Tạp chí Khoa học và Công nghệ - Đại học Đà Nẵng

AI-assisted ECU programming for twining injectors supplying flexible syngas-biogas-hydrogen blend to engines

Bui Van Hung, Duong Trong Chung, Le Khac Binh — Wed, 31 Dec 2025 00:00:00 +0700

The engine utilizing a flexible variable blend of syngas-biogas-hydrogen fuel requires fuel supply via a dual-injector system controlled by a specially designed ECU. The control law for the injectors in the dual-injector system can be described by two piecewise linear functions: the first injector’s opening increases linearly, then is held constant; the second injector begins to open when the syngas concentration in the mixture exceeds 50%, and then increases linearly. Artificial Intelligence (AI) can be employed to program the complex control operation of the dual-injector system based on the provided hardware specifications and the injector opening rule. The adaptability of the ECU to the fuel mixture composition and the engine operating regime can be simulated using the Proteus software. This allows us to save a significant amount of time and effort in ECU testing before manufacturing.

Comparative study of direct torque control and field-oriented control for six-phase permanent magnet synchronous motor

Thi-Anh-Em Bui, Trong Hai Nguyen, Hoang Than, Dai-Qui Vo, Quoc Thien Pham — Wed, 31 Dec 2025 00:00:00 +0700

This study presents a MATLAB-based comparative simulation of a six-phase Permanent Magnet Synchronous Motor (PMSM) under field-oriented control (FOC) and direct torque control (DTC). The results show that FOC maintains a more stable speed response than DTC, exhibiting smaller errors and reduced ripple, particularly during rapid acceleration steps. Mean torque and RMS fluctuation analyses indicate that FOC delivers smoother torque with significantly lower short-term oscillations, whereas DTC provides a faster torque transient at the expense of larger torque fluctuations, especially at low and medium speeds. Phase-current evaluation further confirms improved current regulation with FOC, with RMS and peak-to-peak values lower or comparable to those of DTC at high speed, reflecting reduced oscillatory behavior. These trends remain consistent across 0–3000 rpm, highlighting the trade-off between the faster torque response of DTC and the smoother torque and current performance of FOC in multiphase PMSM drives.

Design of automotive traction power distribution controller considering the ABS actuator

Phan Tan Tai, Tran Van Nhu, Ngo Thanh Ha, Phan Van Tuan — Wed, 31 Dec 2025 00:00:00 +0700

The contact state between the wheel and the road surface significantly affects automotive traction performance. On rough roads, differential slippage between the left and right wheels leads to uneven traction power, which may cause vehicle instability or immobilization, resulting in engine power loss. This study proposes a method for managing traction power by controlling the ABS actuator to generate wheel slip resistance as required. In this paper, a cascade controller is developed to regulate both the speed of the drive wheels and the oil pressure in the wheel brake mechanism, thereby preventing wheel slip. This approach enables the differential to distribute power more effectively to the active axle, enhancing engine efficiency. System evaluation using Matlab/Simulink software demonstrate the high responsiveness of the proposed solution.

Development of a framework for power system operational risk assessment using probabilistic power flow with SCADA data

Luong Phu Nhan, Hoang Minh Ha, Le Thi Da Thao — Wed, 31 Dec 2025 00:00:00 +0700

High penetration of renewable energy poses significant challenges to power system operational security. This paper presents the development of a comprehensive framework for operational risk assessment using a data-driven Probabilistic Power Flow (PPF) approach. The proposed framework automatically identifies the best-fit probability distributions for loads and generation sources, constructs a cluster-based correlation model, and quantifies risks via an enhanced Monte Carlo simulation using Latin Hypercube Sampling (LHS). The framework's effectiveness is validated through a case study on the Vietnamese Central Power System. Results demonstrate that the methodology accurately quantifies operational risks, highlighting the critical importance of correlation modeling. This framework provides a powerful and widely applicable tool to support secure grid operation and planning.

