Articles

Access the latest knowledge in applied science, electrical engineering, computer science and information technology, education, and health.

Advanced Search
Filter Icon

Filters article

Years

FAQ Arrow
Single Year
Range Year
0
0

Source Title

FAQ Arrow

Authors

FAQ Arrow

31,456 Article Results

Edge-aware distilled segmentation with pseudo-label refinement for autonomous driving perception

10.11591/ijra.v14i4.pp376-386
Novelio Putra Indarto , Oskar Natan , Andi Dharmawan
Achieving precise semantic segmentation is essential for enabling real-time perception in autonomous systems, yet leading approaches typically require substantial annotated data and powerful hardware, restricting their use on devices with limited resources. This work introduces an efficient segmentation framework that integrates pseudo-label refinement, knowledge distillation, and entropy-based confidence filtering to train compact student networks suitable for edge deployment. High-quality pseudo-labels are first produced by a robust teacher network, then further improved using a dense conditional random field to boost spatial consistency. An entropy-based selection mechanism removes unreliable predictions, ensuring that only the most trustworthy labels guide the student model's training. The use of knowledge distillation effectively transfers detailed semantic understanding from the teacher to the student, enhancing accuracy without added computational overhead. Experimental results with multiple EfficientNet backbones reveal that this pipeline improves segmentation accuracy and output clarity, while also supporting real-time or near real-time inference on CPUs with limited processing power. Extensive ablation and qualitative studies further confirm the method's robustness and flexibility for real-world edge applications.
Volume: 14
Issue: 4
Page: 376-386
Publish at: 2025-12-01

Study of asymmetrical-multi level inverter using two switching angle techniques

10.11591/ijpeds.v16.i4.pp2570-2581
Dewan Ashikur Rahaman , Tapan Kumar Chakraborty
An inverter is a device that transforms DC power into AC power. Inverters can be categorized into single-level inverters and multilevel inverters. This paper discusses two controlled strategies-equal step angle and sinusoidal switching angle-for a multilevel inverter, highlighting their effectiveness in harmonic mitigation as the number of voltage levels increases. The simulation software used to generate 3-15 level voltage outputs is PSIM, which allows for the adjustment of switching angles based on both equal step and sinusoidal switching values. Various types of DC sources are connected to H-bridge units, with MOSFET driving signals applied via gating blocks. The study demonstrates a notable reduction in total harmonic distortion (THD) when the switching angles are altered in equal and sinusoidal steps. Initially, the output signal generates a square wave without a filter. However, after implementing an LC filter, the output voltage signal more closely resembles an AC signal, and THD values are further reduced. Additionally, the output voltage signal's fast Fourier transform (FFT) is presented.
Volume: 16
Issue: 4
Page: 2570-2581
Publish at: 2025-12-01

Bidirectional AC/DC converter connecting AC and DC microgrids for smart grids

10.11591/ijpeds.v16.i4.pp2549-2561
Nguyen Van Dung , Nguyen The Vinh
This paper proposes a converter connecting two independent AC and DC microgrids in a flexible microgrid and smart grid system. With this converter, basic DC/DC converter types such as Flyback are used to develop the power circuit and controller for the converter that is capable of integrating the operating functions for the operation between microgrids. The converter uses bidirectional switching locking technology to simplify the control algorithm. The energy is converted in two directions, AC/DC and DC/AC, with different working principles of increasing and decreasing voltage according to the standards of the distribution grid and DC microgrid. The TDH value is significantly limited when using the recovery circuit solution. The converter is designed, simulated based on OrCAD software, and tested with a capacity in the range of 2-10 kW. The DC microgrid output voltage is 400 VDC, voltage is 220 VAC.
Volume: 16
Issue: 4
Page: 2549-2561
Publish at: 2025-12-01

A survey on convolutional neural network hardware acceleration through approximate computing multiple and accumulates unit

