating frequency, the dynamic event-triggered communication method is proposed. Compared with existing event-triggered methods, the communication and controller updating frequency of the proposed dynamic event-triggered method have been reduced a lot. Additionally, the proposed method can not only avoid the Zeno behavior, but also obtain the lowest bound of the sampled time interval. Finally, the numerical simulation results and experimental results verify the effectiveness of the proposed control strategy. [ABSTRACT FROM AUTHOR]

Nowadays, with the rapid development of hydrogen powered vehicles, the demand for hydrogen refueling stations (HRSs) in the transportation field is growing. The renewable energy, e.g., wind energy, is used to produce and store hydrogen on site for HRSs, which may be a relatively cheap and clean solution. Considering the wind farm (WF) and HRSs belong to different entities, a new cooperative operation model of the WF-HRSs combined system is proposed. Specifically, the robust optimization methods are implemented to characterize the uncertainties of wind power and market electricity price, respectively, to alleviate risk. To ensure the fairness of profit allocation, the bargaining power is measured by various contribution levels of each stakeholder based on the Nash–Harsanyi bargaining game theory. Furthermore, the model is transformed into two sequential subproblems: energy trading problem (SP1) and the payment bargaining problem (SP2). Accordingly, a solving technique adopting column and constraint generation algorithm is provided to solve the SP1. In addition, for the sake of privacy protection, a distributed approach based on data-centric mode is developed to solve SP2. Finally, the numerical results can validate the effectiveness and scalability for the proposed scheme and algorithms. [ABSTRACT FROM AUTHOR]

In this article, a highly efficient hp -adaptive scheme has been developed for the discontinuous Galerkin time-domain (DGTD) method. In order to achieve the ${h}$ - and ${p}$ -adaptive behaviors, a hierarchical octree-based recursive meshing procedure has been introduced. In the hp -adaptive scheme, the transformation method of the basis function is developed to convert the fields between the parent element and the child elements and between the higher- and the lower-order bases. Instead of the time-consuming field update process according to the governing equations, the error based on modified reference solutions can be rapidly and accurately calculated in the ${h}$ - and ${p}$ -adaptions by using the proposed transformation, thus flexibly manipulating the hp -adaptive procedure. Some 3-D radiation and scattering examples are given to demonstrate the good accuracy and high efficiency of the proposed hp -adaptive DGTD method. [ABSTRACT FROM AUTHOR]

Renewable energy is an advisable choice to reduce fuel consumption and $\rm CO_{2}$ emission. Therein, wind energy and solar energy are the most promising contributors to reach this goal. Although the hybrid wind/solar system has been widely studied, the real-time current sharing based on their maximum capacities is rarely achieved in terms of seconds. Based on this, this paper proposes an accurate current sharing and voltage regulation approach in hybrid wind/solar systems, which is based on distributed adaptive dynamic programming approach. Firstly, the equivalent wind/solar model is built, which is an indispensable preprocessing to achieve the complementary between wind energy and solar energy. Therein, the wind energy and solar energy can output relative current according to their respective capacity ratio, which ensure the maximum utilization ratio of renewable energy source. Furthermore, current sharing and voltage regulation problem is switched into optimal control problem. Under this effect, each source agent aims to obtain the optimal control variable and achieve accurate current sharing/voltage regulation. Moreover, an adaptive dynamic programming approach based on Bellman principle is proposed. It can achieve accurate current sharing and voltage regulation. Finally, the simulation results are provided to illustrate the performance of the proposed adaptive dynamic programming approach. [ABSTRACT FROM AUTHOR]

Periodical validation and calibration to microgrid models are necessary for investigating the behaviors of microgrids and evaluating the stability of connected power grids. However, instead of calibrating the problematic parameters that vary from their actual values, the commonly used methods normally adjust the selected sensitive parameters. The performance of the calibrated models is still in doubt. Therefore, this article proposes a generic approach for microgrid model calibration, which aims to accurately locate the indeed “ill-conditioned” problematic parameters, as well as address the multiple solutions issue. With the application of sensitivity and correlation analysis, the potentially problematic parameters (PPPs), which are sensitive and independent, are selected to ensure the accuracy and uniqueness of parameter calibration results. Then, the problematic parameters are further screened out by comparing the energy distribution of PPPs and model errors in the frequency domain. Also, hybrid dynamic simulation based parameter estimation is employed to adjust the problematic parameters, by which each device in microgrids can be calibrated independently. Finally, the simulation results demonstrate the outstanding performance in parameter calibration of a real microgrid. [ABSTRACT FROM AUTHOR]

