Showing papers on "Voltage published in 2023"

Development of a Bidirectional DC/DC Converter with Dual-Battery Energy Storage for a Hybrid Electric Vehicle System

Abstract: Abstract: This study develops a newly designed, patented, bidirectional DC/DC converter (BDC) that interfaces a main energy storage (ES1), an auxiliary energy storage (ES2), and a DC-bus of different voltage levels, for application in hybrid electric vehicle systems. The proposed converter can operate in a step-up mode (i.e., low-voltage dual-source-powering mode) and a stepdown (i.e., high-voltage dc-link energy-regenerating mode), both with bidirectional power flow control. In addition, the model can independently control power flow between any two low-voltage sources (i.e., low-voltage dual-source buck/boost mode). Herein, the circuit configuration, operation, steady-state analysis, and closed-loop control of the proposed BDC are discussed according to its three modes of power transfer.

Convolutions are competitive with transformers for protein sequence pretraining

Abstract: Pretrained protein sequence language models have been shown to improve the performance of many functional and structural prediction tasks, and are now routinely integrated into bioinformatics tools. However, these models largely rely on the Transformer architecture, which scales quadratically with sequence length in both run-time and memory. As a result, even state-of-the-art models have limitations on maximum sequence length, limiting what protein sequences can be analyzed with protein language models. To address this limitation, we investigated if more efficient convolutional neural network (CNN) architectures, which scale linearly with sequence length, could be equally as effective as pretrained protein language models. We show that with the masked language model pretraining task, CNNs are competitive to and occasionally superior to Transformers across an extensive set of downstream applications, including structure prediction, zero-shot mutation effect prediction, and out-of-domain generalization. We also demonstrate strong performance on sequences longer than those allowed in the current state-of-the-art Transformer models. Our work has important implications for applications built on protein language models, suggesting that computational efficiency can be improved without sacrificing performance simply by using a CNN architecture instead of a Transformer, and emphasizes the importance of disentangling pretraining task and model architecture in future work studying these models.

Vision Permutator: A Permutable MLP-Like Architecture for Visual Recognition

Abstract: In this paper, we present Vision Permutator, a conceptually simple and data efficient MLP-like architecture for visual recognition. By realizing the importance of the positional information carried by 2D feature representations, unlike recent MLP-like models that encode the spatial information along the flattened spatial dimensions, Vision Permutator separately encodes the feature representations along the height and width dimensions with linear projections. This allows Vision Permutator to capture long-range dependencies and meanwhile avoid the attention building process in transformers. The outputs are then aggregated in a mutually complementing manner to form expressive representations. We show that our Vision Permutators are formidable competitors to convolutional neural networks (CNNs) and vision transformers. Without the dependence on spatial convolutions or attention mechanisms, Vision Permutator achieves 81.5% top-1 accuracy on ImageNet without extra large-scale training data (e.g., ImageNet-22k) using only 25M learnable parameters, which is much better than most CNNs and vision transformers under the same model size constraint. When scaling up to 88M, it attains 83.2% top-1 accuracy, greatly improving the performance of recent state-of-the-art MLP-like networks for visual recognition. We hope this work could encourage research on rethinking the way of encoding spatial information and facilitate the development of MLP-like models. PyTorch/MindSpore/Jittor code is available at https://github.com/Andrew-Qibin/VisionPermutator .

Nonisolated Topology for High Step-Up DC–DC Converters

Abstract: A new dc–dc converter structure with high voltage gain is proposed in this paper. The proposed converter consists of two switches that are turned on and off simultaneously. In addition, the two switched-capacitor cells and one energy storage cell are utilized in the structure of the proposed converter. These result in high voltage gains at low values of the duty cycle. Besides, the voltage stress across the power devices is low and below half of the output voltage. Therefore, the MOSFET switches with low $R_{\mathrm {DS-on}}$ and devices with reduced nominal voltage can be used in the proposed converter which in turn reduces the conduction and turn-on losses. The operating principles of the proposed topology are explained, and the analysis of voltage and current stresses of the devices is accomplished. The circuit performance is compared with other solutions in the literature in terms of voltage gain and normalized voltage stress of switches and diodes. Eventually, in order to verify the theoretical analysis, the experimental results are provided.

