Transformer Design Research Articles

DIRformer: A Novel Image Restoration Approach Based on U-shaped Transformer and Diffusion Models

Image restoration (IR) involves the retrieval of missing or damaged image information and represents a significant challenge in the field of visual reconstruction. Currently, U-Net based Diffusion Models (DMs) display favorable results when utilized for IR tasks. However, the Diffusion Model (DM) based on U-Net demonstrates shortcomings in capturing the global context for IR. To address this issue, we propose a Novel Image Restoration Approach Based on U-shaped Transformer and Diffusion Models (DIRformer). DIRformer enhances the modeling capacity for long-range dependencies within DMs. In particular, DIRformer replaces the traditional U-Net downsampling with Patch merging, dedicated to improving detail preservation, and replaces upsampling with Dual up-sample, strategically designed to alleviate checkerboard artifacts. Besides, as a lightweight and versatile transformer-based solution for IR, DIRformer incorporates time and degradation mapping into the transformer design, all while preserving the fundamental U-shaped structural framework. We assess the efficacy of DIRformer in a multi-tasking IR setting across four datasets. The experimental performance illustrates that DIRformer achieves competitive performance on distortion metrics, including PSNR and SSIM. Remarkably, our proposed approach is almost 25× smaller and 2× faster than the existing methods while achieving comparable high performance.

A Real-Life Laboratory Setting for Clinical Practice, Education, and Research in Family Systems Care: Protocol for a Transformational Action Research Study.

Burdening health and illness issues such as physical or mental illnesses, accidents, disabilities, and life events such as birth or death influence the health and functioning of families and contribute to the complexity of care and health care costs. Considerable research has confirmed the benefits of a family systems-centered care approach for patients, family caregivers, families, and health care professionals. However, health care professionals face barriers in working with families, such as feeling unprepared. Family systems-centered therapeutic conversations support families' day-to-day coping, resilience, and health. A family systems care unit (FSCU) was recently established as a real-life laboratory at one of the Swiss Universities of Applied Sciences. In this unit, health care professionals offer therapeutic conversations to families and individual family members to support daily symptom management and functioning, soften suffering, and increase health and well-being. These conversations are observed in real time through a 1-way window by other health care professionals, students, and trainees and are recorded with video for research and education. Little is known about how therapeutic conversations contribute to meaningful changes in burdened families and the benefits of vicarious learning in a real-life laboratory setting for family systems care. In this research program, we aim to deepen our understanding of how therapeutic conversations support families and individuals experiencing burdening health and illness issues and how the FSCU laboratory setting supports the learning of students, clinical trainees, and health care professionals. Here we apply a transformational action research design, including parallel and subsequent substudies, to advance knowledge and practice in family systems care. Qualitative multiple-case study designs will be used to explore the benefits of therapeutic conversations by analyzing recordings of the therapeutic conversations. The learning processes of students, trainees, and professionals will be investigated with descriptive qualitative study designs based on single and focus group interviews. The data will be analyzed with established coding methods. Therapeutic conversations have been investigated in 3 single-case studies, each involving a sequence of 3 therapeutic conversation units. Data collection regarding the second research question is planned. Preliminary results confirm the therapeutic conversations to support families' coping. This renders the FSCU a setting for ethically sensitive research. This program will not only support the health and well-being of families, but also contribute to relieving the financial and workforce burdens in the health and social care system. DERR1-10.2196/53090.

Three-Phase Bridge Fully-Controlled Rectifier Circuit Design Based On Digital Integrated Circuits

This paper aims to design and analyze a three-phase fully controlled bridge rectifier circuit. The paper first introduces the basic principles and advantages of the three-phase fully controlled bridge rectifier circuit and then elaborates on five key aspects of the design: transformer design, thyristor selection, thyristor protection design, trigger circuit design, and trigger circuit power supply design. During the design process, parameters such as the motor’s rated power, voltage, current, and resistance were considered to ensure that the selection of thyristor parameters and protective measures meet the needs of practical applications. In addition, the paper discusses the design of synchronous transformers, phase selection, the application of the KJ004 trigger circuit, and the design of the trigger circuit power supply. Finally, the circuit was simulated using Matlab simulation tools, the circuit design was optimized, and satisfactory pulse waveforms and output waveforms were obtained by adjusting the experimental time. The research in this paper provides theoretical basis and practical guidance for the optimized design of three-phase fully controlled bridge rectifier circuits.

