Search

Great progress has been made in deep learning over the past few years, which drives the deployment of deep learning–based applications into cyber-physical systems. But the lack of interpretability for deep learning models has led to potential security holes. Recent research has found that deep neural networks are vulnerable to well-designed input examples, called adversarial examples . Such examples are often too small to detect, but they completely fool deep learning models. In practice, adversarial attacks pose a serious threat to the success of deep learning. With the continuous development of deep learning applications, adversarial examples for different fields have also received attention. In this article, we summarize the methods of generating adversarial examples in computer vision, speech recognition, and natural language processing and study the applications of adversarial examples. We also explore emerging research and open problems.

The fault current of the high‐voltage and large‐capacity flexible DC power grid has the characteristics of a fast rising rate, high peak value, and large influence area; therefore, fault current suppression equipment suitable for the flexible DC power grid has become a research hotspot. First, the fault mechanism and fault current calculation associated with the DC converter are analysed in this study; subsequently, a multiport superconducting fault‐current‐limiting circuit breaker (SFCLCB) suitable for a flexible DC power grid is designed. The circuit breaker adopts ground drainage to transfer the fault current. It only needs to be equipped with a single‐way conduction power electronic switch group and the number of switches is not affected by the number of DC lines, considerably improving the cost effectiveness of the equipment. In addition, the workflow and current‐limiting principle of SFCLCB are analysed, and the optimal matching method is determined with respect to the resistance of the superconducting current limiter and the power electronic devices of the main circuit breaker. Afterwards, the four‐terminal flexible DC transmission simulation system is observed to exhibit high efficiency in suppressing the fault current and high cost effectiveness based on comparative simulations. Finally, its engineering applicability is verified by the digital physical and hybrid simulation platform.

The utilization of demand response flexibility has become a significant method to cope with the intermittence of renewable energy sources in distributed systems. This paper proposed a new pricing method for demand response resources managed by a distribution system aggregator, which is deduced from analyzing the operating revenue within the timescale from hours to years. In the proposed model, the hourly decision-making of an aggregator is formulated as a newsvendor model and uncertainties in the long-term decisions are modelled by a backward valuation process. It maximizes the benefit of an aggregator by considering the price and quantity uncertainties of distributed load/generation in day-ahead and real-time wholesale electricity markets. Meanwhile, the coexistence of controllable and uncontrollable loads is also considered, where the former refers to electricity consumption from end-users who are equipped with smart devices for energy management, and the latter load demand of passive end-users who have no willingness or capability to participate in the demand response schemes. Finally, numerical studies are carried out to demonstrate the feasibility and effectiveness of the developed model and methods, and the impacts of active end-user percentage on the aggregator operation under the proposed pricing method are also compared and illustrated.

Replacing expensive IGBT with thyristor to design dc circuit breaker (dcCB) is attracting attention recently due to its advantages in terms of low investment and low conduction loss. However, due to the half-controlled characteristic of thyristor, the existing thyristor-based dcCBs (T-CBs) cannot be actively turned off . As a result, the opening process is not fully controllable and reliable. In this letter, a simple active T-CB (AT-CB) topology is proposed. The proposed AT-CB can interrupt both fault current and operation current actively, and the current breaking process is fast, reliable, and fully controllable. Moreover, the proposed AT-CB can be improved to be bidirectional without increasing the conduction loss, and the topology remains simple. Parameters design principle and operation process of AT-CB are presented. Scaled-down experiment verified the effectiveness of the proposed solution.

SCR-based solid-state dc circuit breaker (SCR-DCCB) topologies have been widely studied for replacing the traditional expensive insulated-gate bipolar transistor (IGBT) based DCCB. Yet, the existing SCR-DCCB cannot be actively turned- off since SCR is a half-controlled device. In this letter, an active Z-source DCCB (AZ-DCCB) is proposed by the combination of IGBT and SCR. The proposed AZ-DCCB has advantages of simple and compact topology, economical design, low conduction losses, and active and bidirectional current breaking capacity. The operation process of the proposed AZ-DCCB is discussed in detail, and the designing principle of circuit parameters is also presented. Experimental results verified the effectiveness of the proposed AZ-DCCB.

In this study, the graphene oxide (GO)-TiO2 nanocomposite film prepared by the sol–gel method was used to optimize the photoelectric performance of the perovskite solar cells (PSCs), which structure is based on the carbon electrode and no hole transport layer. Through a series of scientific experiments, it has been proved that the GO–TiO2 nanocomposite film has excellent electrical properties. The electron transport layer which containing 1 wt% of GO makes the PSCs have the most excellent performance. Compared with the controlling PSCs, the photoelectric conversion efficiency of PSCs, which containing 1 wt% GO, increased by 9.3%, from 12.44 to 13.60%. Short-circuit current density (Jsc) increased by 5.99%, from 21.55 to 22.84 mA/cm2, and open circuit voltage remained basically unchanged. It can be seen from the measurement results that the trend of incident photon-to-electron conversion efficiency spectrum is consistent with J–V characteristics, indicating that the PSCs device containing 1 wt% GO has the best electron extraction and transfer performance. However, when too much GO is used, the increased surface charge trap state will lead to the acceleration of carrier recombination and the decrease of electron transport path, and the photovoltaic parameters show a downward trend.