Fault diagnosis in a five-level multilevel inverter using an artificial neural network approach

Abstract: Introduction. Cascaded H-bridge multilevel inverters (CHB-MLI) are becoming increasingly used in applications such as distribution systems, electrical traction systems, high voltage direct conversion systems, and many others. Despite the fact that multilevel inverters contain a large number of control switches, detecting a malfunction takes a significant amount of time. In the fault switch configurations diode included for freewheeling operation during open-fault condition. During short circuit fault conditions are carried out by the fuse, which can reveal the freewheeling current direction. The fault category can be identified independently and also failure of power switches harmed by the functioning and reliability of CHB-MLI. This paper investigates the effects and performance of open and short switching faults of multilevel inverters. Output voltage characteristics of 5 level MLI are frequently determined from distinctive switch faults with modulation index value of 0.85 is used during simulation analysis. In the simulation experiment for the modulation index value of 0.85, one second open and short circuit faults are created for the place of faulty switch. Fault is identified automatically by means of artificial neural network (ANN) technique using sinusoidal pulse width modulation based on distorted total harmonic distortion (THD) and managed by its own. The novelty of the proposed work consists of a fast Fourier transform (FFT) and ANN to identify faulty switch. Purpose. The proposed architecture is to identify faulty switch during open and short failures, which has to be reduced THD and make the system in reliable operation. Methods. The proposed topology is to be design and evaluate using MATLAB/Simulink platform. Results. Using the FFT and ANN approaches, the normal and faulty conditions of the MLI are explored, and the faulty switch is detected based on voltage changing patterns in the output. Practical value. The proposed topology has been very supportive for implementing non-conventional energy sources based multilevel inverter, which is connected to large demand in grid.

A Multipurpose and Power Quality Improved Electric Vessels Charging Station for the Seaports

Abstract: To avoid air and noise pollution on seaports, the use of electric vessels and shore-to-ship power supply based charging station on seaports are in trend, which is also known as cold ironing (CI) infrastructure. In the starting stage, electric vessels charging stations are being implemented on the most busiest commercial ports, where modern electronic devices are used to serve different types of vessels. These devices only take care of the quality of output power. However, on the input side, it creates a serious harmonic issue. During CI, megawatt range of power is supplied to the ships, so generated harmonics and interharmonics in the high-power supply based charging system become more dangerous for the grid. In this article, we propose a novel power quality improved cold ironing (PQICI) system (charging mechanism) to overcome this issue. Moreover, due to inherent multipurpose useability, this novel PQICI is able to supply different types of power to various types of vessels, such as ac power, dc power, 50 Hz frequency, 60 Hz frequency, single phase, three phases, three-wire network, four-wire network, five-wire network, and different ac and dc voltages. Furthermore, this PQICI supply system is capable of serving four vessels at a time. In every type of loading condition with different kinds of power supply mode, the PQICI maintains the power quality on the input side, with fulfilling the varieties of the power demand of vessels. The capability of the developed “PQICI” system is tested for different types of ships and vessels on the simulation platform as well as through real-time hardware-in-the-loop experimentation. On IEEE standard 519, satisfactory test results on both platforms show the efficiency of the developed system and its control mechanism.

Learning Adversarial Transformer for Symbolic Music Generation

Abstract: Symbolic music generation is still an unsettled problem facing several challenges. The complete music score is a quite long note sequence, which consists of multiple tracks with recurring elements and their variants at various levels. The transformer model, benefiting from its self-attention has shown advantages in modeling long sequences. There have been some attempts at applying the transformer-based model to music generation. However, previous works train the model using the same strategy as the text generation task, despite the obvious differences between the pattern of texts and musics. These models cannot consistently produce music samples of high quality. In this article, we propose a novel adversarial transformer to generate transformer to generate music pieces with high musicality. The generative adversarial learning and the self-attention networks are combined creatively. The generation of long sequence is guided by the adversarial objectives, which provides a strong regularization to enforce the transformer to focus on learning of the global and local structures. Instead of adopting the time-consuming Monte Carlo (MC) search method that is commonly used in the existing sequence generative models, we propose an effective and convenient method to compute the reward for each generated step (REGS) for the long sequence. The discriminator is trained to optimize the elaborately designed global and local loss objective functions simultaneously, which enables the discriminator to give reliable REGS for the generator. The adversarial objective combined with the teacher forcing objective is used to guide the training of the generator. The proposed model can be used to generate single-track or multitrack music pieces. Experiments show that our model can generate long music pieces with the improved quality compared with the original music transformers.