10.11591/ijra.v14i4.pp366-375
Suvitha Pathiyadan Sudhakaran , Aathmanesan Thangakalai
Convolutional neural networks (CNNs) are applied to a different range of real-world complex tasks to provide effective solutions with high accuracy. Based on the application's complexity, CNN demands a lot of processing units and memory spaces for its effective implementation. Bringing this computational task to hardware for processing the data to enhance the acceleration helps in achieving real-time performance improvement. Recent studies focused on approximation methodology to overcome this problem. This proposed survey analyzes various recent methods involved in implementing approximating computing-based processing elements and their usage in CNNs. Primarily, the survey focuses on multiple and accumulates (MAC) unit and their various approximation methods, which acts as a fundamental block as a processing element in the CNN layers. Secondly, it focuses on various CNN hardware acceleration architectures and their layers designed using different methods and their wide range of applications. Some of the recent design methods applied to various ranges of applications are also analyzed in the proposed survey. This detailed analysis gives an outlook on effective approximation blocks and the CNN architecture to be effectively used in various designs, with a scope of area in which future improvement can be made.
Volume: 14
Issue: 4
Page: 366-375
Publish at: 2025-12-01

Integration and optimization of grid through ANN-based solar MPPT and battery

10.11591/ijape.v14.i4.pp988-998
Kolli Sujran , Ankala Sirisha , Ganapaneni Swapna , Malligunta Kiran Kumar , Kambhampati Venkata Govardhan Rao
Integration of solar energy into the grid is the most important aspect for achieving sustainable energy systems. This paper presents an artificial neural network-based maximum power point tracking (ANN-MPPT) system with battery storage to enhance grid efficiency. The proposed ANN-MPPT is dynamically adapted to the varying irradiance and temperature, hence ensuring optimal power extraction from the photovoltaic system. Excess energy is stored in batteries during high solar radiation and discharged when solar generation is low or grid demand is high, maintaining a stable power supply. This system enhances the grid performance in terms of supporting real-time energy exchange, load balancing, and grid stability. Efficient management of the energy fluctuations ensures reliability even at times of grid failures. Further, integration of ANN-based MPPT with battery storage reduces dependence on non-renewable sources and harmonizes solar energy utilization. It can be achieved through enabling smarter energy management and thus contributing to the resilience and efficiency of a grid for better integration of renewable energies. The proposed system can tolerate fluctuating grid demands apart from supporting the features of smart grid, hence viable for increasing stability and sustainability in the grid.
Volume: 14
Issue: 4
Page: 988-998
Publish at: 2025-12-01

A hybrid transformer-graph neural networks framework for enhanced physical activity recognition and sedentary behavior analysis

10.11591/ijra.v14i4.pp429-438
Sudarsanam Anandanarayanan , Suvarnalingam Thirumaran
Sedentary behavior has been identified as a major risk factor for chronic diseases such as cardiovascular disorders, obesity, and diabetes. The accurate prediction of sedentary health risks is essential for early intervention and personalized healthcare strategies. This study proposes a novel machine learning-based predictive model that leverages transformer-based architectures and graph neural networks to analyze multidimensional behavioral data. Unlike traditional models, our approach incorporates temporal attention mechanisms to capture long-term dependencies in activity patterns and graph-based learning to model complex relationships between physiological and behavioral factors. The study utilizes real-world datasets, including wearable sensor data and self-reported activity logs, to train and validate the models. Experimental results demonstrate that the proposed framework outperforms conventional machine learning techniques such as random forest and XGBoost, achieving superior predictive accuracy and robustness. The findings highlight the potential of advanced machine learning algorithms in assessing sedentary health risks, enabling proactive health management and intervention strategies.
Volume: 14
Issue: 4
Page: 429-438
Publish at: 2025-12-01