Smart home scheduling, facilitated by advanced metering, monitoring, and manipulation technology, plays an important role in the energy transition in terms of accommodating intermittent renewable energy and improving energy consumption efficiency. The key functionalities of home energy scheduling are usually implemented by leveraging the flexibility of household appliances, such as thermostatically controlled loads (TCLs) and energy storage units, to improve the peak-to-average ratio for utilities and reduce energy bills for customers. However, the consumption patterns of appliances are greatly influenced by a variety of factors, including real-time tariffs, ambient temperature profiles, indoor activities, and solar irradiance. Hence, smart home energy scheduling is a challenging task because most of these impacting factors are stochastic and difficult to predict. To properly model and manage the uncertainty factors associated with smart home appliance scheduling, an economic model predictive control (MPC)-based bilevel smart scheduling scheme is proposed in this paper. The comprehensive modeling of distributed generation and household appliances is performed at the single-household level. The home energy scheduling problem is formulated on two levels, with the upper level emphasizing the economic impact and the lower level focusing on capturing TCL responses. The correlations among different TCLs and their performance under the influence of various uncertainty factors, such as environmental impacts and user behaviors, are considered. The efficiency of the proposed MPC-based bilevel optimization model and the effectiveness of the home energy scheduling strategy in managing uncertainties are validated and illustrated in numerical studies.

Memtransistor, a hybrid structure that integrates the function of memristor and transistor, is a promising device prototype for the realization of complex neuromorphic learning owing to its diverse functionality and additional flexibility in emulating synaptic behaviors. Memtransistor of two-dimensional (2D) chalcogenide materials have received many interests as it has distinctive memristive mechanism quite different from conventional oxide memristors. Here, we report a memtransistor based on the two-dimensional thin films (2DTFs) of non-layered β-In2S3. The In2S3 2DTFs grown by physical vapor deposition method have microscopically visible grain boundaries (GBs) formed by the stacking and interconnecting of 2D In2S3 flakes. The memtransistors of In2S3 2DTFs show tunable bipolar resistive states with resistance ratio up to 105, endurance over 200 cycles, and a retention time of 104 s. Illumination of laser light from visible and near-infrared are able to induce intermediate resistance states in memtransistors, enabling optical-modulated multilevel memory storage. Also, the memtransistors are able to emulate the synaptic function of long-term potentiation (LTP) and long-term depression (LTD) with tunable synaptic weight in response to presynaptic stimuli of drain/gate pulses. Interestingly, the plasticity of LTP and LTD behavior can be switched in a highly tunable manner by simply varying the gate voltages. The diverse optoelectronic properties and controllable functionality of memtransistors based on the emerging 2D In2S3 offer a useful guide to potential application in electronic memory and artificial synapses.

Solar anti-icing/deicing (SADI) is an economically-efficient method of harvesting solar energy as heat for melting and removing ice. However, SADI materials with superior sunlight harvesting performances and high deicing rate remain elusive. Herein, the successful preparation of hierarchically macro/micro-structured polydimethylsiloxane (PDMS)/reduced graphene oxide (rGO) films was demonstrated based on a facile dual-template method. The as-fabricated films exhibited efficient broadband solar light absorption and high energy transfer efficiency (90.4%). When used as anti-icing/deicing materials, these SADI films exhibited excellent anti-icing/deicing performances under one sun illumination (qi = 1.0 kW m−2), i.e., long freezing delay time (tD > 2 h) and ultra-low deicing strength (∼0.2 kPa) even at an ultra-low temperature of −40 °C. The SADI surfaces showed a high deicing rate (>1.05 kg m−2 h−1) under various simulated realistic conditions over a wide range of geographical differences. Moreover, all the frost, rime, and glaze (ice thickness < 2 mm) could be easily removed within 300 s.

The interface of the business system in the power cloud environment is characterized by numerous types and multifarious services. The interface test plays an important role in ensuring the safe and stable operation of the platform. Traditional interface test only considers specific message protocols, so that test case design is weak for business process detecting, test cases are lack of comprehensiveness, test design and implementation have high repeatability. Combined with the operation characteristics of power cloud, this paper proposes an automatic interface test architecture in the power cloud environment in order to improve the efficiency of information system interface test in the power industry.

Increasing milk production is one of the key concerns in animal production. Traditional breeding has gotten limited achievement in the improvement of milk production because of its moderate heritability. Milk production traits are controlled by many genes. Thus, identifying candidate genes associated with milk production traits may provide information that can be used to enhance the accuracy of animal selection for moderately heritable traits like milk production. The genomic selection can enhance the accuracy and intensity of selection and shortening the generation interval. The genetic progress of economically important traits can be doubled with the accuracy of selection and shortening of generation interval. Genome-wide association studies (GWAS) have made possible the screening of several single nucleotide polymorphisms (SNPs) in genes associated with milk production traits in dairy cattle. In addition, RNA-sequencing is another well-established tool used to identify genes associated with milk production in dairy cattle. Although it has been widely accepted that these three methods (GWAS, RNA-seq and DNA sequencing) are considered the first step in the screening of genes, however, the outcomes from GWAS, DNA-sequencing and RNA-seq still need further verification for the establishment of bonafide causal variants via genetic replication as well as functional validation. In the current review, we have highlighted genetic markers identified (2010-to date) for their associations with milk production traits in dairy cattle. The information regarding candidate genes associated with milk production traits provided in the current review could be helpful to select the potential genetic markers for the genetic improvement of milk production traits in dairy cattle.

