Forward Branch Research Articles

MSFE-UIENet: A Multi-Scale Feature Extraction Network for Marine Underwater Image Enhancement

Underwater optical images have outstanding advantages for short-range underwater target detection tasks. However, owing to the limitations of special underwater imaging environments, underwater images often have several problems, such as noise interference, blur texture, low contrast, and color distortion. Marine underwater image enhancement addresses degraded underwater image quality caused by light absorption and scattering. This study introduces MSFE-UIENet, a high-performance network designed to improve image feature extraction, resulting in deep-learning-based underwater image enhancement, addressing the limitations of single convolution and upsampling/downsampling techniques. This network is designed to enhance the image quality in underwater settings by employing an encoder–decoder architecture. In response to the underwhelming enhancement performance caused by the conventional networks’ sole downsampling method, this study introduces a pyramid downsampling module that captures more intricate image features through multi-scale downsampling. Additionally, to augment the feature extraction capabilities of the network, an advanced feature extraction module was proposed to capture detailed information from underwater images. Furthermore, to optimize the network’s gradient flow, forward and backward branches were introduced to accelerate its convergence rate and improve stability. Experimental validation using underwater image datasets indicated that the proposed network effectively enhances underwater image quality, effectively preserving image details and noise suppression across various underwater environments.

Analysis of Power Transmission Characteristics of a Novel Secondary Side LCD Series Forward Converter

A novel secondary-side LCD series forward converter is proposed, and the power transmission characteristics of the excitation inductance of the proposed converter under different modes are analyzed. By analyzing the relationship between the transmission power of forward branch and additional branch and excitation inductance, it is concluded that when the excitation inductance Lm works in CCM (continuous conduction mode), the transmission power of forward branch and additional branch has nothing to do with the value of excitation inductance. When the excitation inductance Lm works in DCM (discontinuous conduction mode), the transmission power of forward branch increases with the increase of excitation inductance, while the transmission power of additional branch decreases with the increase of excitation inductance. At the same time, the influence of power distribution on the efficiency of transformation topology is explored, and the relationship between the transmission power of the forward branch and the current stress of the switching tube was derived, and the parameter design method of magnetic element is put forward. Finally, the circuit model of the novel secondary side LCD series Forward converter is established, and the experimental results verify the correctness of the theoretical analysis and the feasibility of the parameter design method.

Learn by Oneself: Exploiting Weight-Sharing Potential in Knowledge Distillation Guided Ensemble Network

Recent CNNs (convolutional neural networks) have become more and more compact. The elegant structure design highly improves the performance of CNNs. With the development of knowledge distillation technique, the performance of CNNs gets further improved. However, existing knowledge distillation guided methods either rely on offline pretrained high-quality large teacher models or online heavy training burden. To solve the above problems, we propose a feature-sharing and weight-sharing based ensemble network (training framework) guided by knowledge distillation (EKD-FWSNet) to make baseline models stronger in terms of representation ability with less training computation and memory cost involved. Specifically, motivated by getting rid of the dependence of offline pretrained teacher model, we design an end-to-end online training scheme to optimize EKD-FWSNet. Motivated by decreasing the online training burden, we only introduce one auxiliary classmate branch to construct multiple forward branches, which will then be integrated as ensemble teacher to guide baseline model. Compared to previous online ensemble training frameworks, EKD-FWSNet can provide diverse output predictions without relying on increasing auxiliary classmate branches. Motivated by maximizing the optimization power of EKD-FWSNet, we exploit the representation potential of weight-sharing blocks and design efficient knowledge distillation mechanism in EKD-FWSNet. Extensive comparison experiments and visualization analysis on benchmark datasets (CIFAR-10/100, tiny-ImageNet, CUB-200 and ImageNet) show that self-learned EKD-FWSNet can boost the performance of baseline models by large margin, which has obvious superiority compared to previous related methods. Extensive analysis also proves the interpretability of EKD-FWSNet. Our code is available at https://github.com/cv516Buaa/EKD-FWSNet.