Eco-friendly innovation: green energy empowered by IoT

10.11591/ijape.v14.i4.pp903-911
Nikita Amoli , Jitendra Singh , Rahul Mahala , Rajesh Singh , Anita Gehlot , Mahim Raj Gupta
Energy demand is high globally, impacting daily life and promoting sustainable modernization. Goal 9 aims to build an elastic framework for economies, while Goal 15 of the Sustainable Development Goals (SDGs) emphasizes the preservation of terrestrial environment, sustainable woodland management, and biodiversity conservation. The International Energy Agency predicts a significant increase in global renewable capacity, with solar PV being two-third of this growth. Green technology is crucial to combat global warming and Industry 4.0, a digital transformation that aims to create a strong framework for sustainable modernization. The growth of the smart grid is vital, involving energy sources, control techniques, computation, generation, transmission, distribution, and more. Supercapacitors store and deliver energy at high capacity, while green energy transforms fossil fuels into eco-friendly sources using natural resources like hydro, solar, wind, thermal, and biomass. This study explores the efficient use of microprocessors in solar and wind energy, as well as the application of actuators in the green energy sector. Green energy is a sustainable solution to increasing energy needs, reducing dependence on fossil fuels. IoT technologies, including sensors, actuators, microprocessors, and microcontrollers, are used in energy generation, transmission, distribution, and composition.
Volume: 14
Issue: 4
Page: 903-911
Publish at: 2025-12-01

Asymmetrical nine-level hybrid multilevel inverter design and analysis for electric vehicle applications

10.11591/ijape.v14.i4.pp1023-1034
Gerri Ratnaiah , Ramya Ganesan
A novel type of single-phase hybrid multilevel inverter (HMLI) is proposed in this paper. A hybrid system is made up of a multilevel inverter coupled to an H-bridge unit and which can generate nine-level output. To synthesize an output voltage waveform with nine steps, this setup uses merely seven power switches, two diodes, and two DC supplies. A greater number of steps were achieved in output voltage through suggested circuit with a smaller number of components than other existing multilevel inverter (MLI) topologies. A finer output waveform that is closer to a sinusoidal shape is produced with less total harmonic distortion (THD) because of the greater number of steps in the output voltage. Furthermore, it prolongs the switches' lifetime and lowers the voltage stress across them, increasing reliability. In addition, the system produces fewer switches than necessary, resulting in lower power losses and increased efficiency. This guarantees the suggested system's small size and inexpensive cost. A comparison between the suggested topology and the most current MLI topologies has been conducted to highlight the key components of the proposed topology. The suggested topology has been controlled using three distinct controlling schemes are phase disposition-pulse width modulation (PD-PWM), phase opposition disposition-PWM (POD-PWM), and alternative phase opposition disposition-PWM (APOD-PWM).
Volume: 14
Issue: 4
Page: 1023-1034
Publish at: 2025-12-01

Analysis and implementation of computation offloading in fog architecture

10.11591/ijra.v14i4.pp479-492
Prince Gupta , Rajeev Sharma , Sachi Gupta , Adesh Kumar
The fast expansion of connected devices has led to an unparalleled increase in data across sectors like industrial automation, social media, environmental monitoring, and life sciences. The processing of this data presents difficulties owing to its magnitude, temporal urgency, and security stipulations. Computation offloading has arisen as a viable alternative, allowing resource-constrained devices to assign demanding work to more robust platforms, thus improving responsiveness and efficiency. This paper examines decision-making strategies for computing offloading by assessing various algorithms, including a deep neural network with deep reinforcement learning (DNN-DRL), coordinate descent (baseline), AdaBoost, and K-nearest neighbor (KNN). The performance evaluation centers on three primary metrics: system accuracy, training duration, and latency. The computation offloading mitigates these issues by transferring intricate workloads from resource-limited devices to more proficient platforms, thus enhancing efficiency and responsiveness. The evaluation examines accuracy, training duration, and latency as key parameters. The results indicate that KNN attains maximum accuracy and minimal latency, AdaBoost provides a robust balance despite increased training costs, and the baseline underperforms in both efficiency and responsiveness. These findings underscore the trade-offs between computational expense, precision, and real-time application, providing insights for forthcoming IoT and edge-computing systems.
Volume: 14
Issue: 4
Page: 479-492
Publish at: 2025-12-01

A method integral sliding mode control to minimize chattering in sliding mode control of robot manipulator