The integration and transformative design research of plant dyeing techniques in Dong brocade with sustainable materials

Dong brocade, a fabric renowned for its intricate patterns and ethnic symbolism, has been woven by the Dong people for generations, showcasing their cultural significance. Traditional plant dyeing technology is one of the main aspects of Dong brocade but the documentation and understanding of this is still rather limited. With regard to the use of plant dye in Dong brocade, it is not as well explored as it should be since it has a traditional aspect. The main purpose is to investigate and apply the traditional plant dyeing technique to Dong brocade for the improvement of that sustainable concept and the preservation of cultural assets. Therefore, 121 Dong villagers were interviewed to elicit their awareness regarding prehistoric plant dyeing. By observing the dyeing conditions, this study provided accurate perception and learned how to differentiate between natural and synthetic mordants through ethnobotanical perception. The strategy is intended to integrate sustainable products into Dong brocade, employing orthogonal array development to find the right dyeing conditions for corresponding plant dyes. Research revealed that 8 genera of plants which include 7 species are used in dyeing Dong brocade. The findings presented in this work prove the effectiveness of the use of plant dyes in Dong brocade, showing its advantages with 30% of frequency and CI (Color Index) indices, 8% of them being cultural. 5 for ethnic cultural sustainment, developmental and bio-diversity reasons respectively. The unique integration between the traditional dyeing technique in Dong brocade and the utilization of sustainable resources is very promising for the improvement of identity enhancement and embodiment, and the preservation of the environment.

Modeling and Simulation of Wide-Frequency Characteristics of Electromagnetic Standard Voltage Transformer

To achieve the broadband applicability of standard voltage transformers in “dual high” power systems, an equivalent circuit model of the standard voltage transformer is first established. Using the complex magnetic permeability method and utilizing existing core parameters, the excitation impedance values are obtained. Next, based on the equivalent circuit model, the no-load error function of the standard voltage transformer is analyzed, and through simulation, the no-load error response curve of the standard voltage transformer in the frequency range of 20 Hz to 3000 Hz is derived. The simulation results indicate that within the 20 Hz to 700 Hz range, both the no-load ratio error and the no-load angular error meet the accuracy requirements, with the ratio error within ±0.05% and the angular error within 2′. Additionally, the derivation of the error transfer function demonstrates the correlation between the no-load error values and the number of turns and cross-sectional area of the standard voltage transformer. Simulation results, obtained by increasing and decreasing the number of turns and cross-sectional area by 10%, provide valuable insights for the error compensation and structural design of standard voltage transformers.

Enhancement Design of Multi-phase Transformer for Cascaded H-Bridge Motor Driver

Enhancement Design of Multi-phase Transformer for Cascaded H-Bridge Motor Driver

Dynamic Analysis of Metamorphic Mechanisms with Impact Effects During Configuration Transformation

Metamorphic mechanisms have attracted considerable attention owing to their capability to switch their topology to adapt to different operational tasks. One feature of topological change is the re-contact of different bodies, which inevitably causes collisions affecting operation accuracy and service life. Consequently, in this study, a collision incidence matrix was introduced to describe the topology of a system involved in collisions, and a method for reducing the closed-loop system to an open-loop system was proposed. The complex movement of the metamorphic mechanism in a changing topology was classified into two different running stages of the source metamorphic mechanism. Based on the relative coordinate method, dynamic modeling of the source metamorphic mechanism considering the impact effects was conducted. Combining the classical collision theory and Newton–Euler equation, the generated impact impulse and the motion after collision were determined. Subsequently, a dynamic analytical method for the full configuration of metamorphic mechanisms was proposed to reflect the changes in the topological structure in the dynamic model. Finally, two typical metamorphic mechanisms used in packaging and spinning were considered as examples to verify the correctness and effectiveness of the proposed method, and their impact characteristics during configuration transformation were analyzed. The proposed analytical method of internal impact for a variable topology process provides effective theoretical guidance for the stability analysis of configuration transformation and structural design aimed at minimizing impacts.