HiRN: Hierarchical recurrent neural network for video super-resolution (VSR) using two-stage feature evolution

The aim of video super-resolution (VSR) is generate the high-resolution (HR) frames from their low-resolution (LR) counterparts. As one of the fundamental module of VSR, propagation process provides the path of feature map and specifies how the feature map is leveraged. In the recurrent propagation, the latent features can be propagated and aggregated. Therefore, adopting the recurrent strategy can resolve the limitation of sliding-window-based local propagation. Recently, bi-directional recurrent propagation-based latest methods have achieved powerful performance in VSR. However, existing bi-directional frameworks have structured by combining forward and backward branches. These structures cannot propagate and aggregate previous and future latent features of current branch. In this study, we suggest the hierarchical recurrent neural network (HiRN) based on feature evolution. The proposed HiRN is designed based on the hierarchical recurrent propagation and residual block-based backbone with temporal wavelet attention (TWA) module. The hierarchical recurrent propagation consists of two stages to combine advantages of low frame rate-based forward and backward schemes, and multi-frame rate-based bi-directional access structure. The proposed methods are compared with state-of-the-art (SOTA) methods on the benchmark datasets. Experiments show that the proposed scheme achieves superior performance compared with SOTA methods. In particular, the proposed HiRN achieves better performance than all compared methods in terms of SSIM on Vid4 benchmark. In addition, the proposed HiRN surpasses the existing GBR-WNN by a significant 3.03 dB in PSNR on REDS4 benchmark with fewer parameters.

INVESTIGATION OF THE CURRENT-VOLTAGE CHARACTERISTICS OF THE N-CDS/P-CU(IN, GA)SE2 HETEROSTRUCTURE

In this work, current transfer mechanism in the n-CdS/p-Cu(In, Ga)Se2 heterostructure in the forward and reverse current directions at T=300 K is investigated. It has been established that the dark current-voltage characteristics of heterostructures built on a double logarithmic scale are described by power-law dependences type of the I=Vα. In the forward direction of the current, sections of the current-voltage characteristic were observed: α1=1 (ohmic) and α2=2 (quadratic), and for the reverse section of the current-voltage characteristic: α1=1 (ohmic), α2=0.25. From the quadratic section of the forward branch, the value µn∙τn=4.5∙10-10 cm2/V for the CIGS active layer was determined, which is explained by the processes of recombination of charge carriers through simple local centers.

B2GAN: Bidirectional-branch generative adversarial network for text image super-resolution with structure preservation

Text image super-resolution (SR) is one of the most hot topic research in computer vision. The traditional convolutional neural networks usually over-smooth fine structures of SR images while the popular generative adversarial network (GAN) based SR methods may produce undesirable SR results with distorted edges. To cope with these problems, this paper proposes a bidirectional-branch generative adversarial network (B2GAN) for text image SR. The proposed B2GAN consists of two parts: one is a generator, the other is a discriminator. In the generator, a forward branch is used for mapping the low-resolution (LR) image to its corresponding SR image to generate photo-realistic images, while a reverse branch is responsible for mapping the generated SR image to a low-resolution image which is viewed as its original LR image. A loss function is proposed to train a pleasing generator and guarantee a stable convergence to the original LR image. In the discriminator, The VGG16 and the adversarial loss are employed. In the experiments, we establish two text image datasets including Chinese and multi-language characters which are used to test the effectiveness of the proposed B2GAN method. Quantitative associated with qualitative experimental results validate that the proposed B2GAN method can achieve competitive performance and even outperform the state-of-the-arts.