A dc–ac converter for voltage-source converter (VSC)-HVdc technology has a significant influence on the performance of the entire power transmission system. This paper introduces a new dc–ac converter composed of submodules and series insulated-gate bipolar transistor (IGBT) switches. The basic idea of this hybrid solution is to shape the ac voltage by submodules but, per half cycle, reconnect them to different electrical points by IGBT switches. This concept can help to reduce the quantities of submodules, thereby reducing the energy storage requirement significantly. Another advantage is that the IGBT switches can be soft switched by utilizing the high controllability of the submodules. This brings extra benefit of power losses reduction. In this paper, the operating principle of this hybrid converter is explained. Its performances are also presented in detail and compared with those popular VSC dc–ac converters. The feasibility of the new concept is also verified by simulation and experimental results.

Increasing deployment of distributed energy resources (DERs) is re-sculpturing the modern power systems in recent years. Future smart power distribution systems should be competent at accommodating extensive integration of DERs and managing the associated uncertainties at the distribution level. The electricity market has been proved to be an efficient way to employ market signals to direct behaviors of users and DERs with large capacity and homogeneous pattern. However, existing market frameworks cannot effectively handle a large number of small-scale DERs due to their diverse characteristics and arbitrary behavior patterns. In this context, an aggregated model which can represent and manage a diverse collection of DER, load, and storage is proposed. An additional trading platform, namely the energy sharing market, is established to reinforce the coordination and collaboration among various aggregators as well as operators. Energy sharing scheme is applied and a corresponding dynamic dispatch platform is designed to solve the crowdsource problem. The efficiency of the proposed model is validated by the numerical studies, and the market performance and impacts of energy sharing on the power systems are illustrated.

Prediction over heterogeneous data attracts much attention in urban computing. Recently, satellite imagery provides a new chance for urban perception but raises the problem of how to fuse visual and non-visual features. So far, the practice is to concatenate the multimodal features into a vector, which may suppress important features. Therefore, we propose a new ensemble learning framework: (1) An estimator is developed for each predictor to score its confidence, which is input adaptive. (2) By applying the output of each predictor to the input of the corresponding estimator as feedback, the estimator learns the performance of the predictor in the input-output space. When a new input is applied to produce a prediction, the similar situations will be recalled by the estimator to score the confidence of the prediction. (3) Using end-to-end training, the estimator learns the weights automatically to minimize the total loss of the neural networks. With the proposed method, data mining based urban computing and computer vision rendered urban perception can be bridged at the task of commercial activeness prediction, where the prediction based on satellite images and social context data are fused to yield better prediction than those based on single view data in the experiments.

It is vital of practical engineering significance to connect the isolated station into the DC grid smoothly without affecting other stations or triggering protective systems. To this end, an active charging method with damping control of the isolated station is proposed and will be used in a 500 kV-level DC grid in China. First, the analytical expression of starting current is calculated. The key factors influencing the starting current are analysed to provide theoretic evidence for simplifying the calculation model. Second, the selecting expression of start-up resistor is derived by simplifying the main circuit parameters, DC line parameters, control processes and so on. Third, the principle of active charging method is clarified, the selecting principle of decreasing slope and parameters of damping controller are presented, and the required time of charging process is calculated. This method does not need the start-up resistor any more but still ensure no oscillation and impulse during the active charging process. Meanwhile, it is also faster than the traditional method with a better economy. Finally, electromagnetic transient simulation models are built to verify the correctness of the resistor selecting expression and the effectiveness of the active charging strategy with damping control.

In this work, inorganic tin-doped perovskite quantum dots (PQDs) are incorporated into carbon-based perovskite solar cells (PSCs) to improve their photovoltaic performance. On the one hand, by controlling the content of Sn2+ doping, the energy level of the tin-doped PQDs can be adjusted, to realize optimized band alignment and enhanced separation of photogenerated electron–hole pairs. On the other hand, the incorporation of tin-doped PQDs provided with a relatively high acceptor concentration due to the self-p-type doping effect is able to reduce the width of the depletion region near the back surface of the perovskite, thereby enhancing the hole extraction. Particularly, after the addition of CsSn0.2Pb0.8I3 quantum dots (QDs), improvement of the power conversion efficiency (PCE) from 12.80 to 14.22% can be obtained, in comparison with the pristine device. Moreover, the experimental results are analyzed through the simulation of the one-dimensional perovskite/tin-doped PQDs heterojunction.

