Hypertuned-YOLO for interpretable distribution power grid fault location based on EigenCAM

Ain Shams Engineering Journal, Год журнала: 2024, Номер 15(6), С. 102722 - 102722

Опубликована: Март 1, 2024

Ensuring the reliability of electrical distribution networks is a pressing concern, especially given power outages due to surface contamination on insulating components. Surface can elevate conductivity, thereby resulting in failures that lead shutdowns. Addressing this challenge, paper proposes an approach for real-time monitoring grids prevent such incidents. A hypertuned version you only look once (YOLO) model tailored application. We refine model's hyperparameters by integrating genetic algorithm maximize its detection performance. The EigenCAM technique enhances visual interpretability outcomes, providing operators with actionable insights maintenance and tasks. Benchmark tests reveal proposed Hypertuned-YOLO outperforms Detectron (Masked R-CNN), YOLOv5, YOLOv7 models. achieves F1-score 0.867 [email protected] 0.922, validating robustness efficacy.

Язык: Английский

---

Seq2Seq-LSTM With Attention for Electricity Load Forecasting in Brazil

IEEE Access, Год журнала: 2024, Номер 12, С. 30020 - 30029

Опубликована: Янв. 1, 2024

Electricity load forecasting is important to planning the decision-making regarding use of energy resources, in which power system must be operated guarantee supply electricity future at lowest possible price. With rise ability based on deep learning, these approaches are promising this context. Considering attention mechanism capture long-range dependencies, it highly recommended for sequential data processing, where time series-related tasks stand out. a sequence-to-sequence (Seq2Seq) series Brazil, paper proposes long short-term memory (LSTM) with perform forecasting. The proposed Seq2Seq-LSTM outperforms other well-established models. Having mean absolute error equal 0.3027 method shown field applications. contributes by implementing an Seq2Seq data, therefore, more than one correlated signal can used prediction enhancing its capacity when avaliable.

Язык: Английский

---

The Sustainability Concept: A Review Focusing on Energy

Sustainability, Год журнала: 2023, Номер 15(19), С. 14049 - 14049

Опубликована: Сен. 22, 2023

The concept of sustainability, with a focus on energy, has emerged as central tenet in addressing the mounting global challenges environmental degradation and resource depletion. Indicators sustainability focusing energy are crucial tools used to assess monitor progress toward achieving more sustainable system. These indicators provide valuable insights into environmental, social, economic dimensions practices their long-term impacts. By analyzing understanding these indicators, policymakers, businesses, communities can make informed decisions, formulate effective policies, steer efforts future. serve navigational guides, steering world that support both present needs well-being future generations. In this paper, measurement indexes reviewed, factors. discussion presented here is related an assessment possibilities for improving efficiency evaluating measure whether desired levels being achieved.

Язык: Английский

---

Real‐time defect detection method based on YOLO‐GSS at the edge end of a transmission line

IET Image Processing, Год журнала: 2024, Номер 18(5), С. 1315 - 1327

Опубликована: Янв. 12, 2024

Abstract Combining edge devices with intelligent inspection for transmission lines can fulfill the demand real‐time defect detection in field. However, there has been limited research on algorithms suitable low computational power and memory, existing primarily focuses CPU optimization. To address these issues, this paper proposes a method line endpoints based YOLO‐GSS (YOLOv8 Mosaic‐9, G‐GhostNet, S‐FPN, Spatial Intersection over Union (SIoU) modifications). First, authors improve input of YOLOv8 network using Mosaic‐9 to increase number features training phase enhance algorithm robustness. Next, introduce G‐GhostNet S‐FPN backbone neck sections while improving inference speed accuracy. Finally, modify Complete loss function SIoU further Experimental results demonstrate that compared original YOLOv8, proposed achieves 5x Nvidia Jetson NX 7.7% improvement accuracy, meeting requirements field inspections.