Method for Determining the Schottky Diodes Electrical Parameters

When studying the operation of Schottky diodes the most important electrical parameters are the height of the potential barrier, the coefficient of ideality, the saturation current and the series resistance of the material and contacts. These parameters can be determined from the experimental volt-ampere characteristics. The article considers the methods of determining these electrical parameters of Schottky diodes, as well as the factors that affect the accuracy of calculations. The existing methods for calculating the electrical parameters of Schottky diodes are analyzed, namely: the method of Norde, Roderick, Chong, Sato and the method of direct approximation. The Norde method was developed for a coefficient of ideality equal to one for cases where the effect of series resistance on the I–V characteristics makes a significant error in determining the barrier height by simpler methods. A significant disadvantage of this method is that in many cases the coefficient of ideality is not equal to one, even in the case of an ideal diode, which makes an error in the calculation result. The advantage of Roderick's method is the possibility of describing the forward and reverse branches of the I–V characteristics by one dependence, as well as taking into account measurements at voltages less than tripled temperature potential. The disadvantages of this method include the lack of consideration of the effect of series resistance, which may result in additional errors. The main advantage of the Chong method is the determination of the series resistance together with the height of the barrier and the coefficient of ideality, which not only provides additional information about the contact, but also convenient in terms of automation of the calculation process. The disadvantages include the possibility of applying the method only to the voltage range above the tripled temperature potential. The disadvantages of Sato methods and direct approximation include the fact that the calculation is performed at one point of the I–V curve, which can negatively affect the accuracy. It is also shown that these methods have a significant standard deviation of the calculated values from the experimental ones, which is due to the temperature dependence of the height of the potential barrier and the dependence of the coefficient of ideality on the voltage. Also, the reason for the increase in the calculation error of the electrical parameters in all five methods is the decrease in the length of the I–V characteristics in logarithmic coordinates. When using any of the considered methods, the calculation is performed in logarithmic coordinates, which complicates the determination of the boundaries of the I–V section, where the dependence of the parameters of the Schottky diode on the voltage is insignificant. A new algorithm for calculating the electrical parameters of Schottky diodes has been developed. Based on the conjugate gradient method, a method for optimizing the algorithm for calculating the electrical parameters of Schottky diodes was developed, which made it possible to reduce the standard deviation by more than an order of magnitude. The developed algorithm is verified by comparing the calculated volt-ampere characteristics of Schottky diodes with those obtained experimentally. To construct the calculated volt-ampere characteristics, the values of the electrical parameters of Schottky diodes were used, which were determined by the presented algorithm. The results of the calculation are in good agreement with the experimental data. The proposed method can be used both in scientific work to study the properties of semiconductor materials, and in production to control the quality of Schottky diodes.

Electrical and Photoelectric Properties of Organic-Inorganic Heterojunctions PEDOT:PSS/n-CdTe

PEDOT: PSS thin films are widely used as transparent coatings in flexible semiconductor devices including solar cells. However, they are not widely used as transparent coatings in combination with crystal substrates. This work shows the possibility of using PEDOT:PSS thin films as a frontal transparent conducting layer in hybrid organic-inorganic Schottky type heterojunctions of the PEDOT:PSS/n‑CdTe, which were prepared by deposition of PEDOT:PSS thin films (using the spin-coating method) on crystalline cadmium telluride substrates. The current-voltage (in a wide temperature range) and capacitance-voltage (at room temperature) characteristics of heterojunctions were measurement and analyzed. It has been established that PEDOT:PSS/n-CdTe heterojunctions have good diode properties with a high rectification ratio RR≈105, a potential barrier height φ0 = 0.95 eV, and series Rs = 91 Ohm and shunt Rsh = 5.7 × 107 Ohm resistances. Analysis of the forward branches of the I–V characteristics of heterojunctions showed that the dominant charge transfer mechanisms are determined by the processes of radiative recombination at low biases (3kT/e <V <0.3 V) and tunneling through a thin depleted layer at high biases (0.3 V <V <0.6 V). Capacity-voltage characteristics are plotted in the Mott-Schottky coordinate, taking into account the influence of series resistance, measured at a frequency of 1 MHz. Used the C-V characteristic was determined the value of the built-in potential Vc = 1.32 V (it correlates well with the cutoff voltage determined from the current-voltage characteristics) and the concentration of uncompensated donors in the n-CdTe substrate ND-NA = 8.79 × 1014 cm-3. Although the photoelectric parameters of unoptimized PEDOT:PSS/n-CdTe heterojunctions are low, their photodiode characteristics (Detectivity D*> 1013 Jones) are very promising for further detailed analysis and improvement. The proposed concept of a hybrid organic-inorganic heterojunction also has potential for use in inexpensive γ- and X-ray detectors.