Because of the special low-dimensional physical properties of quantum dots (QDs) and the excellent optoelectronic properties of perovskite, the combination of perovskite solar cells (PSCs) and QDs may largely be one of the breakthroughs in device performance. However, little attention has been paid to the effect of the size of QDs on device performance except for bandgap tuning. In this work, non-toxic SnS QDs of different sizes were prepared by varying the reaction temperature of the thermal injection method. Their optical bandgaps with sizes were investigated, and they were implanted into the active layers of the carbon-based PSCs without hole transport layer. The effects of implanting SnS QDs with different sizes on the perovskite film quality and PSCs performance were studied. The cells implanted with SnS QDs of the best size (average 6.9 nm) achieved a photoelectric conversion efficiency of 14.26%, reaching a 12.42% improvement. The improved performance was mainly attributed to the presence of SnS QDs, which provided more nucleation sites for the growth of perovskite grains, and contributed to better quality perovskite films (including higher crystallinity, increased grain numbers and fewer surface defects), thus improving the light utilization, accelerating carrier transfer and reducing carrier recombination in the active layers. The successful implantation of QDs with different sizes in this paper has facilitated the development of the combination of perovskite and QDs in mesoporous PSC.

This work focuses on the electrical properties of Ni/n-GaN Schottky barrier diodes (SBDs) fabricated on GaN bulk substrates with different crystal orientations. For the SBDs on a-plane, m-plane, and s5-plane GaN, the Schottky barrier heights (SBHs) exhibited magnitudes of 0.65 eV, 0.69 eV, and 0.75 eV, respectively. The relatively small SBH of a-plane devices results in a relatively larger reverse leakage current than that of m-plane and s5-plane. In addition, the carrier concentrations extracted by C–V characteristics are comparable for the GaN substrates with different crystal orientations. The temperature-dependent I–V characteristics indicate how the barrier inhomogeneity of the Ni/GaN Schottky contacts on the a-plane GaN is smaller than the others. Based on the XPS spectra results, the discrepancy in Schottky contact characteristics stems from the different polarization charges and/or surface oxides of the different crystal planes. The low GaOx density on a-plane GaN surface contributes to the minimum barrier inhomogeneity of the corresponding SBD devices.

In this work, the groove depth of the vertical GaN trench MOSFETs have been investigated by TCAD simulation. The groove depth is found to have a crucial influence on the specific on-resistance and breakdown voltage. A deep enough groove is required to annihilate the build-in barrier between the p-body and the n-drift region close to the sidewall of the groove, in order to reduce the specific on-resistance. On the other hand, an increased groove depth will also cause electrical field peak and hence degrade the breakdown voltage. Therefore, there is a trade-off between the specific on-resistance and breakdown voltage, and optimization is required. This study has provided a quantitative analysis on the optimization scheme.

Partial fingerprint verification is a challenging task because of the few features contained in small area as well as the large rotation angle and translation between query images and template images. In this paper, we propose a new framework of partial fingerprint verification based on spatial transformer networks (STN) model, where a transform model, i.e., AlignNet network, is proposed to estimate the alignment parameters, and the verification is modeled as a binary classification task. The experimental results on the simulated datasets created from FVC2004 and the real-world dataset FVC2006 DB1 show that our method is invariant to rotation, and also robust to different kinds of scanners, and dramatically outperforms the rank-1 entry of FVC2006 participants. The EER on FVC2006 DB1 of the proposed algorithm is 3.587% compared to that of 5.564%, the best of FVC2006 DB1 entries.

A new photoelectrochemical imaging method termed scanning electrochemical photometric sensor (SEPS) is proposed in this work. It was derived from light-addressable potentiometric sensor (LAPS) and scanning photoinduced impedance microscopy (SPIM) using a front-side laser illumination at a field-effect structure. When the laser beam scans across the sensor substrate, local photocurrent changes at inversion due to the light absorption of analytes can be recorded. It will be shown that SEPS could be used for label-free living cell imaging with micro-resolution as well as real-time quantitative absorption analysis, which would broaden the applications of traditional LAPS/SPIM from potentiometric/impedance measurements to local optical analysis.

The determination and optimization of the rated capacitance of a capacitor voltage transformer (CVT) is very valuable for the optimal design of a CVT since it is a fundamental parameter for solving many problems such as minimizing the effect of stray capacitance (SC), ensuring a high accuracy of voltage measurement, and fast transient response. SC is known to seriously affect the capacitance of an ultra-high-voltage (UHV) CVT resulting in a larger additional measurement error and a higher false operation rate of relay protection; it is usually unknown or poorly quantified and cannot be measured directly. A new effective and reliable method for rated capacitance design of a new UHV equipotential shielding CVT under the interference of SCs is proposed and successfully demonstrated in this work. An equivalent mathematical model is established to derive the CVT rated capacitance and the SCs are extracted using a finite element method. The performance of this method is verified in the experiment to demonstrate the effectiveness of the method.