Evaluation of the effectiveness of applying 3D models of typical automotive parts in education

Tran Van Tung, Tran Cong Chi, Pham Van Tinh — Wed, 31 Dec 2025 00:00:00 +0700

This study evaluated the effectiveness of 3D models in improving learning outcomes in automotive engineering education. The models were created using SolidWorks and exported to HyperText Markup Language (HTML) format via the eDrawings tool, allowing use without specialized software. They were integrated into automotive engineering subjects to support students’ visualization of complex components. To assess effectiveness, a survey of 75 students was conducted focusing on three aspects: Visualization and Comprehensibility (VC), Interest and Engagement (IE), and Retention and Application (RA). Results showed that mean scores were significantly higher than the neutral value (p < 0.001), with Cronbach’s α = 0.813 indicating good reliability and positive correlations among the aspects. The findings confirm that 3D models enhance training effectiveness by improving visualization, interaction, and knowledge retention. The study provides a practical framework for developing visualization tools that help connect theory to practice.

Experimental construction of a cascade control structure for liquid temperature stabilization in a stirred tank

Tran Hieu Nghia, Ngo Quang Thinh, Nguyen Dang Khang, Ngo Manh Tung — Wed, 31 Dec 2025 00:00:00 +0700

In food and beverage production, precise liquid temperature control is crucial for product quality. Similarly, in the nuclear industry, liquid temperature strongly affects efficiency, safety, and lifespan. This paper presents a cascade control method for liquid temperature regulation in a mixing tank, with an inner loop regulating flow rate via pump motor speed and an outer loop controlling temperature. The two-loop structure is compared with a conventional single-loop PI controller on the TecQuipment CE117 experimental module. Experimental results show that the cascade PI improves control quality, reducing settling time by approximately 17–36% under different operating conditions and halving recovery time under pump-stoppage disturbances, compared with the single-loop PI. These results indicate that the proposed cascade architecture is suitable for thermal process applications requiring accurate and robust temperature control.

Experimental study on HEV series truck model

Van Phuc Nguyen, Van Nam Tran, Minh Duc Le — Wed, 31 Dec 2025 00:00:00 +0700

This study presents the design of a serial HEV model and experimental evaluation of performance, fuel consumption measurement, and pollution emission index. Experimental results of fuel consumption measurement on the same cycle show that fuel consumption in HEV mode is reduced by 36.72% compared to ICE mode. The HEV model has a significantly lower emission index than the ICE internal combustion engine during most of the cycle. During the operating stages, the emission time of the ICE experimental cycle always has a period of emission maintenance at 800 rpm. Both cycles reach a sudden emission level of up to about 1.7%, however, the HEV cycle remains essentially HSU emission-free throughout the first 150 minutes and more than
120 minutes of operation at 1500 rpm of the ICE. The smoke coefficient of the test cycle is within the allowable limits of the national standard QCVN 09:2024/BGTVT.

Reliability-based design optimization of V-belt drives using inverse reliability analysis and monte carlo simulation

Tran Van Thuy — Wed, 31 Dec 2025 00:00:00 +0700

V-belt drives are widely used in industrial and agricultural machinery, yet conventional deterministic design often ignores uncertainties in materials, operating conditions, and tolerances. This study proposes a Reliability-Based Design Optimization (RBDO) framework for V-belt drives, integrating Inverse Reliability Analysis with a Genetic Algorithm and validated by Monte Carlo Simulation. Design variables include driver and driven pulley diameters, belt length, and number of belts, with objectives of minimizing mass and maximizing transmitted power under a reliability target of R* = 99.9%. Results indicate that RBDO reduces system mass by 18.2% and increases power from 4.75 kW to 5.65 kW compared to deterministic optimization, while ensuring reliability. The proposed framework provides a practical tool for lightweight, reliable, and high-performance V-belt drive design.