The Case Study of the Big Latvian Energy Companies’ Compliance with the Principles of Sustainable Development

The rationale of the study is connected with the fact that society has increasingly begun to discuss its responsibility to the generation to come. Without economic aspects, social and environmental issues have become increasingly apparent because there is more and more relevant question whether the natural system will be able to withstand the increasing loading associated with human activities. In Latvia there have been made attempts to promote the principle of sustainable development, its adoption and implementation. On the basis of the above stated rationale of the subject, the objective of the study is to estimate the big Latvian energy companies’ compliance with the principles of sustainable development. The reduction of energy sources and their capacity, greenhouse gas emission caused by energy production, and features of the energy market put forward branch enterprises in front of the principle of sustainable development and the idea of its implementation. To determine the big Latvian energy companies’ contribution to sustainable development, in the survey of representatives of the business environment there have been included questions about the big Latvian energy companies’ compliance with the principles of sustainable development. In the research paper there have been used the statistical analysis of data and the processing method.

Computer-Aided Diagnostic System for Digital Mammography

In this work, Computer-Aided Detection (CADe) and Computer-Aided Diagnosis (CADx) systems are developed and tested using the public and freely available mammographic databases named MIAS and DDSM databases, respectively. CADe system is used to differentiate between normal and abnormal tissues, and it assists radiologists to avoid missing a breast abnormality. At the same time, CADx is developed to distinguish between normal, benign and malignant breast tissues, and it helps radiologists to decide whether a biopsy is needed when reading a diagnostic mammogram or not. Any CAD system is constituted of typical stages including preprocessing and segmentation of mammogram images, extraction of regions of interest (ROI), features removal, features selection and classification. In both proposed CAD systems, ROIs are selected using a window size of 32×32 pixels, then a total of 543 features from four different feature categories are extracted from each ROI and then normalized. After that, the selection of the most relevant features is performed using four different selection methods from MATLAB Pattern Recognition Toolbox v.5 (PRtool5) named Sequential Backward Selection (SBS), Sequential Forward Selection (SFS), Sequential Floating Forward Selection (SFFS) and Branch and Bound Selection (BBS) methods. We also utilized Principal Component Analysis (PCA) as the fifth method to reduce the dimensions of the features set. After that, we used different classifiers such as Support Vector Machines (SVM), K-voting Nearest Neighbor (K-NN), Quadratic Discriminant Analysis (QDA) and Artificial Neural Networks (ANN) for the classification. Both CAD systems have the same implementation stages but different output. CADe systems are designed to detect breast abnormalities while CADx system indicates the likelihood of malignancy of lesions. Finally, we independently compared the performance of all classifiers with each selection method in both modes. The evaluation of the proposed CAD systems is done using performance indices such as sensitivity, specificity, the area under the curve (AUC) of the Receiver Operating Characteristic (ROC) curves, the overall accuracy and Cohen-k factor. Both CAD systems provided encouraging results. These results were different corresponding to the selection method and classifier.