This paper presents a comprehensive analysis and a novel solution for the discharge problem on the half-bridge submodules’ (HBSMs’) capacitors in the hybrid modular multilevel converter (MMC)-based high-voltage direct current transmission systems under a long-term (e.g., several minutes or longer) pole-to-ground fault condition. The mechanism of this issue is analyzed, and then a solution based on fundamental-frequency reactive circulating current (FFRCC) injection is proposed. The influence of FFRCC injection is analyzed in terms of the fluctuated arm capacitor energy and the current stress of semiconductors. Moreover, an enhanced arm voltage modulation method based on HBSMs and full-bridge submodules coordinate control is proposed to accelerate the capacitor voltage balance of HBSMs. Finally, the performance of the proposed control strategy is verified by the simulation results.

Reducing the capacitance of submodules (SM) and handling short-circuit faults in HVdc systems are crucial for economical and reliable operation of modular multilevel converters (MMC). This paper proposes an SM capacitance reduction method for full-bridge submodule (FBSM)-MMC, which combines circulating current injection and ac voltage boosting. An optimized circulating current control method considering the limitation of fixed semiconductor current rating is introduced. By adding second-harmonic voltage to the reference arm voltage, the capacitor voltage ripple can be controlled to be a targeted value under all operating conditions. To further decrease the SM capacitor voltage ripples, the negative voltage state of FBSM is fully utilized to boost the ac voltage under fixed dc voltage with the modulation index up to 1.414. Consequently, the capacitance requirement for ac voltage boosting FBSM-MMC can be significantly reduced as compared to the conventional FBSM-MMC. Experimental results confirm the feasibility and validity of the proposed SM capacitance reduction method.

Pole-to-ground (PTG) fault is one of the most common DC-side faults in multi-terminal high-voltage direct-current (HVDC) systems. During the PTG fault, the loss of system transmission power is a challenging issue. This study proposes a PTG fault ride through strategy for low-impedance grounded radial multi-terminal HVDC systems. It is realised by the coordinated operation between half-/full-bridge hybrid modular multilevel converters and mechanical DC switches. The mechanical DC switch is utilised to isolate the fault and a ground return path is formed to transmit the active power during PTG fault. If the fault is located at the sending terminal, half of the rated transmission power can be kept at the entire fault duration. If the fault is located at the receiving terminal, the successfully restarted active power regulator stations can further return to normal operation after the isolation of the PTG fault. With the proposed strategy, the loss of transmission power during PTG fault is minimised, which can improve the transient angle stability of the AC grid. PSCAD/EMTDC simulation is performed to verify the proposed strategy.

In this study, anti-solvent carbon quantum dots (ASCQDs), generated via pulsed laser irradiation in chlorobenzene (CB), were adopted as an additive for defect passivation of the grain boundaries of CH3NH3PbI3 in hole-conductor-free, carbon-counter-electrode perovskite solar cells (PSCs). The size of the ASCQDs can be easily regulated by different laser fluence. The improvement of perovskite films’ morphology was revealed by scanning electron microscopy (SEM) and atomic force microscopy (AFM), which was beneficial for carrier transport. Higher photoluminescence (PL) intensity as well as prolonged carrier lifetime were proved in the ASCQDs optimized perovskite film, confirmed the laser induced ASCQDs can decrease nonradiative recombination via defect passivation, which can be further verified by the decreased defect density. As a result, the introduction of ASCQDs (1064 nm-50 mJ pulse−1 cm−1-10 min) led to a champion efficiency of 14.95%, which improved 14.04% when compared with pure CB treated PSCs (13.11%). Furthermore, the ASCQDs treated PSCs can still retain 93% and 81% of the initial power conversion efficiency after 600 h in the glovebox and in the atmosphere with humidity of 50–70%, respectively.

In this study, NaGdF4:Yb3+, Er3+@NaGdF4:Eu3+ (NGEY@NGE) core-shell rare earth nanoparticles are prepared by a simple hydrothermal method. After testing, the emission peaks are observed at 415, 525, 540 and 660 nm with an excitation light of 980 nm due to the up-conversion luminescence from rare-earth nanoparticles. Under light excitation at 395 nm, the rare-earth nanoparticles undergo down-conversion luminescence with emission peaks at 590, 610 and 695 nm. The resulting nanoparticles are introduced into the TiO2 mesoporous layer of hole-conductor-free perovskite solar cells (PSCs) based on carbon counter electrodes to broaden the spectral response of the cells from the visible range to the near-infrared and near-ultraviolet regions, which can guarantee a high light absorption efficiency and corresponding accelerated generation of electron-hole pairs. The results suggest that the mixing of NGEY@NGE nanoparticles into mesoporous TiO2 layer improves the power conversion efficiency of PSCs to 14.21%, which is 10.6% higher than that of the control cells, which can pave an effective way to improve the performance of PSCs with a broadened response range.