Photovoltaic-thermal solar-assisted heat pump systems for building applications: Integration and design methods

Abstract: • PVT-SAHP is a suitable solution to meet nZEB standards in the residential sector. • PVT-SAHP may cover a high fraction of building thermal needs, exploiting RES. • Multi-source configurations have high performances, robustness and flexibility. • Advancement in technology and manufacturing led to a new generation of PVT collectors. The photovoltaic-thermal collector is one of the most interesting technology for solar energy conversion, combining electric and thermal energy production in a single device. Vapour-compression heat pump is already considered the most suitable clean technology for buildings thermal energy needs. The combination of these two technologies in an integrated ”photovoltaic-thermal solar-assisted heat pump” (PVT-SAHP) system allows reaching a high fraction of the building thermal needs covered by renewable energy sources and to improve the performances of both the photovoltaic-thermal collector and the heat pump. The first is cooled down increasing its energy conversion efficiency, while providing low-temperature thermal energy to the second, which benefits from a higher evaporation temperature. The review study presents the state-of-art of photovoltaic-thermal solar-assisted heat pump systems intended to cover thermal energy needs in buildings, with a particular focus on the integration methodologies, the possible configurations, the use of different sources and the design of sub-system components. These issues are addressed by much scientific research, to improve the reliability and applicability of this technology, as an option for the building decarbonization. This study aims to present PVT-SAHP systems in an organic and critical way to propose a useful tool for future research developments. More in detail, the work highlights the fact that the integration of photovoltaic-thermal collectors as evaporator of the heat pump in direct-expansion systems allows the highest heat recovery and performances. However, the distinction of the two circuits lead to more reliable, flexible and robust systems, especially when combined with a second heat source, being able to cover both heating and cooling needs. The implementation of real-time control strategy, as well as the continuous development of the compressor and refrigerant industries is positively influencing this technology, which is receiving more and more attention from scientific research as a suitable solution for nearly zero energy buildings.

Voltage Sensorless Based Model Predictive Control With Battery Management System: For Solar PV Powered On-Board EV Charging

Abstract: This work deals with a novel voltage sensorless based model predictive control (VSPC) scheme for continuous and quick maximum power harvesting (MPH) from a photovoltaic (PV) array for a solar-powered on-board electric vehicle (EV) charging system. In VSPC, the first model predictive control (MPC) is used with a PV array to predict the system state in the horizon of time and to eliminate the voltage sensor. An adaptive concept is used for deciding the operating point, which accelerates the tracking process and improves dynamic performance during irradiation changes and shading pattern changes in partially shaded conditions. Moreover, VSPC also takes care of the EV charging process using the EV -provided battery management system (BMS) command or threshold safety limits of the EV battery. The working principle of VSPC is based on a prediction of the future behavior of the system. It realizes on a selected time horizon, in an arbitrary number of samples, which is decided according to the complexity of the fitness function. In order to minimize or maximize the fitness function, it predicts the voltage of the solar PV array as well as tunes the present control signal, which forces it to converge or reach the convergence criteria. Moreover, the cost and response time of current sensors are lower than voltage sensors. Therefore, the VSPC control gives a fast response and low power oscillations in steady-state compared to conventional techniques. This control technique is verified on a developed prototype of the PV system in different shading and irradiance conditions, as well as the system stability, is analyzed through the Bode plot. The system performance is also compared with the state-of-the-art methods.

Sequential Subspace Optimization Design of a Dual Three-Phase Permanent Magnet Synchronous Hub Motor Based on NSGA III

Abstract: The multiobjective optimization design of dual three-phase permanent magnet synchronous hub motors (PMSHMs) is challenging due to the high dimension and huge computation cost of finite-element analysis (FEA). A new multiobjective optimization strategy is proposed for dual three-phase PMSHMs in this article. All design parameters are divided into two subspaces according to the Pearson sensitivity analysis results to improve optimization efficiency. A new training method is adopted to improve the accuracy of the approximate model. By improving a multiobjective intelligent optimization algorithm, nondominated sorting genetic algorithm (NSGA) III, a new algorithm is proposed, which will greatly shorten optimization time. It is found that the proposed optimization method can significantly improve the performance, such as smaller torque ripple and higher maximum torque for the investigated PMSHM, while the computation resources are reduced. A prototype based on the optimization results is manufactured, and experiments are conducted on the platform to verify the accuracy of the optimization results and the FEA. The effectiveness of optimization and the accuracy of the simulation are verified by the experimental results.