Reverse Unfolding of Petri Nets and its Application in Program Data Race Detection

The unfolding technique can partially alleviate the state explosion in Petri nets through branching processes. However, all states of a system are still contained in its unfolding net. To deal with some practical problems, only the coverability determination of a specific state is needed. In view of this, reducing the scale of the unfolding net is feasible. This study proposes a target-oriented reverse unfolding algorithm for the coverability determination of 1-safe Petri nets, which combines a heuristic technique to reduce the scale of unfolding nets, thereby improving the efficiency of coverability determination. Furthermore, the reverse unfolding is applied to the formal verification of concurrent programs, and their data race detection is converted into the coverability determination of a specific state in 1-safe Petri nets. The experiment compares the efficiency between forward nfolding and reverse unfolding in the coverability determination of a Petri net. The results show that when the Petri net has more forward branches than backward branches, reverse unfolding is more efficient than forward unfolding. Finally, the key factors influencing the efficiency of reverse unfolding are analyzed.

String Perturbation Theory on the Schwinger-Keldysh Time Contour.

We perform the large-N expansion in the Schwinger-Keldysh formulation of nonequilibrium quantum systems with matrix degrees of freedom and study universal features of the anticipated dual string theory. We find a rich refinement of the topological genus expansion: In the original formulation, the future time instant where the forward and backward branches of the Schwinger-Keldysh time contour meet is associated with its own world sheet genus expansion. After the Keldysh rotation, the world sheets decompose into a classical and quantum part.

The Research of Piezo Phototron Effect in A2B6 Film Structures

The paper presents the results of the study of a piezo phototron effect in hetero structures based on A2B6 films. The field of study of the piezo phototron effect in semiconductor thin films is new. The use of materials that have several properties at the same time, namely semiconductors, piezoelectrics and photoelectric properties, allows us to create a fundamentally new class of semiconductor devices, especially in the field of sensors and photovoltaics. Obtaining films of the A2B6 group, which include CdS, CdTe, CdSe, ZnS, ZnSe, ZnO with an existing piezoelectric effect, is an urgent task. In this work, the object of study is diode structures based on them. Namely, CdS films were grown by thermal evaporation on a flexible molybdenum substrate. A thin layer of degenerate CuxS semiconductor was deposited on top of the CdS films, which allowed it to form a diode barrier structure. For the study of photodiode properties, measurements were made of the current-voltage characteristics (VAC) in the dark and light modes. Measurements were made using a power source, a voltmeter and an ammeter. The limits of voltage measurements ranged from -2 V to 2 V to prevent possible breakdown of structures. A white LED with a power output of 30mW / cm2 was used as the light source. From the obtained CVCs it was obtained that a rectifying barrier is formed between the semiconductor films and the structure has a diode characteristic. The magnitude of the reverse current increases during illumination. Pressure was applied to the diode structures to investigate the piezoelectric effect. The pressure applied to the sample was changed by means of levers weighing 10, 20, 50, 100 and 200 grams. As a result of the measurements of the CVC during the application of pressure, it was found that the current on the forward and reverse branches of the CVC showed an increase in both the dark mode of measurement and in the light. This indicates the presence of a piezoelectric effect in this heterojunction. The calculated values of the sensitivity to pressure showed that the current on the back and forward branches of the CVC changes at different intervals. The highest sensitivity in the range 0–77.31 absolute units is observed for reverse current without illumination. With additional illumination, the sensitivity to the pressure on the back branch of the CVC decreases and is 0–7.73. All measurements made show the presence of a piezophototron effect, which allows the creation of new functional devices for measuring pressure and luminance based on this heterojunction.

