This paper presents a proactive blockage detection algorithm and performance for sub-terahertz (sub-THz) communication systems. It is a critical technology for next-generation wireless networks. Human body blockage is a significant challenge for millimeter wave and sub-THz systems.
Read more
Recent advancements in artificial intelligence and computer vision have led to the automation of many conventional surveillance techniques, especially in smart city applications. However, object detection systems often struggle in poorly lit environments and may suffer from slow processing speeds. To address these restrictions, this paper proposes an adaptive illumination correction technique using an enhanced Logarithmic Image Processing model. The method improves object visibility in low-light video frames and enhances detection performance. Additionally, a customized deep convolutional neural network is developed to accurately detect objects after applying the illumination correction. The combined framework is evaluated on standard datasets and demonstrates superior robustness to varying illumination conditions compared to existing state-of-the-art methods. The results show significant improvements in metrics such as accuracy, recall, precision, and F-measure, proving the effectiveness of the proposed approach.
Read more
The essential parts of the circuit breaker working system is the relay which is susceptible to performance deterioration when operating in extreme conditions like salt spray. This can eventually impair the circuit breakers functionality and compromise the power system's ability to function effectively and consistently. This study used an existing test structure to conduct rapidly deteriorating tests on a specific design of additional switch under salt spray in order to assess the relay's lifetime condition effectively. As distinctive parameters impacted by salt spray, contact resistance, pickup voltage, release voltage, pickup time, release time and coil resistance of the relay were examined. The results demonstrated that the presence of salt spray reduced the coil effectiveness of the relay, necessitating a higher operating voltage in order to supply the electromagnetic force required for functioning. Moreover, actual and predicted lifetime from individual parameter is calculated. However, coil deterioration also led to a reduction in the electromagnetic force that the step voltage produced, which lengthened the operating period. A Random Forest algorithm model was then created that was appropriate for determining the relay status in this investigation. The existing Gradient Boosting, XGBoost algorithm is compared with proposed method to analyze the performance metrics such as accuracy, precision, recall, F1-score, R2 score, specificity, sensitivity, Root Mean Square Error (RMSE) and Mean Absolute Error (MAE) is determined.
Read more
This paper presents an experimental exploration of position control techniques applied to a Rotary Motion Servo Plant (SRV02), employing output feedback controllers. The transfer function model of the SRV02 is derived through first principles. Output feedback controllers, specifically Position Velocity (PV) and Multirate Output Feedback (MOF), are formulated and designed. The study evaluates the efficacy of these controllers in real-world scenarios for positioning control of the SRV02. A comparative analysis between PV and MOF controllers was conducted, assessing their static and dynamic characteristics. The results, obtained from both simulation and experimental setups, illustrate the superiority of MOF over PV controller in achieving precise position control of the SRV02. The findings of this study not only validate the effectiveness of MOF in position control applications but also provide insights into the practical implementation of output feedback controllers for enhancing the performance of rotary motion servo systems like SRV02.
Read more
Accurate and efficient road obstacle detection is crucial for ensuring the safety and reliability of autonomous and assisted driving systems. However, current object detection models, including YOLOv8, often encounter difficulties in accurately detecting small obstacles and balancing detection precision with inference speed. To address these challenges, we propose an improved YOLOv8-based detection framework featuring three key enhancements: (1) the backbone is redesigned using RepVGG modules to accelerate inference through structural re-parameterization without compromising feature representation; (2) an additional small-object detection head at the P2 feature level is introduced to significantly enhance sensitivity towards small-scale obstacles; (3) a Swin Transformer module is integrated into the final stage of the backbone to improve global contextual understanding and semantic feature extraction. Extensive experiments conducted on a comprehensive, multi-source road obstacle dataset demonstrate that our proposed model achieves superior performance with a precision of 0.812, recall of 0.782, mAP@0.5 of 0.822, and mAP@0.5:0.95 of 0.588. Additionally, it maintains a real-time inference speed of 180 FPS on an NVIDIA H800 GPU. The results confirm the efficacy of our approach, particularly in enhancing small object detection in complex real-world traffic scenarios.