Multifunctional Control Strategy for a Hybrid Solid-State Transformer Applied to Modern Distribution Electric Grids

This paper presents the control and hardware design for a Hybrid Solid-State Transformer (HSST) applied to modern distribution systems. The HSST combines the advantages of conventional transformers, such as high efficiency and low cost, with those of Solid-State Transformers (SST), such as multifunctionality and fast dynamic response. Real-time simulation using a Typhoon HIL404 device is performed to validate the proposed multifunctional control strategy. The corresponding results validate the HSST’s capability to provide, while only processing partial system power, a regulated output voltage with grid voltage harmonic suppression and, at the same time, current load reactive and harmonic compensation to ensure a high-power factor and improved power quality to the grid.

My ideal teacher: the Gen Z’s choice

This study explored the Gen Z’s ideal teacher. The research employed an explanatory sequential mixed method. The intent of the transformative mixed methods design, to seek reality from their preferred teaching styles, on how do they perceived an ideal teacher, and on how can teachers meet their expectations. The findings revealed that the most preferred teaching styles and most effective way in teaching these students is experiential learning. In addition, as perceived by the Gen Z students, an ideal teacher should have positive values and they must be emotionally supportive and technologically literate. Furthermore, it was unveiled that a teacher must imposes skill enhancement, be conscious of today’s trend, adapt to the digital revolution and have an awareness of individual differences and needs to meet the Gen Z expectations. This also provides information and serves as a guide on how teachers meet the Gen Z expectations based on the overall data gathered from the respondents. Through this, an instructional input was crafted containing the different teaching styles that the student prefers and the different activities that can help teachers accommodate the said teaching styles.

Novel modular multilevel matrix converter topology for efficient high-voltage AC-AC power conversion

Introduction. This paper delves into the practical application of multilevel technology, particularly focusing on the capacitor-clamped converter as a promising solution for medium-to-high voltage power conversion, with specific emphasis on direct AC-AC switching conditions. Problem. The limitations of conventional single-cell matrix converters (MC) in efficiency and performance for medium-to-high voltage power conversion applications are well-recognized. Goal. The primary objective is to investigate the performance of the 3 phase modular multilevel matrix converter (3MC) with three flying capacitors (FCs) modeling. This investigation utilizes the Venturini method for gate pulse generation, aiming to compare the performance of the 3MC with standard converter designs. Methodology. To achieve the research goal, the Venturini method is adopted for generating gate pulses for the 3MC, representing a departure from conventional approaches. Detailed simulations employing MATLAB/Simulink are conducted to comprehensively evaluate the performance of the 3MC in comparison to conventional converter designs. Results. The simulation outcomes reveal a significant reduction of 73 % in total harmonic distortion (THD) achieved by the 3MC. This reduction in THD indicates improved robustness and suitability for medium-to-high voltage power conversion systems necessitating direct AC-AC conversion. These results highlight the efficacy of the 3MC in enhancing power conversion efficiency and overall performance. Originality. This paper contributes novel insights into the practical implementation of multilevel technology, particularly within the realm of capacitor-clamped converters. Furthermore, the utilization of the Venturini method for gate pulse generation in the 3MC represents an original approach to enhancing converter performance. Practical value. The research findings present significant advancements in multilevel transformer technology, offering valuable guidance for optimizing transformer design in various industrial and renewable energy applications. These contributions serve to enhance the development of reliable and efficient power systems, addressing critical needs in the energy sector. References 54, tables 3, figures 4.

Stable successive Neural Image Compression via coherent demodulation-based transformation

Neural Image Compression (NIC) has made significant strides in recent years. However, the existing NIC methods demonstrate instability issues during iterative re-compression cycles, which can degrade image quality with each cycle. This paper introduces a novel framework aimed at enhancing the stability of NIC methods. We first conducted a theoretical analysis and identified that the instability in current NIC methods stems from a lack of idempotency in transformations. Drawing from the domain of signal processing, we then examined the principles of idempotency in coherent demodulation techniques. This examination led to the identification of three foundational principles that inform the design of stable transformations: the cosine function, parameter sharing, and low-pass filtering. Leveraging these insights, we propose the innovative Coherent Demodulation-based Transformation (CDT), which is designed to address the stability challenges in NIC by incorporating these principles into its architecture. The experimental results suggest that CDT not only significantly improve the re-compression stability but also preserves the codec’s rate–distortion performance. Furthermore, it can be broadly applied in current NIC structures. The effectiveness of the module endorses the viability of designing transformation networks based on Coherent Demodulation principles, playing a crucial role in enhancing stability of NIC. The code will be available at https://github.com/baoyu2020/Stable_SuccessiveNIC.