Superconducting transition in resistivity and magnetization of the solid solution (Pb0.4Sn0.6)0.8In0.2Te

The temperature and magnetic field dependences of electrical resistivity ρ(Т, Н) and magnetization m(Т, H) of the semiconductor solid solution (Pb0.4Sn0.6)0.8In0.2Te were studied in the temperature range T = 2 K - 300 K (including the region of the superconducting transition of the compound T ≤ 4 K) and in magnetic fields H ≤ 140 kOe. The critical parameters of superconducting transition were determined: the critical temperature Tc ≈ 4 K and the second critical magnetic field at T → 0 K, Hc2(0) ≈ 27.5 kOe. The magnetic field - magnetization dependences m(H) at T < Tc region exhibit a hysteresis loop associated with the magnetic flux capturing by a polycrystalline superconducting sample. In magnetic fields close to the second critical Hc2(T) at temperatures T < 3 K, an additional maximum of the diamagnetic response on the forward branch and a symmetric paramagnetic maximum on the reverse branch were observed on the m(H) dependences, which we interpret as the peak-effect in magnetization of the superconducting solid solution. In the (Pb0.4Sn0.6)0.8In0.2Te sample, at the temperature Т = 2 K and the peak-effect field Нpeak = 16.5 kOe, the magnetization peak amplitude was ~ 6% of the maximum value of the Meissner response at H ≈ 0 Oe. On the basis of the obtained data the possible nature of the peak-effect in the magnetization of solid solutions (PbzSn1-z)1-хInхTe, at Т < Tc was discussed, influence of the vortex structure dynamics of the superconducting compound, the lead content in the solid solution, the level of In doping and external parameters was considered.

The universal algorithm for calculating SPICE models’ parameters of semiconductor diodes

The methods and procedures for calculating SPICE models’ parameters of semiconductor diodes for the forward and reverse branches of the VAC are described. Recommendations for planning the experiment and selecting initial approximations of parameters for forward and reverse switching of the diode are given. Existing methods for calculating SPICE models’ parameters do not allow obtaining their statistical characteristics, and for some parameters, methods for estimating them from experimental data have not been developed at all. This article offers a universal method for evaluating all the necessary parameters of SPICE models of semiconductor diodes and their statistical characteristics.

Features of Glow Discharge burning between a hollow cathode and a mesh anode

The article is devoted to an experimental study of the features of the combustion of a glow discharge between a hollow copper cathode and a mesh anode in a longitudinal air flow. The cylindrical channel of the discharge chamber was formed by five glass sections separated by copper plates, which served as probes for measuring the plasma potential. Air flowed through the hollow cathode into a cylindrical channel. The anode was a metal grid of a copper plate 1.5 mm thick with holes uniformly located on it with a diameter of 1 mm. The distance between the hollow cathode and the mesh anode was 6.5 cm. Air pressure P, its velocity V, discharge current I and its voltage U varied respectively in the ranges: P = (6. 4 - 19) kPa, V = (0 - 10) m / s, I = (5-100) mA, U = (1. 0 – 2. 0) kV. The current-voltage characteristic had a negative differential resistance. By changing the ballast resistance, the forward and reverse branches of the current-voltage characteristic were removed. It was found that in the range of discharge currents I = (40-80) mA, these branches do not coincide, that is, hysteresis phenomena were observed. In the ascending branch of the current-voltage characteristic in the region of the current strength I = 80 mA, the discharge voltage increases sharply from U = 1. 2 kV to U = 1.6 kV. In this case, the discharge glow becomes more intense, wider and brighter, and the structure of light and dark zones along the discharge chamber also changes. Characteristic photos of the discharge are given. The formation of luminous and dark gaps inside the positive column of the discharge and their movement towards the anode was detected.