Skeleton-based action recognition is widely used in varied areas, e.g., surveillance and human-machine interaction. Existing models are mainly learned in a supervised manner, thus heavily depending on large-scale labeled data, which could be infeasible when labels are prohibitively expensive. In this paper, we propose a novel Contrast-Reconstruction Representation Learning network (CRRL) that simultaneously captures postures and motion dynamics for unsupervised skeleton-based action recognition. It consists of three parts: Sequence Reconstructor (SER), Contrastive Motion Learner (CML), and Information Fuser (INF). SER learns representation from skeleton coordinate sequence via reconstruction. However the learned representation tends to focus on trivial postural coordinates and be hesitant in motion learning. To enhance the learning of motions, CML performs contrastive learning between the representation learned from coordinate sequences and additional velocity sequences, respectively. Finally, in the INF module, we explore varied strategies to combine SER and CML, and propose to couple postures and motions via a knowledge-distillation based fusion strategy which transfers the motion learning from CML to SER. Experimental results on several benchmarks, i.e., NTU RGB+D 60/120, PKU-MMD, CMU, and NW-UCLA, demonstrate the promise of the our method by outperforming state-of-the-art approaches. [ABSTRACT FROM AUTHOR]

In this article, we propose an equivalent circuit theory-assisted deep learning approach to accelerate the design of metasurfaces. By combining the filter equivalent circuit theory and a sophisticated deep learning model, designers can achieve efficient metasurface designs. Compared with most existing metasurface generative design methods that rely on arbitrarily generated training dataset (TDS), the proposed method can adaptively produce highly relevant and low-noise training samples under the guidance of filter equivalent circuit theory, resulting in a significantly narrowed target solution space and improved model training efficiency. Furthermore, we select the variational autoencoder (VAE) as a generative model, which can compress the raw training samples into a lower-dimensional latent space where optimization methods, such as genetic algorithm, can be more efficiently executed to find the optimal design than a brute-force search. To verify the effectiveness of the proposed method, we apply it in the creation of three examples of frequency selective surfaces (FSSs), presenting wide-band, dual-band, and band-stop responses. Experimental results show that the proposed method can realize much faster and more stable convergence than deep learning design methods without domain knowledge. [ABSTRACT FROM AUTHOR]

In recent years, a large variety of short-term load forecasting methodologies have been proposed, but a common drawback of them is that require a large amount of historical data to train the model. However, in reality, the cases in limited historical data would be more frequently occurred, for example, a new house or substation is built. To tackle this issue, a combining forecasting method of deep residual neural network (ResNet)-based transfer learning for residential loads is proposed in this article. To improve the accuracy and reliability of transfer learning, a novel deep ResNet with dual skip connections (ResNet-DSC) is proposed as the base model. Then with sparse training data, a Bayesian probability weighted averaging (approach is proposed, to address the model combination parameters estimation problem. In addition, probability density function forecasting is also delivered by the method. For demonstration, a series of experiments, with actual residential load data, was performed. With the comparison with other transfer learning and nontransfer learning approaches, the effectiveness and improvement of the proposed method have been validated. [ABSTRACT FROM AUTHOR]

In the electrical engineering field, there are two types of system which can potentially generate resonance under excitation, one is the electrical system with inductor and capacitor, the other is the mechanical system with spring-mass characteristic. A lot of research on active damping control algorithms for grid-connected inverters with LCL filter and inverter-driven machine with multirotating masses have been demonstrated. However, research works for these two systems were carried out independently and there is a lack of systematic comparison for modelling and control between these two systems. This article will unify the mathematical models and active damping control algorithms for these two systems. It is found that the mathematical models and control structures are fundamentally the same. The existing or future potential active damping control algorithms used in electrical system can be applied in mechanical system and vice versa to avoid reinventing the wheel. Parameter sensitivity analysis for controller and feedback gains was performed for electrical systems in the discrete z-domain. For mechanical systems, it is found that a substantial electromagnetic torque overshoot was introduced when applying the active damping control and it was analyzed quantitatively with various damping coefficients to guide the inverter design. Finally, experimental tests were done to verify the findings. [ABSTRACT FROM AUTHOR]