Read more
It is a real challenge to bring reliable and stable power to remote locations, critical facilities such as telecommunication towers, because it is costly and challenging to expand the conventional power distribution networks. In this paper, a new power supply system is proposed that can overcome this problem by exploiting common overhead ground wire (OHGW) energy transmission lines that run high-voltage (HV). The principle of action is that the system isolates a portion of the OHGW and applies an alternating current (AC) voltage to it by capacitively coupling it to the phase conductors of a transmission line. To take an example, a 4-km span of OHGW on a 735 kV line can produce a maximum power of 25 kW of power which is modelled as a 52.4 kV Thevenin source with the capacitive impedance as 26nF. The main problem is to achieve a stable output voltage (within 5%) across a large load range (0 to 25 kW) without the need for complex control systems, power electronics and so on. To obtain good regulation, this problem is solved using a mix of passive components, nonlinear inductances, and a unique magnetic component, IVACE. Simulation studies have proved that the system could sustain a load voltage of about 62 kV, and the system would be fault resistant and could recover fast with a load disturbance. Harmonic analysis: The value of the total harmonic distortion (THD) is 6.5 per cent at no load and 4.7 per cent at nominal load, with the third harmonics as the most important factor. This method has the advantage as compared to other means of power supply to remote regions, where expansion of the HV transmission network is not needed, and the method is a low-cost and low-maintenance method that can be implemented quickly.
Read more
The increasing demand for energy and environmental concerns necessitate the integration of Renewable Energy Sources (RES), particularly Photovoltaic (PV) systems, into modern power grids. This paper presents an efficient grid connected PV system utilizing a dual-phase Interleaved Boost Converter (IBC) and an intelligent Maximum Power Point Tracking (MPPT) approach based on Adaptive Neuro Fuzzy Inference System (ANFIS). To enhance the controller's performance, optimization algorithms Class Topper Optimization (CTO), Falcon Optimization Algorithm (FOA), and Pufferfish Optimization Algorithm (POA) are employed. The dual phase IBC ensures reduced ripple, improved voltage gain, and stable power conversion. Simulation results in MATLAB/Simulink demonstrate that the proposed system achieves 94% efficiency, 0.95% Total Harmonic Distortion (THD), and a maximum MPPT tracking efficiency of 97.05% using the POA-ANFIS controller.
Read more
In Wireless Sensor Networks (WSNs), sensor nodes with limited resources must manage energy consumption and network efficiency. This work proposes an AM-DRL framework that adjusts transmission power, duty cycling, and clustering methods in response to real-time network conditions. Our methodology implements Multi-Agent Reinforcement Learning (MARL) to allow distributed decision-making by individual sensor nodes, which improves network lifetime while ensuring QoS parameters such as latency, packet delivery ratio, and throughput are met. The framework steers learning on the path of the energy-efficient decisions by introducing a novel hybrid reward function that accounts for energy expenditure, network topology, and data garnered. Transfer learning allows the model to be applied to different WSN configurations with minimal retraining efforts. AM-DRL outperformed other tested methods at saving energy and enhancing network lifetime and data transmission, during extensive simulations and real-world trials, proving more efficient than energy-saving approaches using reinforcement learning, and clustering. This paper provides scalable and intelligent WSN energy optimization solutions for industrial IoT, environmental monitoring, and smart city infrastructure.
Read more
In Wireless Sensor Networks (WSNs), sensor nodes with limited resources must manage energy consumption and network efficiency. This work proposes an AM-DRL framework that adjusts transmission power, duty cycling, and clustering methods in response to real-time network conditions. Our methodology implements Multi-Agent Reinforcement Learning (MARL) to allow distributed decision-making by individual sensor nodes, which improves network lifetime while ensuring QoS parameters such as latency, packet delivery ratio, and throughput are met. The framework steers learning on the path of the energy-efficient decisions by introducing a novel hybrid reward function that accounts for energy expenditure, network topology, and data garnered. Transfer learning allows the model to be applied to different WSN configurations with minimal retraining efforts. AM-DRL outperformed other tested methods at saving energy and enhancing network lifetime and data transmission, during extensive simulations and real-world trials, proving more efficient than energy-saving approaches using reinforcement learning, and clustering. This paper provides scalable and intelligent WSN energy optimization solutions for industrial IoT, environmental monitoring, and smart city infrastructure.