Absence of percolation for Poisson outdegree-one graphs

Nous considérons des graphes orientés dont l’ensemble des sommets est donné par un processus ponctuel de Poisson et tels que chaque sommet admette une et une seule arête sortante. Le résultat principal de ce papier est l’absence de percolation pour de tels graphes satisfaisant deux hypothèses. L’hypothèse Shield stipule que l’état du graphe localement ne dépend que de son voisinage, tandis que l’hypothèse Loop prétend que toute branche orientée du graphe échoue sur un cycle dès qu’un ensemble (fini) de sommets bien choisis est ajouté au processus de Poisson. Plusieurs modèles intéressants satisfont ces deux hypothèses générales et, par conséquent, ne percolent pas. Nous résolvons ainsi dans Theorem 3.1 une conjecture de Daley et al. sur l’absence de percolation pour le “line-segment model” (Conjecture 7.1 de (Probab. Math. Statist. 36 (2016) 221–246), également discutée dans (Ann. Inst. Henri Poincaré Probab. Stat. 52 (2016) 127–145)). Dans ce modèle bidimensionnel, un segment pousse depuis chaque point du processus de Poisson jusqu’à ce qu’il heurte un autre segment, stoppant ainsi sa croissance. Les directions dans lesquelles poussent les segments sont choisies uniformément sur la sphère et indépendamment les unes des autres. Enfin, un autre modèle dit de navigation est présenté et satisfait aussi les hypothèses Shield et Loop.

Depth-dependent ferromagnetic spin structure and asymmetric magnetization reversal in exchange-biased Fe/FeMn bilayers

Depth-dependent spin rotation of the Fe spins upon magnetisation reversal of the ferromagnetic layer in polycrystalline Fe/FeMn exchange-biased bilayers was investigated by using 57Fe probe layers and angular hysteresis loops constructed from in-field 57Fe conversion electron Mössbauer spectroscopy (CEMS) measurements and magnetometry. Our results provide evidence for the formation of an exchange-spring-like domain wall structure inside the ferromagnetic Fe layer during magnetisation reversal along the forward branch of the hysteresis loop (H values varying from positive to negative), and coherent reversal along the recoil branch (H varying from negative to positive). This result implies asymmetric magnetisation reversal for each branch of the hysteresis loop, as indicated by depth-resolved CEMS and vibrating sample magnetometry. These results are relevant for the development of more precise quantitative models for exchange bias in polycrystalline metallic systems with Mn-based antiferromagnets such as those used in spintronic applications.

Experimental study of transient processes in a Glow Discharge

The article experimentally investigated the features of a glow discharge with a hollow cathode in a longitudinal air flow. The discharge chamber consisted of glass sections separated by copper plates serving as probes. The channel formed by such sections and probes had a cylindrical shape. Air flowed through the hollow cathode into the cylindrical channel. The anode served as a metal grid. Probes and electrodes were made of copper. The distance between the electrodes was 6.5 cm. The air pressure P, its speed V, the discharge current I and its voltage U changed respectively in the ranges: P = (6, 4 - 19) kPa, V = (0 - 10) m / c, I = (5-100) mA, U = (1, 0 - 2, 2, 0) kV. Volt-ampere characteristic had negative differential resistance. It was found that in the range of discharge currents I = (40-80) mA, the forward and reverse branches of the current-voltage characteristics do not match. When the current decreases (reverse branch), the discharge voltage rises sharply. This is due to the process of transition of a glow discharge into its non-stationary form, which was observed visually and was recorded photographically for all the characteristic modes of discharge. It was also shown that the flow of air allows a reverse transition to the usual form of discharge.

Power Loss Comparison in a BOOST PFC Circuit Considering the Reverse Recovery of the Forward Diode

The nature of diode reverse recovery is analyzed in this paper, and the reverse recovery loss is evaluated in a BOOST PFC converter using a silicon (Si) or silicon carbide (SiC) diode in the forward branch. Mathematical models of the forward conduction and reverse recovery losses are established to assess the influence of Si and SiC diodes. To characterize and quantify the losses related to diode reverse recovery, an 85~265V AC to 400V DC, 2kW BOOST PFC prototype is built with switching frequencies of 65kHz. It is found that the reverse-recovery inherent in a Si diode cannot be neglected. The switching loss is substantially smaller when the diode is a SiC one. In order to investigate further, a double pulse test rig is established, with the switch and the diode being either Si or SiC. The experimental results demonstrate that with a SiC diode, not only the diode conduction losses but also the transistor turn-on loss is greatly reduced.