A multi-stage LSTM federated forecasting method for multi-loads under multi-time scales

Expert Systems with Applications, Journal Year: 2024, Volume and Issue: 253, P. 124303 - 124303

Published: May 30, 2024

Language: Английский

---

Model predictive controller based design for energy optimization of the hybrid shipboard microgrids

Ocean Engineering, Journal Year: 2025, Volume and Issue: 323, P. 120545 - 120545

Published: Feb. 8, 2025

Language: Английский

---

Monitoring Daily Activities in Households by Means of Energy Consumption Measurements from Smart Meters

Journal of Sensor and Actuator Networks, Journal Year: 2025, Volume and Issue: 14(2), P. 25 - 25

Published: Feb. 27, 2025

Non-Intrusive Load Monitoring (NILM) includes a set of methods orientated to disaggregating the power consumption household per appliance. It is commonly based on single metering point, typically smart meter at entry electrical grid building, where signals interest, such as voltage or current, can be measured and analyzed in order disaggregate identify which appliance turned on/off any time. Although this information key for further applications linked energy efficiency management, it may also applied social health contexts. Since activation appliances related certain daily activities carried out by corresponding tenants, NILM techniques are interesting design remote monitoring systems that enhance development novel feasible healthcare models. Therefore, these foster independent living elderly and/or cognitively impaired people their own homes, while relatives caregivers have access additional about person’s routines. In context, work describes an intelligent solution deep neural networks, able household, starting from disaggregated provided commercial meter. With identified, usage patterns behaviour monitored long term after training period. way, every new day assessed statistically, thus providing score how similar routines learned during interval. The proposal has been experimentally validated means two commercially available monitors installed real houses tenants followed routines, well using well-known database UK-DALE.

Language: Английский

---

Smart Home Energy Prediction Framework using Temporal Kolmogorov-Arnold Transformer

Energy and Buildings, Journal Year: 2025, Volume and Issue: unknown, P. 115529 - 115529

Published: March 1, 2025

Language: Английский

---

Real-time monitoring system for evaluating the operational quality of rice transplanters

Computers and Electronics in Agriculture, Journal Year: 2025, Volume and Issue: 234, P. 110204 - 110204

Published: March 11, 2025

Language: Английский

---

Production prediction of pumping wells based on multi-mode transfer learning

Journal of Physics Conference Series, Journal Year: 2025, Volume and Issue: 2975(1), P. 012009 - 012009

Published: March 1, 2025

Abstract In the modern petroleum industry, it is difficult to establish an intelligent oil well production prediction model due limited number of samples for actual operating conditions and uneven distribution between different conditions. To address this issue, article proposes a new method predicting production. Firstly, by analyzing working process pumping well, key parameters describing fault are proposed, dynamic simulation lifting unit under established; Then, grey wolf optimizer (GWO) algorithm used optimize parameters, so that can adapt commonalities various Finally, target pre trained using source domain model, small sample data from added optimization accurately predict electric power The experimental results show simulate production, providing important reference applying artificial intelligence technology traditional energy industry.

Language: Английский

---

Artificial intelligence in smart homes: innovative approaches and application opportunities

Published: April 27, 2025

Artificial intelligence-supported smart home technologies are evolving rapidly, offering users enhanced living standards. This article analyzes the current state of AI-integrated homes, exploring both academic literature and product applications on market. The primary goal is to understand technological development trends how theoretical research aligns with real-world products. explains AI enhances automation, management, human-robot interaction in homes. results indicate a delay between advancements market implementations, suggesting that AI-powered systems will become more widespread near future.

Language: Английский

---

Intelligent modeling and analysis of hybrid organic Rankine plants: Data-driven insights into thermodynamic efficiency and economic viability

Engineering Applications of Artificial Intelligence, Journal Year: 2025, Volume and Issue: 143, P. 109946 - 109946

Published: Jan. 8, 2025

Language: Английский

---

Deep learning technology: enabling safe communication via the internet of things

Frontiers in Communications and Networks, Journal Year: 2025, Volume and Issue: 6

Published: Feb. 4, 2025

Introduction The Internet of Things (IoT) is a new technology that connects billions devices. Despite offering many advantages, the diversified architecture and wide connectivity IoT make it vulnerable to various cyberattacks, potentially leading data breaches financial loss. Preventing such attacks on ecosystem essential ensuring its security. Methods This paper introduces software-defined network (SDN)-enabled solution for vulnerability discovery in systems, leveraging deep learning. Specifically, Cuda-deep neural (Cu-DNN), Cuda-bidirectional long short-term memory (Cu-BLSTM), Cuda-gated recurrent unit (Cu-DNNGRU) classifiers are utilized effective threat detection. approach includes 10-fold cross-validation process ensure impartiality findings. most recent publicly available CICIDS2021 dataset was used train hybrid model. Results proposed method achieves an impressive recall rate 99.96% accuracy 99.87%, demonstrating effectiveness. model also compared benchmark classifiers, including Cuda-Deep Neural Network, Cuda-Gated Recurrent Unit, (Cu-DNNLSTM Cu-GRULSTM). Discussion Our technique outperforms existing based evaluation criteria as F1-score, speed efficiency, accuracy, precision. shows strength detection highlights potential combining SDN with learning assessment.

Language: Английский

---

Barriers to the adoption of energy management systems in residential buildings

Facilities, Journal Year: 2024, Volume and Issue: 42(15/16), P. 107 - 125

Published: July 30, 2024

Purpose This study aims to examine the challenges in implementation of energy management systems residential buildings lower running cost and achieve a better energy-efficient building. Design/methodology/approach adopted mixed research method. Quantitative data was gathered by issuing questionnaire 20 Delphi experts, while qualitative acquired through Systematic Literature Review. Data received analyzed using descriptive analysis Findings The findings revealed that main barriers incorporating (EMSs) consist lack awareness systems, commitment management, knowledge about funds for resistance technology property owners managers, distrust owners, high initial technologies, shortage technicians nonexistence local manufacturers equipment, incentives efficient repair costs technologies. Research limitations/implications specific focus on may limit applicability commercial or industrial sectors. Further is warranted accommodate other energy-consuming Practical implications People’s perceptions, either wrong correct, affect their ability make an informed decision adopt denying them opportunity reap associated benefits. Therefore, there urgent need industry stakeholders government increase educational opportunities managers tenants importance systems. Originality/value presents potential obstacles problematic areas residents encounter these Consequently, they will be able well-informed choice when installing Moreover, elucidates identification novel perspectives also unexamined impede widespread use buildings.

Language: Английский

---