A Multilevel Inverter With Minimized Components Featuring Self-Balancing and Boosting Capabilities for PV Applications

Abstract: The cascaded H-bridge multilevel inverter (CMI) attracts much attention as a versatile converter in photovoltaic (PV) applications. Requiring several isolated dc sources and many switches are the main demerits of the CMI. In PV applications with the CMI, PV modules can be used as the isolated dc sources, which, however, may contribute to intermodule and grid leakage currents due to the module stray capacitors. In this context, a switched-capacitor (SC)-based cascaded half-bridge multilevel inverter is proposed in this paper to address the above issues. The proposed topology only requires one dc source, and it achieves the minimum number of switches, spontaneous capacitor charging, voltage boosting, and continuous input current. The intermodule leakage currents can also be eliminated in the proposed topology. The feasibility and effectiveness of the proposed topology are validated through simulations and experimental tests.

GPT-4 Technical Report

Abstract: We report the development of GPT-4, a large-scale, multimodal model which can accept image and text inputs and produce text outputs. While less capable than humans in many real-world scenarios, GPT-4 exhibits human-level performance on various professional and academic benchmarks, including passing a simulated bar exam with a score around the top 10% of test takers. GPT-4 is a Transformer-based model pre-trained to predict the next token in a document. The post-training alignment process results in improved performance on measures of factuality and adherence to desired behavior. A core component of this project was developing infrastructure and optimization methods that behave predictably across a wide range of scales. This allowed us to accurately predict some aspects of GPT-4's performance based on models trained with no more than 1/1,000th the compute of GPT-4.

Transformer-resonant circuit for battery charging

Abstract: The article discusses the issues of charging storage batteries using a transformer-resonant circuit, where a capacitor connected in parallel between the phases of a three-phase power supply is used to compensate for the negative section of the characteristics of a three-element oscillatory circuit. The batteries are charged from a multi-output three- phase matching transformer, which makes it possible to increase the charging power. The charging current stabilization mode is provided with compensation of the negative section of the resonant circuit characteristic.

Vision Transformers for Single Image Dehazing

Abstract: Image dehazing is a representative low-level vision task that estimates latent haze-free images from hazy images. In recent years, convolutional neural network-based methods have dominated image dehazing. However, vision Transformers, which has recently made a breakthrough in high-level vision tasks, has not brought new dimensions to image dehazing. We start with the popular Swin Transformer and find that several of its key designs are unsuitable for image dehazing. To this end, we propose DehazeFormer, which consists of various improvements, such as the modified normalization layer, activation function, and spatial information aggregation scheme. We train multiple variants of DehazeFormer on various datasets to demonstrate its effectiveness. Specifically, on the most frequently used SOTS indoor set, our small model outperforms FFA-Net with only 25% #Param and 5% computational cost. To the best of our knowledge, our large model is the first method with the PSNR over 40 dB on the SOTS indoor set, dramatically outperforming the previous state-of-the-art methods. We also collect a large-scale realistic remote sensing dehazing dataset for evaluating the method's capability to remove highly non-homogeneous haze. We share our code and dataset at https://github.com/IDKiro/DehazeFormer.

Dual Boost Five-Level Switched-Capacitor Inverter With Common Ground

Abstract: This brief presents a five-level common ground type (5L-CGT) inverter topology with double voltage boosting. This topology uses only eight switches, two capacitors that can be charged equal to the input source. The advantageous feature of this topology is the CGT configuration between the input source and output load, which ensures minimal common-mode voltage and leakage current. A detailed analysis is done to select and size capacitors, along with quasi-soft charging to minimize the charging current ripples across the capacitors. The performance of the proposed 5L-CGT inverter is tested for inductive load, varying the modulation index/load/input voltage under experimental test conditions. A detailed comparison is done with the existing state-of-art that shows the merits of the proposed topology in terms of equal voltage stress across the capacitor with voltage boost. The power loss analysis for the inverter is carried out using PLECS with efficiency to be 97.20%, and the experimental efficiency is found to be 96.54% for the 900W inverter prototype.