Read more
Early detection of oil spills is crucial and essential for marine environments to minimize environmental harm and enable quick responsive measures. Oil spills can cause significant ecological and financial losses, which emphasizes the need for an efficient monitoring system. This paper presents the use of YOLO deep learning algorithms to enhance the oil spill detection speed and accuracy. A robust and high-quality dataset is taken, consisting of images extracted from Roboflow. To maximize the data quality, preprocessing techniques such as label normalization, contrast enhancement and noise reduction were used. The proposed YOLO algorithms were trained using Adam and SGDM optimizers with an initial learning rate of 0.01, 0.001 and 0.0001. Among the adopted YOLO models, YOLOv9 yielded impressive results with an mAP@0.5 of 94.45%, precision of 95.6%, recall of 93.3% and F1 Score of 94.44%. The recommended system, which incorporates deep learning technologies into marine environment monitoring, greatly improves the marine surveillance systems for oil spill detection and emergency response capabilities by enabling real-time monitoring.
Read more
The variability in output power from renewable energy sources (RES), load and system disturbances in autonomous microgrid (MG) can lead to significant frequency deviations. Traditional PID controllers often struggle to perform adequately under these challenging conditions. To address these issues, this paper recommends a multistage PID (MSPID) controller designed explicitly for microgrid frequency management. The parameters of this controller are optimized using a Bald Eagle Search (BES) algorithm. The controller is rigorously tested on an autonomous hybrid microgrid, considering various parametric and systemic disturbances. The results demonstrate notable improvements in error reduction, settling time, and overshoot, highlighting the effectiveness of the proposed controller compared to multilevel PID(MLPID) and traditional PID controllers found in existing literature.
Read more
The evolution of wireless communication systems has intensified the demand for compact, efficient antennas tailored to specific applications. This research introduces an S-shaped microstrip patch antenna for wireless communication, emphasizing its compactness and operational efficiency. The antenna designed and simulated using Computer Simulation Technology (CST) software, achieves a gain of 1.75 dBi, 95% radiation efficiency, an operational bandwidth of 200 MHz (2.37 GHz to 2.57 GHz), and an excellent impedance matching with compact dimensions of 29 mm × 20 mm optimized to achieve resonance within the desired frequency of 2.45 GHz and make the antenna suitable for space-constrained environments. Detailed analysis of return loss, voltage standing wave ratio (VSWR), and radiation characteristics are presented. The antenna is fabricated, and the simulated results were verified by practical measurements of a Vector Network Analyzer (VNA). The measured resonance frequency and return loss proved the appropriateness of the antenna to the desired ISM band of 2.45 GHz. Though some minor deviations were noted between the simulated and practical performance. Comparative results with recent antenna designs highlight the proposed S-shaped antenna's performance and potential for specific applications in constrained environments requiring highly efficient narrowband antennas.
Read more
This paper presents a dual-fed approach to the design of a wide bandwidth microstrip patch antenna (MPA) of patch dimensions (l × w = 42 mm × 37 mm) and substrate dimensions (l × w × h = 100 mm × 100 mm × 3.14 mm); featuring two symmetrically inverted U-slots of dimensions (l × w × h = 25.62 mm × 15.5 mm × 2 mm) and two coaxial probes (7 mm each) adjacent to each slot. The substrate material utilized is Rogers-RT/duroid-5880, exhibits a dielectric constant of 2.2, makes it extremely efficient in reducing lossy nature of generated output parameters. The antenna demonstrates broad -10 dB impedance bandwidth ≈ 44% from 2.55 GHz - 3.85 GHz centered at a frequency of 3 GHz within the S band (2-4 GHz) region. A radiation intensity of 12 dB is observed alongside impeccable efficiency of 99.98%.
Read more
In the context of heterogeneous cloud radio access networks (H-CRANs), there's a growing challenge in achieving efficient resource allocation due to the surge in data and complexities related to asynchrony in federated learning frameworks. This research presents an innovative methodology that blends federated averaging with asynchronous federated deep reinforcement learning (AF-DRL). By enabling individual agents to interact with distinct environments and subsequently adjust policy parameters, local updates are consolidated at a central point to generate a global update.
Read more