Synergetic effect of zinc oxide, iron oxide, and bentonite/zeolite P1 for the effective crystal violet removal based on Fenton-like oxidation

Wed, 31 Dec 2025 00:00:00 +0700

In this study, a ZnO/α-Fe₂O₃/bentonite/zeolite P1 (ZFBP) composite was successfully synthesised using natural bentonite sourced from Vietnam. The influence of zinc concentration on the Fenton-like catalytic efficiency toward crystal violet (CV) degradation and the physicochemical characteristics of ZFBP was evaluated through various analytical methods. At Fe₂O₃ and ZnO loadings of 15.14 wt.% and 7.34 wt.%, respectively, 92.54% degradation of CV was achieved in 40 min. Kinetic analysis revealed that the CV removal followed a pseudo-first-order model, with a reaction rate constant of 0.06 1/min under the optimal conditions. The removal mechanisms comprised of adsorption and the Fenton-like oxidation reaction, with the synergistic effect of the ZFBP components actively supporting the good removal efficiency. Furthermore, the composite demonstrated remarkable reusability after 10 cycles. These findings highlight the promising potential of ZFBP as an effective material for the treatment of organic dyes.

A practical approach to building legal knowledge graphs from legal texts for legal consultation systems

Huynh Le Huy, Khuat Thanh Tung, Le Thi My Hanh — Wed, 31 Dec 2025 00:00:00 +0700

In the context of applying artificial intelligence to the legal domain, building legal question-answering (QA) systems requires a structured, queryable, and inferable knowledge foundation. This paper proposes a semi-automated method to extract and map legal knowledge from the Penal Code of Vietnam into a knowledge graph, supporting criminal law QA systems. The solution includes five main steps: (1) extracting legal text from the PDF files, (2) preprocessing and normalizing the text into individual articles, (3) using large language models (LLMs) to generate fundamental knowledge components based on the predefined rule set, (4) mapping these components into nodes and edges to construct the knowledge graph, and (5) optimizing and visualizing the graph. Preliminary results show that the model can accurately represent legal relationships such as violations, penalties, applicable subjects, and inter-article references, which will form the groundwork for the future automated legal QA applications.

Development of a computational tool for sEMG decomposition

Pham Nam Son, Tran Ngoc Dung — Wed, 31 Dec 2025 00:00:00 +0700

Motor unit decomposition from HD-sEMG (high density surface electromyography) provides valuable information for reviewing the neuromuscular system. The purpose of this study was to implement and validate the framework of the convolutional blind source separation proposed by Negro [1]. We tested on the Hyser (High densitY Surface Electromyogram) dataset with healthy subjects that perform low force finger movements during the execution of a task performed with a system of HD sEMG of 256 channels. Motor units were extracted using 3-second sliding windows with strict quality criteria (silhouette score ≥0.85). Decomposition yielded 4,536 quality motor units with a mean silhouette score of 0.926±0.023. The test-retest reliability indicated moderate stability of quality assessed parameters (ICC=0.52-0.59), whereas appropriately variable firing characteristics (ICC<0.10), confirming that the algorithm produced consistent outputs while still able to detect the temporal discharge variability.

Determinants of 5G adoption: a comparative analysis of Vietnam and the United States

Wed, 31 Dec 2025 00:00:00 +0700

Synthesizing TAM, UTAUT, and DOI frameworks, this study compares 5G acceptance drivers between consumers in Vietnam and the United States (N=522). PLS-SEM analysis evaluated the impact of perceived usefulness, ease of use, social influence, environmental and health awareness, and perceived value on usage intention and acceptance. Findings indicate that perceived usefulness, ease of use, social influence, and environmental awareness positively affect intention. While health awareness strengthens intention, perceived value unexpectedly exerts a negative moderating effect on the intention-acceptance relationship. Crucially, the moderating role of nationality proved insignificant, suggesting cross-market consistency in adoption behaviors. These results extend technology acceptance theories and imply that strategies promoting 5G should prioritize communicating functional benefits, ease of use, and distinct value propositions to overcome adoption barriers globally.

Factors influencing the acceptance of using generative AI to support English language learning among university students: a case study in Vietnam

Vo Nguyen Da Thao — Wed, 31 Dec 2025 00:00:00 +0700

This study investigates factors influencing Vietnamese university students’ acceptance of generative AI tools (e.g., ChatGPT, Gemini) in English learning. Using a mixed methods design and based on TAM and UTAUT frameworks, it proposed a model with eight factors affecting students’ “Intention to Use AI.” Linear regression analysis indicated that these factors explained 34.8% of the variance (R² = 0.348). “Perceived Usefulness” (PU) had the strongest effect (β = 0.458, p < 0.001), followed by “Perceived Ease of Use” (PEOU, β = 0.116), “Institutional and Instructor Support” (IIS, β = 0.125), “Concern about Accuracy” (CAA, β = 0.135), and “Self-Regulated Learning” (SRL, β = 0.133). “Social Influence”, “Trust in AI”, and “Attitude” were not significant (p > 0.05). The study highlights the importance of institutional AI policies and recommends fostering students’ critical thinking to prevent over-reliance on AI.