Optimal Design and Analysis of High-Frequency Isolation Transformer for Switched-Mode Power Converters

High-frequency (HF) transformers have gained great interest in recent years due to the advent of powerful soft magnetic materials with low core loss in semiconductor power switches. Also, the optimal design of the HF transformer is a significant issue for high-performance energy conversion systems. In this paper, a 40W 50/12.5/25 V universal input and two output discontinuous-conduction mode (DCM) flyback transformer is designed by using mathematical calculations and analyzed via 3D ANSYS/Maxwell simulation including electromagnetic and loss analysis. It is shown that the simulation results accounting for hysteresis losses, eddy current losses, copper losses, and magnetic flux density determine the accuracy of the mathematical model calculation. Analyzes are performed at 100 kHz frequency levels. Results obtained will include core magnetic flux density, core/copper losses, leakage/magnetizing inductances, windings parasitic capacitances, input/output voltage, current values, and all design parameters. Finally, the proposed HF transformer's overall efficiency is calculated and presented. Significantly, the HF transformer achieves 97.8% efficiency thanks to the transformer's core and coil selection, B-H and B-P characteristics, one-to-one dimension design, and mesh operation. The dynamic and mathematical results of the designed transformer demonstrate the design and efficiency success

Symmetrical Transformer Design for Suppressing Even-Order Components in CM Noise Applied to Converter With Fixed-Frequency PWM Strategy

Symmetrical Transformer Design for Suppressing Even-Order Components in CM Noise Applied to Converter With Fixed-Frequency PWM Strategy

The Effect of Digital Transformation, User Experience, and User Interface on Customer Loyalty in Technology Startups in Bandung

This study investigates the influence of digital transformation, user experience (UX), and user interface (UI) on customer loyalty in technology startups in Bandung, Indonesia. A quantitative approach was employed, with data collected from 150 respondents using a Likert scale ranging from 1 to 5. The data were analyzed using SPSS version 26, and multiple regression analysis was used to determine the significance of the relationships between the variables. The results reveal that digital transformation, UX, and UI all have positive and significant effects on customer loyalty, with UX being the strongest predictor. These findings provide important insights for technology startups in Bandung, suggesting that enhancing digital processes, optimizing UX, and refining UI design can foster greater customer loyalty. This study contributes to the growing body of literature on the role of digital transformation and customer-centric design in building long-term customer relationships in the startup ecosystem.

Leveraging Service Design for Higher Education Transformation

How do we help today’s college students make the most of their time, energy and money invested? This question is at the heart of ongoing discussions among educators, policymakers and stakeholders. Many solutions lie in new programmes, policies and structures, particularly to support minoritised populations. But are they really working?

Building-block-flow computational model for large-eddy simulation of external aerodynamic applications

Computational fluid dynamics is an essential tool for accelerating the discovery and adoption of transformative designs across multiple engineering disciplines. Despite its many successes, no single approach consistently achieves high accuracy for all flow phenomena of interest, primarily due to limitations in the modeling assumptions. Here, we introduce a closure model for wall-modeled large-eddy simulation to address this challenge. The model, referred to as the Building-block Flow Model (BFM), rests on the premise that a finite collection of simple flows encapsulates the essential missing physics necessary to predict more complex scenarios. The BFM is designed to: (1) predict multiple flow regimes, (2) unify the closure model at solid boundaries and the rest of the flow, (3) ensure consistency with numerical schemes and gridding strategies by accounting for numerical errors, (4) be directly applicable to arbitrary complex geometries, and (5) be scalable to model additional flow physics in the future. The BFM is utilized to predict key quantities in five cases, including an aircraft in landing configuration, demonstrating similar or superior capabilities compared to previous state-of-the-art models. The design of BFM opens up new opportunities for developing closure models that can accurately represent various flow physics across different scenarios.