Static and Dynamic Analysis of a Bistable Frequency Up-Converter Piezoelectric Energy Harvester

Abstract: Using energy harvesting to convert ambient vibrations efficiently to electrical energy has become a worthy concept in recent years. Nevertheless, the low frequencies of the ambient vibrations cannot be effectively converted to power using traditional harvesters. Therefore, a frequency up-conversion harvester is presented to convert the low-frequency vibrations to high-frequency vibrations utilizing magnetic coupling. The presented harvester consists of a low-frequency beam (LFB) and a high-frequency beam (HFB) with identical tip magnets facing each other at the same polarity. The HFB, fully covered by a piezoelectric strip, is utilized for voltage generation. The dynamic behavior of the system and the corresponding generated voltage signal has been investigated by modeling the system as a two-degrees-of-freedom (2DOF) lumped-parameter model. A threshold distance of 15 mm that divides the system into a monostable regime with a weak magnetic coupling and a bistable regime with a strong magnetic coupling was revealed in the static analysis of the system. Hardening and softening behaviors were reported at the low frequency range for the mono and bistable regimes, respectively. In addition, a combined nonlinear hardening and softening behavior was captured for low frequencies at the threshold distance. Furthermore, a 100% increment was achieved in the generated voltage at the threshold compared to the monostable regime, and the maximum generated voltage was found to be in the bistable regime. The simulated results were validated experimentally. Moreover, the effect of the external resistance was investigated, and a 2 MΩ resistance was found to be optimal for maximizing the generated power. It was found that frequency up-converting based on magnetic nonlinearity can effectively scavenge energy from low-frequency external vibrations.

Amplitude-phase characteristics of the resonant circuit when using it in the control circuit of thyristors voltage stabilizers

Abstract: This article examines the expansion of the falling section, the amplitude-phase characteristics of parametric circuits in order to determine the possibility of their application to control thyristors in the construction of new, simple and reliable voltage stabilizers. In parametric circuits connected to a voltage source with a low internal resistance, with a certain combination of parameters, the excitation of oscillations at the fundamental frequency is observed, the initial phase of which has a shift with respect to the phase of the applied voltage. Moreover, the phase of the excited oscillations depends on the magnitude of the applied electromotive force. By combining the connection circuits of linear and nonlinear elements, amplitude-phase characteristics of various shapes are obtained. This article reveals the general patterns in the change in the shape of the amplitude-phase relationships when varying various parameters, discusses the stability of stationary oscillations and their harmonic composition. The study reveals the feasibility of using a resonant circuit to control thyristors voltage stabilizers.

Degradation and Recovery of GaN HEMTs in Overvoltage Hard Switching Near Breakdown Voltage

Abstract: GaN high electron mobility transistors (HEMTs) have limited avalanche capability and usually fail catastrophically in voltage overshoot up to their dynamic breakdown voltage (BV _dyn ). This article presents the first comparative study of the parametric shift and recovery of three mainstream GaN HEMTs in repetitive overvoltage switching near their BV _dyn . In each switching cycle, a voltage overshoot up to 90% of BV _dyn was applied during the turn- off process. As the switching prolongs, all devices showed shifts in threshold voltage and saturation current, and these shifts saturated in less than 1-million cycles. These shifts are believed to be induced by the trapping of the holes generated in the impact ionization (I. I.). The device's poststress recovery was found to be dominated by the hole detrapping and through-gate removal, which highly depends on the gate stack. The gate injection transistor showed a fast natural recovery benefitted from the efficient hole removal through the Ohmic gate. The hole detrapping in the Schottky-type p-gate HEMT can be described by the Poole–Frenkel emission, allowing for the accelerated recoveries at negative gate bias and high temperatures. The hole removal in the metal-insulator-semiconductor (MIS) HEMT is blocked by the gate insulator, preventing a natural recovery. The MIS-HEMT can be recovered by applying positive gate and substrate biases, which facilitate the hole recombination in the channel. This article shows the good overvoltage robustness of all three GaN HEMTs and unveils effective methods for their poststress recovery, as well as suggests the significant impacts of I. I. and hole dynamics on the overvoltage ruggedness of GaN HEMTs near BV _dyn .