Referring Expression Comprehension (REC) aims to localize an image region of a given object described by a natural-language expression. While promising performance has been demonstrated, existing REC algorithms make a strong assumption that training data feeding into a model are given upfront, which degrades its practicality for real-world scenarios. In this paper, we propose Continual Referring Expression Comprehension (CREC), a new setting for REC, where a model is learning on a stream of incoming tasks. In order to continuously improve the model on sequential tasks without forgetting prior learned knowledge and without repeatedly re-training from a scratch, we propose an effective baseline method named Dual Modular Memorization (DMM), which alleviates the problem of catastrophic forgetting by two memorization modules: Implicit-Memory and Explicit-Memory. Specifically, the former module aims to constrain drastic changes to important parameters learned on old tasks when learning a new task; while the latter module maintains a buffer pool to dynamically select and store representative samples of each seen task for future rehearsal. We create three benchmarks for the new CREC setting, by respectively re-splitting three widely-used REC datasets RefCOCO, RefCOCO+ and RefCOCOg into sequential tasks. Extensive experiments on the constructed benchmarks demonstrate that our DMM method significantly outperforms other alternatives, based on two popular REC backbones. We make the source code and benchmarks publicly available to foster future progress in this field: https://github.com/zackschen/DMM. [ABSTRACT FROM AUTHOR]

Scene graph generation (SGG) is built on top of detected objects to predict object pairwise visual relations for describing the image content abstraction. Existing works have revealed that if the links between objects are given as prior knowledge, the performance of SGG is significantly improved. Inspired by this observation, in this article, we propose a relation regularized network (R2-Net), which can predict whether there is a relationship between two objects and encode this relation into object feature refinement and better SGG. Specifically, we first construct an affinity matrix among detected objects to represent the probability of a relationship between two objects. Graph convolution networks (GCNs) over this relation affinity matrix are then used as object encoders, producing relation-regularized representations of objects. With these relation-regularized features, our R2-Net can effectively refine object labels and generate scene graphs. Extensive experiments are conducted on the visual genome dataset for three SGG tasks (i.e., predicate classification, scene graph classification, and scene graph detection), demonstrating the effectiveness of our proposed method. Ablation studies also verify the key roles of our proposed components in performance improvement. [ABSTRACT FROM AUTHOR]

The output-based event-triggered cooperative output regulation problem is addressed for constrained linear heterogeneous multiagent system in this article. In light of the robust control theory, $H_{\infty }$ leader-following consensus with respect to exogenous signals, including both disturbance to be rejected and reference state of leader to be tracked, is guaranteed. Meanwhile, the system performance alleviates degradation through a model recovery anti-windup technique while encountering input saturation. Furthermore, the follower’s self-state observer, the leader-state observer, and the anti-windup auxiliary system are integrated into a comprehensive system, and a unified event-triggering mechanism of full states is addressed. A fixed lower bound of sampled interval is adopted such that the frequency of data transmission gets reduced and no Zeno-behavior happens. Both the input and output of the follower’s controller and anti-windup compensator hold constant, respectively, during the event-triggered intervals such that the resulting output-based event-triggered controller can be directly implemented in a digital platform. Finally, a simulation example is provided to illustrate the effectiveness. [ABSTRACT FROM AUTHOR]

Vehicle-to-everything communication (V2X) and mobile edge computing (MEC) have introduced significant technological innovations for autonomous ground vehicles (AGVs), enabling most driving task computations to be performed with edge servers with higher quality and effectively reducing computational power consumption for AGVs. For reducing more computational power consumption which can reach hundreds of watts, sending more offloading tasks to edge servers can result in high latency and congestion when the network resources are finite and even cause serious traffic accidents. Besides, in V2X network, multiple transmission paths, including vehicle-to-vehicle (V2V) links and vehicle-to-infrastructure (V2I) links, can provide more valuable resources for sending offloading tasks, but they increase the complexity of resource management. To improve the utilization efficiency of network resources and ensure low power consumption for AGVs, first, this paper analyzes mathematical models of bandwidth, delay and power consumption based on different computing modes and transmission paths. Second, we propose an allocation algorithm for network resource allocation, transmission paths and computing modes to minimize power consumption according to the states of the V2X-MEC network measured and predicted by the navigation and positioning system. Third, a traversal approach simplified by a road segmentation method is adopted to solve for the optimal parameters. Through simulations, we compare our solution against an allocation approach based on the shortest path method, minimum average delay method and game theory method to illustrate the effectiveness of our proposed method in reducing the power consumption and transmission delay of uplinks. [ABSTRACT FROM AUTHOR]