Design and Optimization of Air-Core HTS Pulse Transformer for Series-Type Hybrid Circuit Breaker (S-HCB)

Design and Optimization of Air-Core HTS Pulse Transformer for Series-Type Hybrid Circuit Breaker (S-HCB)

The method and application of social practice in the clubhouse.

The theory of social practice describes a therapeutic community process for people living with serious mental illness, while the methods involve engaging people to become collaborators and contributors to a social environment. Confusion in the mental health field surrounding the applied methods of social practice-as occurring within the clubhouse model-has not been sufficiently addressed. This article aims to outline the methodology of social practice, as well as provide guidance on its practical application. Specific constructs of social practice are defined, and empirical support is provided to emphasize how practices to support recovery are applied in the clubhouse. Five key elements of social practice-that is, transformational social design, engagement, relationship development, natural feedback and intervention, and transitional environments-are described. Practical examples of therapeutic techniques associated with each element are also provided. Social practice addresses the social determinants of health by focusing on five key elements that support individual and community recovery. Here, we detail social practice techniques utilized in the clubhouse as a way of unifying theory, providing practical guidance to mental health professionals and improving the measurement of community as therapy for serious mental illness. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

Divergent dual C-H isotopic fractionation pattern during anaerobic biodegradation of toluene within Aromatoleum species under nitrate-reducing conditions

Toluene is a pollutant frequently detected in contaminated groundwater, mostly due to leakage from underground gasoline storage tanks and pipeline ruptures. Multi-element compound-specific isotope analysis provides a framework to understand transformation processes and design efficient remediation strategies. In this study, we enriched an anaerobic bacterial culture derived from a BTEX-contaminated aquifer that couples toluene and phenol oxidation with nitrate reduction and the concomitant production of carbon dioxide and biomass. The 16S rRNA gene amplicon data indicated that the toluene-degrading consortium was dominated by an Aromatoleum population (87 ± 2 % relative abundance), and metagenome sequencing confirmed that the genome of this Aromatoleum sp. encoded glycyl-radical enzyme benzylsuccinate synthase (BssABC) and phenylphospate synthase (PpsA1BC) homologous genes involved in the first step of toluene and phenol transformation, respectively. Carbon and hydrogen isotopic fractionation were εbulk, C = - 3.5 ± 0.6 ‰ and εrp, H = - 85 ± 11 ‰, respectively, leading to a dual C-H isotope slope of ΛH/C = 26 ± 2. This value fits with a previously reported value for a consortium dominated by an Azoarcus species (ΛH/C = 19 ± 5) but differs from that reported for Aromatoleum aromaticum (ΛH/C = 14 ± 1), both of which grow with toluene under nitrate-reducing conditions. Overall, this suggests the existence of different BssABC enzymes with different mechanistic motifs even within the same Aromatoleum genus.

MCGnet with multi-level gated unit transformations: A time series prediction-based model for unknown pattern abnormality detection

The task of anomaly detection in data patterns remains challenging due to the diverse of fault patterns, which often render existing models ineffective when encountered with unknown or unseen data patterns in the training data. Furthermore, insufficient samples, the presence of chaotic behavior, unbalanced anomaly patterns are potential limiting factors in the practical application. While prediction error-based approaches attempt to address these challenges, the discriminability of abnormal patterns in existing scenarios is not yet sufficient, thus limiting the ability to identify abnormal phenomena. To overcome these limitations, we proposed a novel model called MCGnet. The backbone architecture of MCGnet employed a novel design of continuous multi-layer circular convolution transformations. This design enhances the expressive capability of the model by expanding the differences among data patterns. Additionally, we introduced in parallel the newly proposed temporal-frequency gated Gaussian unit, which generates richer feature representations. Subsequently, the obtained feature representations were fed into an explicit causal network to learn complex patterns of data along temporal dependencies. The MCGnet not only solves high-precision time series prediction problems but also significantly improves the discriminability between normal and abnormal patterns. Since the proposed model is accurately modeled using only a small amount of normal data, it bypasses the limiting factors such as insufficient fault samples and sample imbalance, and is able to effectively deal with various unknown data patterns. The accuracy of the proposed method was verified through simulation data (Duffing system) and its potential for engineering applications was demonstrated through real-world data (rotor system).