10.11591/ijra.v14i4.pp345-355
Mai Hoang Nguyen , Truc Thi Kim Nguyen
This paper presents an improved sliding mode control (SMC) strategy for robotic manipulators by introducing a novel exponential integral-based adaptive gain law, referred to as integral sliding mode control (ISMC). The proposed approach dynamically adjusts the switching gain KKK in real-time, based on the accumulated system error, thereby effectively reducing chattering while preserving system robustness. Unlike many existing methods, the ISMC strategy eliminates the need for state observers or complex estimation techniques, simplifying implementation. Theoretical analysis is provided using Lyapunov stability theory, ensuring global convergence. Simulation results on 2-DOF and 3-DOF robotic arms demonstrate superior tracking accuracy and smoother control signals compared to conventional SMC approaches. This work contributes a lightweight yet effective SMC enhancement with practical benefits for real-world robotic applications.
Volume: 14
Issue: 4
Page: 345-355
Publish at: 2025-12-01

Performance enhancement using sensor and sensorless control techniques for a modified bridgeless Ćuk converter-based BLDC motor in EV applications

10.11591/ijape.v14.i4.pp769-782
W. Margaret Amutha , S. Premalatha , M. Karthikeyan
This work proposes a solar photovoltaic (PV)-powered, modified bridgeless Ćuk converter tailored for electric vehicle applications. It overcomes limitations such as high ripple, reduced power density, significant switching losses, and complex circuit structures in traditional designs. The system integrates a boost converter with a bridgeless Ćuk topology to ensure a reliable and efficient direct current (DC) power output. Performance evaluation includes sensor-based and sensorless speed control techniques-pulse width modulation (PWM), proportional integral derivative (PID), back electromotive force (EMF), and spider controllers-under both no-load and full-load scenarios. Key parameters such as rise time, overshoot, settling time, and steady-state error are analyzed. MATLAB/Simulink simulations indicate that the spider controller delivers superior dynamic behavior and stability. A 48 W, 1500 rpm hardware prototype confirms the simulation outcomes, demonstrating the practical viability and effectiveness of the proposed converter.
Volume: 14
Issue: 4
Page: 769-782
Publish at: 2025-12-01

A hybrid one step voltage-adjustable transformerless inverter for a one-phase grid incorporation of wind and solar power

10.11591/ijape.v14.i4.pp951-959
Bonigala Ramesh , Madhubabu Thiruveedula , Rahul Inumula , C. Poojitha Reddy , Mohammad Abdul Khadar , K. Sri Sai Hareesh
This paper presents a hybrid one-step voltage-adjustable transformerless inverter designed to efficiently integrate both solar photovoltaic (PV) and wind energy sources into a single-phase grid. The primary objective is to enhance power conversion efficiency while minimizing system complexity and cost. The proposed architecture combines a buck-boost DC-DC converter with a full-bridge inverter in a compact and modular design, enabling voltage regulation across a wide input range typical of hybrid renewable systems. By grounding the PV negative terminal, the system effectively eliminates leakage currents and ensures compliance with IEEE harmonic standards. The inverter operates with reduced switching losses and supports multiple operational modes tailored for variable solar and wind conditions. Simulation of a 300 W prototype demonstrates reliable performance, achieving a total harmonic distortion (THD) below 1%, validating its compatibility with grid requirements. Key contributions include the development of a unified topology for hybrid energy sources, in-depth analysis of energy storage components, and implementation of efficient modulation strategies. This work addresses significant challenges in renewable energy integration and provides a scalable solution for next-generation grid-connected hybrid power systems.
Volume: 14
Issue: 4
Page: 951-959
Publish at: 2025-12-01

Design and implementation of QUADRESCUE: A ROS-based quadruped robot for disaster response support