Using voltage-source converter high-voltage direct current (VSC-HVDC) to interconnect weak ac power systems has drawn much attention. Under the traditional vector control for the VSC that the response of AC voltage loop cannot match that of the active power loop may limit power transmission capability of the VSCs that interconnect with weak power systems. To enhance the power transmission capability and stability, this paper proposes a power compensation control that is based on the analysis of characteristics of power–voltage (PV) curves. This method can improve the damping control performance of the active power and the ac voltage by adjusting the compensated values adaptively. A small-signal model of the integrated system is established to study the performance of the proposed control method. The impact of the compensation coefficients and steady-state operating points on the system stability is studied. Motivated by an on-going back-to-back project in China, corresponding electromagnetic transient simulation models are constructed in MATLAB/Simulink environment to verify the effectiveness of the proposed method on the performance of operating limits, damping performance, disturbance withstanding capability, and fault ride-through capability.

A series of Ce 3+ , Tb 3+ co-doped Na 6 Ca 3 Si 6 O 18 phosphors have been synthesized by traditional solid state reaction. The emission spectra of the Na 6 Ca 3 Si 6 O 18 :Ce 3+ ,Tb 3+ phosphors show a blue broad band at 450 nm of the Ce 3+ and a green line peaking at 550 nm. Energy transfer from Ce 3+ to Tb 3+ ions has been investigated. Results indicate that the color of these phosphors changes from blue to green by adjusting the ratio of Ce 3+ to Tb 3+ . In addition, the mechanism of the energy transfer from the Ce 3+ to Tb 3+ ions is demonstrated to be a dipole-quadrupole mechanism. The energy transfer efficiency as well as the critical distance is also estimated. By optimizing the Tb 3+ concentration, the absolute PL quantum yield as high as 43% can be achieved. The results show that this phosphor has potential applications as a single-phased phosphor for UV white-light LEDs.

Samples of the Ba9 (Lu2-x Yx )Si6 O24 :Ce3+ (x = 0-2) blue-green phosphors were synthesized by solid-state reactions. All the samples exhibited a rhombohedral crystal structure. As the Y3+ concentration increased, the diffraction peaks shifted to the small angle region and the lattice parameters increased due to the larger ionic radius of Y3+ (r = 0.900 A) compared with that of Lu3+ (r = 0.861 A). Under 400 nm excitation, samples exhibited strong blue-green emissions around 490 nm. The emission bands had a slight blue shift that resulted from weak crystal-field splitting with increasing Y3+ concentration. Luminescence intensity and quantum efficiency (QE) decreased with increasing Y3+ concentration. The internal QE decreased from 74 to 50% and the external QE decreased from 50 to 34% as x increased from 0 to 2. The thermal stability of the Lu series was better than that of the Y-series. The excitation band peak around 400 nm matched well with the emission light from the efficient near-ultraviolet (NUV) chip. These results indicate promising applications for these NUV-based white light-emitting diodes.

The voltage source converter (VSC) based HVDC technology have been built for transferring power over long distances with high voltage in past decades. Recently, the development of modular multilevel converter (MMC) technology makes HVDC could be used for higher voltage and greater power capacity, which provides a possibility to build MMC DC grid. In this paper, the flexible sub-module voltage modulation strategy of MMC in a four-terminal MMC DC grid will be investigated and two schemes to handle the occurrence of sub-modules fault will be proposed. A four-terminal DC grid model is built in the software PSCAD/EMTDC in order to verify the effective of the flexible sub-module voltage modulation strategy applied to MMC DC grid.

As the increasing penetration of modular multilevel converter (MMC)-based high-voltage direct current (HVDC) into bulky power transmission systems, the performance of MMC on dealing with DC faults, especially on temporary DC fault in overhead transmission lines, has been becoming more and more significant. A comprehensive overview of MMC on DC fault ride-through (FRT) capability is discussed. Compared with DC fault blocking capability considered as a passive DC FRT strategy focusing on isolating DC fault from AC sides, the DC FRT capability emphasises remaining converters continuous operating during DC faults to regulate and support the connected AC grids. The principle and prerequisite of MMC on DC FRT are analysed, and then improved MMC topologies with DC fault handling capability are summarised and a corresponding comparison among them is conducted. Finally, applications of MMC with DC FRT capability on HVDC systems and its relative control strategies are presented.

The selective area growth (SAG) technique has been proved to be an effective method to achieve trench gate normally-off AlGaN/GaN MOSFET. In this paper, the two-dimensional electron gas (2DEG) transport properties of SAG AlGaN/GaN heterostructure are investigated for samples with different insertion layers. As the thickness of the GaN insertion layer increases, the mobility of 2DEG drops at first and then increases dramatically. This is explained by a shift of the location of the 2DEG from the GaN template into the GaN insertion layer. Interface contamination at the template/insertion layer interface is found to be the reason of the degradation of the transport properties. Through optimizing the thickness of the SAG GaN insertion layer, 2DEG mobility is similar to the value of as-grown AlGaN/GaN heterostructure. The proper thickness of SAG GaN insertion layer for device fabrication is 24 nm referring to the variation in 2DEG transport properties.