Numerical Modeling of Branching-Streamer Propagation in Ester-Based Insulating Oil Under Positive Lightning Impulse Voltage: Effects From Needle Curvature Radius

Abstract: Ester-based insulating oil provides the possibilities for controlling pollutant emissions and enhancing transformer safeties. A 2-D needle-plate electrode branching-streamer model has been established to numerically investigate the propagation characteristics of initial streamer branches and the effects from needle radius in ester-based insulating oil under lightning impulse (LI) voltage. The results indicate that by adding the source term that affects charge carrier density variation to the current continuity equation, the revised model could better capture the branching-streamer morphology in the experiment than the existing model. The accumulation of positive ions in the streamer branches promotes the electric field distortion, introducing the maximum electric field at the head of the streamer branches, which causes the appearance of the branching streamer due to the localized charge carrier surge (CCS). The electric field at the needle zone is affected by the curvature radius, and the smaller curvature radius enhances the electric field, allowing more space charge to be generated, which leads to easier distortion of the electric field at the head of the streamer branches and promotes forward propagation of the streamer branches. In addition, the smaller curvature radius contributes to the generation of the secondary streamer.

State-Space Modeling and Control of Grid-Tied Power Converters With Capacitive/Battery Energy Storage and Grid-Supportive Services

Abstract: Advances in the fields of renewable generation, electric vehicles, and energy storage systems push forward the research on ac–dc and dc–ac grid-tied power converters. However, the variabilities of power converters create new challenges in modeling and control. Existing state-space models fail to accurately describe various types of grid-tied converters (GTCs), particularly those with grid-supportive services, which are increasingly required by upcoming grid codes. As such, this article first proposes to classify GTCs into four basic types according to their ac and dc characteristics. Subsequently, corresponding detailed state-space models of GTCs are introduced, which serve as a useful tool for stability analysis. On top of that, this article further proposes control implementations of grid-supportive services related to active and reactive power control, including droop, inertia, and oscillation damping. Finally, simulation and experimental results demonstrate the effectiveness of the proposed models and grid-supportive services.

Transient Electrochemical Modeling and Performance Investigation Under Different Driving Conditions for 144Ah Li-ion Cell with Two Jelly Rolls

Abstract: Recently, the automotive industry has experienced rapid growth in powertrain electrification, with more and more battery electric vehicles (BEV) and hybrid electric vehicles being launched. Lithium-ion batteries play an important role due to their high energy capacity and power density, however they experience high heat generation in their operation, and if not properly cooled it can lead to serious safety issues as well as lower performance and durability. In that way, good prediction of a battery behavior is crucial for successful design and management. This paper presents a 1D electrochemical model development of a 144 Ah prismatic rolled cell using the GT-Autolion software with a pseudo 2D approach. The model correlation is done at cell level comparing model results and test data of cell open circuit voltage at different temperatures and voltage and temperature profile under different C-rates and ambient temperatures. After the cell level validation, a lumped battery pack model was submitted to different drive cycles: EPA, CLTC-P and WLTC having the battery current as input and comparing battery voltage and SOC. The cell level model presented good correlation with test data and the battery model proved capable of accurately make performance predictions. This model be used for further advanced investigations, such as cell degradation, thermal runaway and gas generation.

Flux-Weakening Controller Design of Dual Three-Phase PMSM Drive System With Copper Loss Minimization

Abstract: The performance of flux-weakening controllers in the multiphase permanent magnets (PM) machines depends on the accurate fundamental voltage limit derivation, and the minimal copper loss design in the flux-weakening operation region has not been well discussed. In this article, the accurate fundamental voltage limit with considering harmonic current suppression is derived at first. Then, two control concepts are addressed, and the corresponding flux-weakening control strategies are designed with the gradient descent method. In the first control strategy, both the fundamental and harmonic voltage vectors are feasible, resulting in the larger amplitude of the phase currents. The second control strategy obtains higher bus voltage utilization by an additional transition stage. At the same time, the harmonic currents cannot be further suppressed in the flux-weakening operation region. Third, the copper loss per electrical cycle is calculated, and a switching scheme is designed to obtain lower copper loss in the flux-weakening operation region. Finally, both strategies are successfully implemented in a dual three-phase PMSM. Moreover, the proposed switching scheme obtains minimal copper loss in the whole flux-weakening operation region.