The main factors limiting the power scaling of high-peak-power high-brightness quasi-continuous wave (QCW) fiber laser oscillator and a possible solution are analyzed in this paper. The impacts of fiber length and grating parameters on the output power are studied by rate equations. Under the optimized parameters, a high-brightness QCW fiber laser oscillator with a peak power of 7.3 kW is presented experimentally by using a 24 meter long spindle-shaped Yb-doped fiber with a constant core cladding ratio, and that is the best performance of peak power and brightness in near-single-mode QCW fiber lasers at present to the best of our knowledge. The corresponding pulse energy is 0.57 J and the beam quality M2 factor is about 1.43 under the condition of repetition frequency of 1 kHz and pulse width of 100 μs. The details are analyzed, and also the limiting factors of further power scaling, strategies of suppressing nonlinear effects and the principle of optical device parameter optimization. [ABSTRACT FROM AUTHOR]

The improvements in signal-to-noise ratio (SNR) are indispensable to obtain the desired output signal of partial discharge (PD) under square wave pulsewidth modulation (PWM) voltage. In this article, the effect of complex PWM voltage parameters (rise time, fall time, frequency, duty cycle) on the SNR of PD tests is investigated. The results revealed that the sensor's high-frequency gain should be increased to improve the SNR under faster switching. PD impulses are more likely to be submerged in the residual interference under high-frequency voltage. The SNR can be improved by only detecting the PD at the rising front because the PD signal overlaps with commutation noise at the falling front of short duty cycle voltage, such as 0.03% duty cycle. The above results can reference the PWM parameters and improve the SNR in PD tests for rotating machines controlled by power electronics. [ABSTRACT FROM AUTHOR]

In this communication, a novel message passing interface (MPI) parallel algorithm for nodal discontinuous Galerkin time-domain (NDGTD) method has been developed. A unified MPI + MPI technique has been introduced for extreme parallelism on a large-scale computer cluster. Through the data transmission between CPU nodes using MPI persistent nonblocking two-side communication and the direct data connection between processors in the same node via MPI shared memory windows, a two-layered parallel architecture is implemented to minimize the communication. To further accelerate the solution of the multiscale problems, the local time stepping (LTS) technique has been employed in the NDGTD method. A fast time step estimation method has been presented in this communication. With high overlap between the information transmission and the data calculation, the proposed MPI + MPI scheme overcomes the degradation of the parallel efficiency of the pure MPI technique in the scenario of the LTS technique and the large-scale CPU cores. Up to 94% parallel efficiency in 6400 CPU cores is achieved for the average single-core loading about 1700 finite elements, and 18 times acceleration for time step estimation can be obtained with the fourth-order basis function. Three practical complex examples are given to demonstrate a good performance of the proposed method. [ABSTRACT FROM AUTHOR]

Maintaining a balance between convergence and diversity is particularly crucial in evolutionary multiobjective optimization. Recently, a novel dominance relation called “strengthened dominance relation” (SDR) is proposed, which outperforms the existing dominance relations in balancing convergence and diversity. In this article, two points that influence the performance of SDR are studied and a new dominance relation, which is mainly based on SDR, is proposed (CSDR). An adaptation strategy is presented to dynamically adjust the dominance relation according to the current generation number. The CSDR is embedded into NSGA-II to substitute the Pareto dominance, labeled as NSGA-II/CSDR. The performance of our proposed method is validated by comparing it with five state-of-the-art algorithms on commonly used benchmark problems. NSGA-II/CSDR outperforms other algorithms in the most test instances considering both convergence and diversity. [ABSTRACT FROM AUTHOR]

In this study, a nano-displacement measurement system is proposed and demonstrated both theoretically and experimentally, which was based on a modified Mach-Zehnder (M-Z) interferometer using two conjugated orbital angular momentum (OAM) beams. In contrast to the previous M-Z-based OAM interferometer, a reflection module is inserted into the reference arm instead of a simple mirror. As a result, the effect of the transverse position-dependence phase-shift caused by the dove prism can be clearly eliminated and a stable and robust (off-axis insensitive) petal-like interference pattern can be obtained successfully. More importantly, a significant rotation angle of the petal-like pattern vs. the tiny displacement of the tested object can be clearly observed. In accordance with the modified measurement setup, a novel phase-demodulation method enabling to quickly and accurately characterize the rotation angle of the petal-like interference-patterns is proposed and demonstrated also. A tiny displacement ranging from 50 to 800 nm with resolution of $\sim 50$ pm has been measured successfully. The proposed approach may find applications in not only the ultra-high precision displacement sensor, but also the temperature, strain, and refractive index sensors. [ABSTRACT FROM AUTHOR]