Due to the rapid development of the voltage-source converter (VSC), the demands for HVDC cables have increased significantly. Although more than ten VSC-HVDC projects are under construction and many more are in operations globally, the testing criteria for HVDC XLPE cables are mainly referenced to test guideline TB 496 recommended by CIGRE. Due to the large stray capacitance of the VSC-HVDC cables, the switching impulse overvoltage stress of the cables is the main electrical parameter concerned, a unified overvoltage factor of 2.1 p.u. being generally accepted. However, the switching impulse overvoltage stress of a VSC-HVDC cable system depends on not only the VSC topology but also the cable parameters and the volt-ampere characteristics of the dc cable surge arresters used. This paper studies the switching impulse overvoltage stresses for the dc cable system of a symmetrical monopole VSC-HVDC transmission system using the modular multilevel converter technology. The studies were performed according to the key technical parameters for a typical 1000-MW/320-kV VSC-HVDC transmission scheme, the XLPE dc cable links being assumed. The results show that a switching impulse overvoltage level of 2.3 p.u. should be applied during the type test of the overvoltage capability of the HVDC cables.

The hybrid multilevel voltage-source converter, combining semiconductor switches and power module cells, is a new family for VSC-HVDC applications. The hybrid multilevel converter in parallel arrangement (PHMC), in terms of converter investment, is very adaptable for cable transmission systems. However, the dc voltage of the converter, synthesized by the output voltage of three-phase cell stacks, is strongly coupled with ac voltage variations, which influences the stability of converter operation. This paper proposes a dc voltage compensation strategy for the parallel hybrid multilevel converter. This strategy, implemented by introducing an isolation period, can effectively cope with the dc voltage deviation, especially under ac network faults. Simulation results in PSCAD/EMTDC verify the effectiveness of the proposed strategy under various circumstances.

The influence of different AlGaN Impurity Blocking Layer (IBL) thickness in the GaN/Si (111) structure with GaN/AlN SLs buffer on the material and electrical properties of GaN/Si (111) system was studied in detail. It is found that the insertion of AlGaN IBL can increase the (102) FWHM and decrease the (002) FWHM. Meanwhile, AlGaN IBL with an optimized thickness can further improve the surface roughness and strain-state of GaN-Si (111) system. By using Secondary Ion Mass Spectroscopy, it is found that AlGaN IBL have a strong effect in blocking the Si donor impurities originating from the Si substrate during the high temperature growth, which can decrease the leakage current while the breakdown voltage can be dramatically increased.

In this paper, Ag@SiO2 core-shell nanoparticles (NPs) with different shell thicknesses were prepared experimentally and introduced into the photosensitive layer of mesoscopic hole-conductor-free perovskite solar cells (PSCs) based on carbon counter electrodes. By combining simulation and experiments, the influences of different shell thickness Ag@SiO2 core-shell nanoparticles on the photoelectric properties of the PSCs were studied. The results show that, when the shell thickness of 0.1 wt% Ag@SiO2 core-shell nanoparticles is 5 nm, power conversion efficiency is improved from 13.13% to 15.25%, achieving a 16% enhancement. Through the measurement of the relevant parameters of the obtained perovskite film, we found that this gain not only comes from the increase in current density that scholars generally think, but also comes from the improvement of the film quality. Like current gain, this gain is related to the different shell thickness of Ag@SiO2 core-shell nanoparticles. Our research provides a new direction for studying the influence mechanism of Ag@SiO2 core-shell nanoparticles in perovskite solar cells.

Stress voltages on time-dependent breakdown characteristics of GaN MIS-HEMTs during negative gate bias stress (with V GS < 0, V D = V S = 0) and off-state stress (V G < V Th, V DS > 0, V S = 0) are investigated. For negative bias stress, the breakdown time distribution (β) decreases with the increasing negative gate voltage, while β is larger for higher drain voltage at off-state stress. Two humps in the time-dependent gate leakage occurred under both breakdown conditions, which can be ascribed to the dielectric breakdown triggered earlier and followed by the GaN layer breakdown. Combining the electric distribution from simulation and long-term monitoring of electric parameter, the peak electric fields under the gate edges at source and drain sides are confirmed as the main formation locations for per-location paths during negative gate voltage stress and off-state stress, respectively.

Model stealing attack may happen by stealing useful data transmitted from embedded end to server end for an artificial intelligent systems. In this paper, we are interested in preventing model stealing of neural network for resource-constrained systems. We propose an Image Encryption based on Class Activation Map (IECAM) to encrypt information before transmitting in embedded end. According to class activation map, IECAM chooses certain key areas of the image to be encrypted with the purpose of reducing the model stealing risk of neural network. With partly encrypted information, IECAM can greatly reduce the time overheads of encryption/decryption in both embedded and server ends, especially for big size images. The experimental results demonstrate that our method can significantly reduce time overheads of encryption/decryption and the risk of model stealing compared with traditional methods.

Phase change materials (PCMs) could significantly contribute to improving the efficiency of thermal energy storage by offering high energy density and constant temperature. This thermal energy storage is crucial for enabling successful implementation of sustainable heating solutions by improving their flexibility and optimizing their performance. This paper provides an overview of most promising PCMs for low-temperature thermal energy storage and discusses their possible integration with sustainable heating and cooling systems for buildings. The use of PCM-based thermal energy storage is still in the research and development phase with a limited number of demonstration projects. However, the first experiences show promising results mostly for improving the efficiency of space heating and cooling by coupling with existing conventional solutions.

