International Journal of Electrical Engineering and Computing

Editorial

Nataša Popović — Tue, 24 Dec 2024 18:35:16 +0000

Impressum

Nataša Popović — Tue, 24 Dec 2024 18:33:39 +0000

Lightning protection of the 110 kV substation at the wind park Ivan Sedlo against transients caused by direct lightning strikes into connected overhead line

Mladen Banjanin, Velibor Crnovčić, Mihael Tomšič — Wed, 25 Dec 2024 09:51:24 +0000

This paper deals with the lightning protection of 110 kV substation at the wind park Ivan Sedlo against transients caused by direct lightning strikes into connected 110 kV overhead line. Lightning protection level (LPL) I as per standard IEC 62305-1 is selected for this substation to reduce probability of equipment failure to low level. Calculations are done in the EMTP-ATP software. Three critical scenarios are analysed: shielding failure at the connected 110 kV overhead line, lightning strike to the top of first tower in front of the substation and lightning strike to the top of gantry tower at the entrance to the substation. It is presented that LPL I for analysed configuration of substation can be achieved by installing two sets of surge arresters. First set of surge arresters is installed in the transmission line bay, while second set of surge arresters is installed in the transformer bay. Transferred lightning transients through the power transformer are also calculated. It is presented that one additional set of surge arresters is necessary at the 30 kV side of the power transformer, to protected it against transferred lightning transients. Energy duty of all three sets of surge arresters is calculated. At the 110 kV voltage level energy stress of surge arresters is significant and strongly depends on the substation grounding resistance value. Estimated energy stress of set of surge arresters at the 30 kV voltage level is very low.

MQTT Protocol in Smart Home Environments: Principles of Operation and Application

Tijana Begović, Vasilije Čabarkapa, Milica Ivković, Božidar Popović — Wed, 25 Dec 2024 09:59:01 +0000

Internet of Things as a revolutionary concept in technology of modern days has led to the popularization of “smart” things. It introduces low-cost and reliable solutions to provide independent operations of devices and machines. Smart home systems have become one of the most important implementations as they significantly improve the quality of people’s lives. Since communication plays a key role in these systems, this article provides insights into Message Queuing Telemetry Transport protocol application in smart home systems. Although it is not the only protocol used, MQTT stands out in terms of reliability, efficiency, and simplicity. The article provides major MQTT characteristics, advantages and its application in a particular smart home system for ambient parameters monitoring and control. The system integrates NodeMCU based sensor node with mobile and web applications via MQTT protocol. Results of the performance analysis of MQTT use in this system are given. Tests have shown great connection establishment speed, a high percentage of successfully sent messages, and a large number of messages sent per minute. Thanks to the flexibility of the MQTT protocol, the system is easy to upgrade and expand.

Optimization and Construction of a Deep Learning Model for Breast Cancer Segmentation

Luka Glišić, Ivana Berković, Biljana Radulović — Wed, 25 Dec 2024 10:03:52 +0000

The research focuses on optimizing and developing a breast cancer segmentation model utilizing multi-GPU arrays to maximize hardware resource utilization. By exploring various architectural strategies, the study aims to enhance resource allocation efficiency despite existing limitations. Detailed evaluations using mammography datasets have demonstrated significant improvements in tumor detection capabilities. This technology holds the potential to revolutionize breast cancer detection, a critical advancement given the global impact of the disease. Training data analysis confirms the scalability of these results across diverse hardware configurations, ensuring high efficiency and reliability. The study employs modern architectures, contributing valuable insights to the field of breast cancer segmentation and advancing medical imaging technologies. The development of efficient and clinically viable solutions is imperative in contemporary medical image analysis. While state-of-the-art deep learning architectures offer impressive capabilities, their substantial computational demands pose barriers to widespread clinical adoption. This research addresses the need for solutions that efficiently process large datasets while maintaining diagnostic accuracy, facilitating integration into clinical workflows and reducing operational costs.

Cost vs. Performance: Raspberry Pi and ESP32 in Face Detection Tasks

Marko Malović, Elma Avdić, Nikola Davidović — Wed, 25 Dec 2024 10:07:16 +0000

A cost-effective, scalable solution that leverages low-cost embedded systems for computationally intensive tasks like face detection is critical for the growing demands of IoT applications, motivating the development of innovative hybrid architectures. This paper proposes such a framework, combining the ESP32 CAM as a distributed image capture unit with light preprocessing and the Raspberry Pi as a centralized processing unit. By addressing the limitations of standalone implementations—low computational capability of the ESP32 CAM and high deployment costs of the Raspberry Pi—our framework achieves a balance between affordability and performance. We present the analysis of the devices, evaluating metrics such as frame rate, detection accuracy, and cost-efficiency. The results highlight the potential of the hybrid system to significantly lower costs while enabling scalable, real-time face detection in IoT scenarios. This study contributes to the ongoing research by proposing an adaptable, resource-optimized framework suitable for diverse use cases, from smart surveillance to retail monitoring, paving the way for more efficient IoT-based vision systems.

A System for Remotely Monitoring and Controlling Temperature within an Egg Artificial Incubator

Nataša Popović, Božidar Popović — Wed, 25 Dec 2024 10:10:47 +0000

Abstract— According to a report by the Food and Agriculture Organization of the United Nations, poultry meat is the leading contributor to global growth in total meat production. This growth is facilitated by the short production cycle of poultry, advancements in genetics, animal healthcare, feeding methods, breeding techniques, and improved environmental conditions, enabling producers to adapt quickly to market demands. Proper hatching practices are critical to maximizing the number of healthy chicks and minimizing their mortality rate. There are two primary methods of hatching: natural incubation, where hens incubate eggs by sitting on them, and artificial incubation, which uses incubators. In artificial incubation, monitoring and controlling key environmental factors, particularly temperature and humidity, are essential. Maintaining the incubator's temperature within an optimal range is crucial, with literature recommending a range of 37°C to 38°C and an ideal temperature of 37.8°C. This paper introduces a system for remotely monitoring and controlling temperature within an artificial incubator. The system integrates hardware and software based on the Internet of Things (IoT) framework. For this purpose, a proportional-integral-derivative (PID) controller is designed to maintain the desired temperature, with temperature data stored in a database and accessible online. The proposed system offers a cost-effective yet highly efficient solution, primarily designed for small-scale hatcheries. With modifications, it can be adapted for broader agricultural applications, including greenhouses, silos, and stables. By contributing to digital agriculture, the system supports sustainable farming practices, which are essential for the future of agriculture.