10.11591/ijra.v14i4.pp387-398
Sanjay Deshmukh , Ojas Chanakya , Om Gabani , Kashish Patni , Asmita Deshmukh
Search and rescue (SAR) operations in hazardous environments demand robotic systems capable of traversing complex terrains while ensuring responder safety. Traditional wheeled platforms often fail in debris-laden areas, and fully autonomous quadrupeds remain financially out of reach for many rescue agencies. This paper presents the design and development of QUADRESCUE, a modular operator-assisted quadruped robot built to bridge the gap between affordability and capability in disaster response. QUADRESCUE delivers core SAR functionalities including remote visual inspection, real-time terrain mapping via an RGB-D camera, payload transport, and GPS-based survivor localization. Built with a robust three degrees of freedom (3DoF) per leg design, the robot uses inverse kinematics algorithms to precisely control twelve servo motors for stable locomotion across uneven terrain. The system integrates the robot operating system (ROS) for seamless operation, real-time joystick control for easy navigation, an IMU for orientation sensing, and a GPS module with 3-meter accuracy. Field evaluations demonstrate 80–94% success rates on challenging surfaces, substantially outperforming wheeled counterparts 19% to 39% with a 200-meter control range and 45 minutes of runtime. QUADRESCUE offers a lightweight, cost-effective, and repairable solution that combines practical usability with advanced performance, making it well-suited for real-world deployment in emergency rescue situations.
Volume: 14
Issue: 4
Page: 387-398
Publish at: 2025-12-01

SHIELD: Security based hybrid autonomous deep learning network for load balancing in cloud

10.11591/ijra.v14i4.pp439-449
Loga Priyadarshini Kathirmalaiyan , Nithya Muthu
Load balancing in the Internet of Things (IoT) enhances the efficiency of the system by dynamically allocating tasks across devices and cloud resources. However, task scheduling struggles with unpredictable tasks, scalability, security risks, and unauthorized access control. To overcome these limitations, a novel security-based hybrid autonomous deep learning network for load balancing in cloud (SHIELD) framework has been proposed for secure task scheduling in cloud resources. Initially, the data received from the IoT devices is passed under certain security constraints to ensure the authenticity of the data. These privacy-preserved data are fed to the task scheduling module, which is employed by the dual DL Network to generate a schedule for resource management. Finally, cloud resources employ optimal allocation of tasks based on the generated schedule to ensure secure load balancing. The proposed framework is simulated by using Cloud Simulator 7G (CloudSim7G). The SHIELD framework is assessed by such metrics, including accuracy, recall, precision, F1-score, and specificity. In comparison, the proposed SHIELD framework achieves a privacy overhead of 14% outperforms the existing QODA-LB, Best-KFF, SPSO-TCS, and VMMISD techniques by achieving 10%, 11%, 12%, and 13% respectively.
Volume: 14
Issue: 4
Page: 439-449
Publish at: 2025-12-01

Effect on saturated and unsaturated fatty acids on various vegetable oils on droplet combustion characteristic

10.11591/ijape.v14.i4.pp980-987
Dony Perdana , Muhamad Nur Rohman , Mochamad Choifin
Vegetable oils have composed of triglycerides, which one consist of 3 fatty acids combined with glycerol. Each saturated and unsaturated fatty acid has a different effect on burning characteristics. This study aimed to investigated effect of fatty acids at ceiba pentandra and jatropha oils on the flame behavior of the droplet combustion process. The combustion characteristic was observed by an ignited droplet at the junction using a thermocouple and a high-speed camera (120 fps). Results showed that a higher saturated fatty acid content resulted in long-life and steady flames. This is because more oleic and linoleic acid carbon atoms leave the droplet area and react with air. Jatropha oil produces a higher temperature of 780 °C than ceiba pentandra oil. Temperature of a vegetable oils flame is influenced by number of carbon chains, double bond, and heating value. Ceiba pentandra oil has a higher burning rate of 0.185 mm/s than jatropha oil at 0.155 mm/s. The chain content of polyunsaturated fatty acids has significant effect on rate of combustion, which is due to the weak van der Waals dispersion forces, such that heat absorption is more active and energetic. The highest flame height for ceiba pentandra oil is 55.03 mm compared to for jatropha oil it is 46.82 mm. Long-chain unsaturated double bonds and glycerol cause micro-explosions. This micro-explosion caused the shape of the flame to split and expand so that evaporation occurred faster, thus increasing the size of the flame.
Volume: 14
Issue: 4
Page: 980-987
Publish at: 2025-12-01
1 2 3
...
2097 2098
Show 194 of 2098

Discover Our Library

Embark on a journey through our expansive collection of articles and let curiosity lead your path to innovation.

Explore Now
Library 3D Ilustration