High temperature operation(HTO) experiments were performed on industrial GaN HEMTs devices. Several degradation failure modes of DC parameters such as transconductance reduction, threshold voltage shift, and gate leakage current increase were identified. The vertical failure localization and analysis of the devices were carried out by electrical method. The results show that the increase of the degradation of the die attach layer is the main reason for the increase of the junction temperature and thermal resistance of the devices. Furthermore, the results are analyzed by three-dimensional X-ray imaging confirming the degradation of the die attach layer. The voids in the die attach layer expanded during the high temperature operation stress experiments, led to a worse heat dissipation performance, which resulting in increased junction temperature, and further accelerate the performance degradation of the devices.

Deep neural networks can model images with rich latent representations, but they cannot naturally conceptualize structures of object categories in a human-perceptible way. This paper addresses the problem of learning object structures in an image modeling process without supervision. We propose an autoencoding formulation to discover landmarks as explicit structural representations. The encoding module outputs landmark coordinates, whose validity is ensured by constraints that reflect the necessary properties for landmarks. The decoding module takes the landmarks as a part of the learnable input representations in an end-to-end differentiable framework. Our discovered landmarks are semantically meaningful and more predictive of manually annotated landmarks than those discovered by previous methods. The coordinates of our landmarks are also complementary features to pretrained deep-neural-network representations in recognizing visual attributes. In addition, the proposed method naturally creates an unsupervised, perceptible interface to manipulate object shapes and decode images with controllable structures. The project webpage is at http://ytzhang.net/projects/lmdis-rep, Comment: 48 pages

People can percept social attributes from streetscapes such as safety, richness, and happiness by means of visual perception, which inspires the research in terms of urban perception. To the best of our knowledge, this is the first work focused on revealing the relationship between visual patterns of satellite images as well as streetscapes and commercial activeness. We propose to make use of bag of features (BoF) in the context of computer vision and sparse representation in the sense of machine learning to predict commercial activeness of urban commercial districts. After obtaining the urban commercial districts via clustering, we predict the commercial activeness degrees of them using four image features, namely, Histogram of Oriented Gradients (HOG), Autoencoder, GIST, and multifractal spectra for satellite images and street view images, respectively. The performance evaluation with four large-scale datasets demonstrates that the presented computational framework can not only predict the commercial activeness with satisfactory precision compared with that based on Point of Interest (POI) data but also discover the visual patterns related.

Different urban regions usually have different commercial hotness due to the different social contexts inside. As satellite imagery promises high-resolution, low-cost, real-time, and ubiquitous data acquisition, this study aims to solve commercial hotness prediction as well as the correlated social contexts mining problem via visual pattern analysis on satellite images. The goal is to reveal the underlying law correlating visual patterns of satellite images with commercial hotness so as to infer the commercial hotness map of a whole city for government regulation and business planning. We propose a novel deep learning-based model, which learns semantic information from raw satellite images to enable predicting regional commercial hotness. First, we collect satellite images from Google Map and label such images with POI categories according to the annotations from OpenStreetMap. Then, we train a model of deep convolutional networks that leverage raw images to infer the social attributes of the region of interest. Finally, we use three classical regression methods to predict regional commercial hotness from the corresponding social contexts reflected in satellite images in Shanghai, where the applied deep features are learned from the examples of Beijing to guarantee the generality. The result shows that the proposed model is robust enough to reach 82% precision at average. To the best of our knowledge, it is the first work focused on discovering relations between commercial hotness and satellite images. A web service is developed to demonstrate how business planning can be done in reference to the predicted commercial hotness of a given region.

In this paper, the selective area growth (SAG) technique is used to regrow thin AlGaN/GaN heterostructure on access region for realizing trench gate normally-off AlGaN/GaN MOSFET. Heavy background doping is found in SAG AlGaN/GaN heterostructure, which is not expected for its degradation on device performance. The background doping originates from SiO2 residuals at SAG interface. Through reducing the deposition temperature of SiO2 mask, background doping can be efficiently suppressed. As a result, the 2 dimensional electron gas transport property of SAG AlGaN/GaN heterostructure improves greatly, which is as good as the as-grown AlGaN/GaN heterostructure. Moreover, the performance of normally-off SAG AlGaN/GaN MOSFET improves greatly by suppression the Si residual on GaN template.

The influence of AlN/GaN superlattices (SL) buffer on the characteristics of AlGaN/GaN-on-Si (1 1 1) template was studied in detail. There existed an optimized Relative AlN Thickness (RAT) in the superlattices buffer which can not only further filtering the edge- and screw-type dislocations to the upper epilayer and lead to a good crystal quality with narrowest (0 0 0 2) and (1 0 −1 2) full width of half maximum (FWHMs), 439″ and 843″, but also improve the surface roughness to enhance the Two dimensional electron gas (2DEG) mobility and superior electrical properties were achieved. Moreover, an optimized RAT in SL can induce a proper compressive stress to the subsequently grown GaN epilayer and protect it from crack during the cooling step, which can also lead to a better wafer bending.