To enhance the resilience of power systems toward the temporal and spatial impacts caused by extreme weather events, e.g., hurricanes, the flexibility of cloud data centers (CDCs) is treated as a kind of efficient demand response. Since the workloads of CDCs have the shifting capacity between different locations and time slots, a day-ahead unit commitment problem including data centers is proposed to explore the integrated spatial and temporal flexibility of the workloads to its full extent through task migration. Considering the uncertainty of probability density functions, the line failure rates, workload arrival rates, power loads are integrated into an ambiguity set. The scheduling process of generators and CDCs is modeled as a two-stage distributionally robust optimization problem, which is reformulated as a large-scale deterministic mixed-integer linear programming problem and solved by the multicuts Benders decomposition method. The performance of the proposed scheduling strategy is tested in both the IEEE 24-bus RTS system and the three-area RTS-96 system. The results reveal that the method could mitigate the adverse impacts of hurricanes by enhancing the resilience of power systems and decreasing the dropping workloads of CDCs. [ABSTRACT FROM AUTHOR]

We propose and demonstrate, with hardware, the first experimental technique to measure a vertical component of FinFET source resistance. Forward bias is applied to the well-to-source p-n junction, and the forward voltage at constant current density, ${V}_{{\text {fb}}}$ , is measured as a function of source-current, ${I}_{\text {ds}}$. The source-current, ${I}_{\text {ds}}$ , is varied by variation of ${V}_{\text {gs}}$ , and an effective vertical resistance is calculated, ${R}_{\text {ve}} =$ dVfb/dIds. Significant self-heating effects of the p-n junction are observed, and we demonstrate a technique to correct ${V}_{\text {fb}}$ for the local rise in temperature. We find that our technique gives well-behaved results for vertical source resistance. Quantitative ${R}_{\text {ve}}$ values are found to be consistent with technology characterization of (linear region) total device resistance. [ABSTRACT FROM AUTHOR]

The square-wave voltage signal injection method with the proportional-integral (PI) observer is commonly adopted to estimate the rotor position of interior permanent magnet synchronous motor (IPMSM). However, the saliency estimation method with PI observer is a continuous injection method and the polarity identification method is a discrete one, which causes extra transition time and complexity. To solve this problem, a fast unipolar sequence-pulse injection method is proposed to unify the process of saliency and polarity identification and the complexity of the algorithm as well as the identification time are reduced. At the same time, a four-stage search algorithm is proposed to ensure the fast convergence and improve the accuracy by reducing the number of the pulse voltage signals. The experiment on motor drive platform verifies the effectiveness of the proposed method to IPMSMs with different saliency ratios. [ABSTRACT FROM AUTHOR]

Accurate and reliable machine performance degradation tracking and remaining useful life (RUL) prognosis establish the foundation for predictive maintenance scheduling toward improved safety and productivity of machine operations. In general, machine performance degradation exhibits a nonlinear and nonhomogeneous pattern that arises from time-varying degradation rate and abrupt performance changes. To address this challenge and improve the generalizability of degradation modeling, in this article, we present a stochastic modeling technique based on the Lévy process, which generalizes system variations as the accumulations of successive and jump increments. The developed Lévy process model consists of two terms: a linear Brownian motion term for capturing the gradual degradation with time-varying rates and a nonhomogenous compound Poisson process term for capturing transient performance changes. By calculating the moments of the characteristic function of the Lévy model, explicit expressions for the probability distributions of predicted performance degradation and RUL are derived. To obtain the time-varying parameters in the Lévy model, Markov chain Monte Carlo is investigated. The developed technique is evaluated through simulation and run-to-failure tests of roller and ball bearings, and the good performance of the developed Lévy model is confirmed. [ABSTRACT FROM AUTHOR]