Strategic competition between China and Japan (2012–2025) and implications for Vietnam

Nguyen Khoa Tuan — Wed, 31 Dec 2025 00:00:00 +0700

China-Japan relations constitute one of the most complex bilateral relationships in the Asia-Pacific region. The period from 2012 to 2025 marks a significant transformation from latent competition to comprehensive strategic confrontation across three domains: political-diplomatic, security-defense, and economic-technological. Employing qualitative analytical methods and a realist theoretical framework, this article elucidates the principal drivers and manifestations of competition, including the Senkaku/Diaoyu Islands dispute, the arms race, and semiconductor technology rivalry. Based on this foundation, the study assesses the impact of Sino-Japanese competition on Vietnam's strategic space and proposes a "flexible hedging" orientation to safeguard national interests. Vietnam must remain steadfast in its independent and self-reliant foreign policy, diversify international relations, enhance its national defense self-reliance, and leverage ASEAN's role in regional security governance.

The impact of feedforward strategy on essay writing skills of English-majored students at Dong Nai Technology University

Nguyen Thi Thanh Van, Lai Ngoc Kha — Wed, 31 Dec 2025 00:00:00 +0700

This study explores the impact of the feedforward strategy on the enhancement of essay writing abilities. In contrast to conventional feedback correcting mistakes, feedforward emphasises the importance of offering constructive suggestions and guidance prior to or during the writing process. Limited information exists regarding the application of feedforward as a strategy that facilitates enhanced performance. This investigation employed an action research framework utilising a mixed-method approach to assess the influence of feedforward on the writing performance of English-majored students in writing IELTS Task 2 essays. Students who received feedforward support demonstrated notably enhanced writing quality. Insights from the semi-structured interviews revealed that participants held favourable views regarding the use of feedforward as a formative strategy. The findings indicate that the integration of feedforward techniques can significantly improve learners’ academic writing abilities while also fostering increased motivation and confidence in their writing endeavours.

Modifications of Vietnam's cultural policies during the participation in the ASEAN Socio-Cultural community (ASCC)

Do Thi Minh Trang — Wed, 31 Dec 2025 00:00:00 +0700

Since the country's reunification in 1975, Vietnam has implemented numerous policies to preserve, conserve, and promote its national cultural values. The Party and State of Vietnam have recognized culture as a crucial factor in stabilizing society and strengthening national security and politics, as demonstrated by their cultural policies throughout subsequent historical periods. In particular, since joining ASCC (2015), Vietnam has made many adjustments in the promulgation and implementation of cultural policies within the framework of ASCC cooperation. The article analyzes Vietnam's cultural perspectives and policy adjustments before and after joining the ASEAN Community, from cautious steps following the unification period to innovation during the “open door” period, and then to strong integration within the ASEAN Community, all aimed at achieving a sustainable future, thereby enhancing Vietnam's position in the international zone.

Some semantic features of final particles in the Thai language

Phung Thi Huong Giang — Wed, 31 Dec 2025 00:00:00 +0700

Final particles in the Thai language are a highly important linguistic element and appear very frequently in everyday conversation. These particles are characterized by complex semantic features that are difficult to grasp and analyze. Depending on the communicative context and the interlocutors involved, final particles can convey different meanings and communicative intentions. Although the number of final particles is not as large as that of other word classes in Thai, they play a crucial role in the language’s grammatical system due to their ability to express the speaker’s emotions, mood, or attitude in relation to the listener. This article aims to examine the use of final particles in Thai through real-life examples, analyze them from both grammatical and semantic perspectives, and identify their most generalizable functional characteristics. It is hoped that this study will help learners and users of Thai better understand these particles and use them appropriately in daily communication.