Язык: Английский

---

Bootstrap aggregation with Christiano–Fitzgerald random walk filter for fault prediction in power systems

Electrical Engineering, Год журнала: 2024, Номер 106(3), С. 3657 - 3670

Опубликована: Янв. 4, 2024

Язык: Английский

---

A cloud 15kV-HDPE insulator leakage current classification based improved particle swarm optimization and LSTM-CNN deep learning approach

Swarm and Evolutionary Computation, Год журнала: 2024, Номер 91, С. 101755 - 101755

Опубликована: Окт. 13, 2024

Язык: Английский

---

A Review of Automation and Sensors: Parameter Control of Thermal Treatments for Electrical Power Generation

Sensors, Год журнала: 2024, Номер 24(3), С. 967 - 967

Опубликована: Фев. 1, 2024

This review delves into the critical role of automation and sensor technologies in optimizing parameters for thermal treatments within electrical power generation. The demand efficient sustainable generation has led to a significant reliance on plants. However, ensuring precise control over these remains challenging, necessitating integration advanced systems. paper evaluates pivotal aspects automation, emphasizing its capacity streamline operations, enhance safety, optimize energy efficiency treatment processes. Additionally, it highlights indispensable sensors monitoring regulating crucial parameters, such as temperature, pressure, flow rates. These enable real-time data acquisition, facilitating immediate adjustments maintain optimal operating conditions prevent system failures. It explores recent technological advancements, including machine learning algorithms IoT integration, which have revolutionized capabilities control. Incorporating innovations significantly improved precision adaptability systems, resulting heightened performance reduced environmental impact. underscores imperative nature generation, their enhancing operational efficiency, reliability, advancing sustainability

Язык: Английский

---

A deep learning-based approach for axle counter in free-flow tolling systems

Scientific Reports, Год журнала: 2024, Номер 14(1)

Опубликована: Фев. 10, 2024

Abstract Enhancements in the structural and operational aspects of transportation are important for achieving high-quality mobility. Toll plazas commonly known as a potential bottleneck stretch, they tend to interfere with normality flow due charging points. Focusing on automation toll plazas, this research presents development an axle counter compose free-flow collection system. The is responsible interpretation images through algorithms based computer vision determine number axles vehicles crossing front camera. You Only Look Once (YOLO) model was employed first step identify vehicle wheels. Considering that several versions available, select best model, YOLOv5, YOLOv6, YOLOv7, YOLOv8 were compared. YOLOv5m achieved result precision recall 99.40% 98.20%, respectively. A passage manager developed thereafter verify when passes camera store corresponding frames. These frames then used by image reconstruction module which creates complete containing all axles. From sequence frames, proposed method able passing scene, count axles, automatically generate appropriate charge be applied vehicle.

Язык: Английский

---

Artificial intelligence approaches for accurate assessment of insulator cleanliness in high-voltage electrical systems

Electrical Engineering, Год журнала: 2024, Номер unknown

Опубликована: Авг. 27, 2024

Язык: Английский

---

Wavelet CNN‐LSTM time series forecasting of electricity power generation considering biomass thermal systems

IET Generation Transmission & Distribution, Год журнала: 2024, Номер 18(21), С. 3437 - 3451

Опубликована: Окт. 15, 2024

Abstract The use of biomass as a renewable energy source for electricity generation has gained attention due to its sustainability and environmental benefits. However, the intermittent demand poses challenges optimizing in thermal systems. Time series forecasting techniques are crucial addressing these by providing accurate predictions availability generation. Here, wavelet transform is applied denoising, convolutional neural networks (CNN) used extract features time series, long short‐term memory (LSTM) perform predictions. result mean absolute percentage error equal 0.0148 shows that CNN‐LSTM promising machine‐learning methodology forecasting. Additionally, this paper discusses importance model evaluation validation strategies assess performance models real‐world applications. major contribution related improving using hybrid method outperforms other based on deep learning. Finally, future research directions potential advancements systems outlined foster continued innovation sustainable

Язык: Английский

---