Recently, convolutional neural networks (CNNs) have been successfully developed and used in the classification of polarimetric synthetic aperture radar (PolSAR) images. However, they often suffer from some problems, such as time-consuming, unsatisfactory detail-preservation, and bad effectiveness given limited training samples. Focusing on these problems, we propose a complex-valued CNN (CV-CNN)-based algorithm for PolSAR image classification in this article. On the one hand, a superpixel-oriented (SPO) scheme is employed to reduce the computational cost of the algorithm and preserve image details simultaneously, which takes superpixels instead of single pixels as classification units. In particular, to meet the input requirement of CV-CNN, three alternative methods of superpixel regularization are designed and compared. On the other hand, considering that both measured data (MD) and manually designed polarimetric features (PFs) have their own advantages, the hybrid data (HD) combining them is employed to drive CV-CNN, which is helpful to improve the effectiveness of the algorithm. We perform experiments on three actual PolSAR image data sets acquired by AIRSAR and Radarsat-2 systems as well as a semisimulated data set. The experimental results demonstrate that, compared to conventional pixel-oriented methods, the proposed SPO scheme is much more time-efficient and is also beneficial to detail preservation. Moreover, the CV-CNN driven by HD generally obtains consistently better classification results than that driven by pure MD or manually designed PFs. [ABSTRACT FROM AUTHOR]

A compact interlinking converter modular (ICM) for hybrid ac/dc microgrids with a distributed storage (DS) bus for storage integration is proposed in this article. The advanced ICM for hybrid ac/dc/ds microgrids involves the nine-switch topology with compact configuration and decreased number of active operating switches, which structurally equivalents to afford a dc interconnection port for dc subgrid, an ac interconnection port for ac subgrid, a ds interconnection port for ds subgrid, and two dc–dc converters for the interconnection ports for parallel DSs. In addition to the popular advantage of component elements saving, the suggested ICM, aggregating various interlinking converter cells (ICCs) into an ICM, is also proficient of participating computations from all ICC sensors to enhance the overall modular performance without building extra communication connection links between ICCs. With this in mind, a decentralized power management control strategy is realized by employing a coordination power control methodology to systematically handle various distributed generations (DGs), DSs, and ICM in a decentralized manner with improved communication fault ride-through capability. The controller hardware-in-the-loop experiments on a real-time digital simulator platform validates the achievement of the advanced ICM application in hybrid ac/dc/ds microgrids and the effectiveness of the associated decentralized power management scheme. [ABSTRACT FROM AUTHOR]

To achieve an individual and continuous power distribution for multireceiver wireless power transfer (WPT) systems, a novel multifrequency modulation method has been proposed. In the proposed method, a look-up table and delta-sigma modulation scheme are introduced to generate a mixed-frequency driving voltage pulse by synthesizing it from the given voltage pulses. The components of synthesized driving voltage pulse have continuously and individually varying amplitudes of multiple specific frequencies. With the proposed modulation method, only a standard full-bridge inverter is employed in the transmitting side to achieve power distribution among receivers, leading to a simple configuration of the transmitting source. Moreover, the proposed method has a small calculation burden and can be easily extended to the system with more receivers. When compared with the existing multifrequency modulation method, a lower switching frequency is obtained in the proposed method. Finally, the effectiveness of the proposed modulation method is verified theoretically and experimentally. [ABSTRACT FROM AUTHOR]

The electric vehicle technology is one of the most promising candidates to reduce fuel consumption and $\rm CO_2$ emission. Although electric vehicles have been widely promoted by governments around the world, their development is seriously hampered due to charger unavailability and range anxiety. Based on this, this paper designs an energy interaction converter between two electric vehicles, which is controlled through disturbance observer based sliding mode control algorithm. For this converter, three main demands should be satisfied, i.e., high power density, weak source and constant power load. Therein, weak source whose minimum short circuit ratio (SCR) belongs to Jia et al., 2020 and Wang et al., 2020, is always called weak grid. Firstly, the equivalent impedance switching process is introduced to eliminate the impact of weak source. Meanwhile, the equivalent six channel interleaved floating dual boost converter is chosen to satisfy the high power density demand, whose generalized state-space function is further built to provide an indispensable preprocessing for following controller design. Moreover, in order to solve the problem regarding low frequency/sub-synchronous oscillation caused through constant power load feature regarding the energy consumption vehicle and weak source feature regarding the energy supply vehicle, a disturbance observer based sliding mode control algorithm is proposed through using generalized state-space function to provide standard DC power with both constant voltage and power. Furthermore, the proportional-resonant controller is proposed to solve the current sharing problem among six parallel channels, which reduces the heat loss and improves the service life of the device. Finally, simulation and experimental results verify the high performance of the proposed control algorithm. [ABSTRACT FROM AUTHOR]

In this paper, the dependence of XLPE electrical breakdown on carrier mobility controlled by both electric field and temperature is evaluated by simulations and experiments. The molecular chain displacement model incorporating the thermoelectric effect of carrier mobility is examined to elucidate the experimental results. Trapping and de-trapping characteristics of space charge and mobility dynamics are modeled using hopping and Poole-Frenkel mechanisms. The experimental outcomes reveal that the DC electrical breakdown reduces with corresponding increment in temperature and thickness. Trapping parameters are estimated from experiments for simulations. The DC breakdown strength from experiments represents an inverse power relation with both the insulation thickness and temperature. The simulation results of molecular chain model with hopping mobility express more permanence with experimental results in comparison with Poole-Frenkel mobility. [ABSTRACT FROM AUTHOR]

The allocation of fault current limiters (FCLs) is increasingly challenging in transmission systems these days. Specifically, the utilized deterministic expected short-circuit fault (SCF) scenarios are prone to cause over-configuration of FCLs. Moreover, the well-established local switching framework (LSF) renders inappropriate FCL switching and may further harm the system safe operation. Aiming at the above deficiencies, a novel 5G-based centralized switch FCL (CSF) framework as well as a method to allocate such flexible FCLs optimally is proposed in this paper. In the proposed CSF, the FCLs are switched by a FCL dispatching (FD) model considering system security constraints of both fault current and voltage sags. By exploiting the fast communication capability of 5G network as well as an off-line fault scanning strategy, the FD model is enabled to give online FCL switching schemes to meet the fast requirement of power system protection. Moreover, considering the probabilistic characteristic of SCFs, a bi-level FCL allocation model is established, in which the upper-level model sites and sizes FCLs considering the installation and expected switching costs while the lower-level model determines the optimal switched FCLs under each specific SCF scenario. Finally, numerical results are provided to verify the proposed allocation model, including its defending effect against SCFs in terms of fault current limiting, voltage sags relieving, as well as its cost-effectiveness. [ABSTRACT FROM AUTHOR]

Modular multilevel converters (MMCs) in high-voltage dc applications usually adopt a distributed control architecture to manage a large number of submodules (SMs) through a communication network. The communication congestion and network disconnection might lead to communication interruption (CI) and eventually cause the system to malfunction. In this article, a resilient operation strategy is proposed and studied to ride-through the CI fault, in order to prevent frequent fault SM bypassing, replacement, or even system shutdown. The analysis of the MMC distributed control system with the presence of CI indicates that the insertion index of the faulted SM might become constant, which distorts the output current and results in overvoltage of the communication interrupted SM (CI-SM). The CI-SM capacitor voltage prediction can be used to determine the MMC safe operation period after CI occurs. During the safe operation period, the CI-SM power balance is sustained by utilizing prestored phase signals to generate a sinusoidal insertion index according to its capacitor voltage tracking error. Two operation modes are proposed and analyzed to ensure the MMC stable operation under various conditions. The system protection is sensibly used only if the CI duration exceeds a safe operation period, which avoids frequent SM cut-off. Good agreement of the CI-SM capacitor voltage is achieved between the theoretical and simulation results. The effectiveness and robustness of the proposed MMC resilient operation are experimentally confirmed. [ABSTRACT FROM AUTHOR]

This article proposes a circulating current and second harmonic current (SHC) suppression method by introducing virtual impedance into the circulating current and output current feedback control loops of the dc solid-state transformer (DCSST). The virtual impedance is employed to adjust impedance in the circulating current feedback control loop to decrease the circulating current when the transmission efficiency of each dual active bridge is different. And a SHC suppression method, which introduces the virtual impedance containing a bandpass filter, is adopted to reduce the SHC and improve the dynamic performance of the DCSST. Moreover, a relatively independent modular control method, which is conducive to the modular expansion of the DCSST, is proposed. Compared with traditional DCSST control methods, the proposed method can simultaneously realize the output current sharing and effectively reduce the SHC with modular control. Finally, a DCSST prototype is built and the results of the experiment verify the validity and effectiveness of the proposed control strategy and solutions. [ABSTRACT FROM AUTHOR]