A Review on Machine/Deep Learning Techniques Applied to Building Energy Simulation, Optimization and Management

Thermo, Год журнала: 2024, Номер 4(1), С. 100 - 139

Опубликована: Март 6, 2024

Given the climate change in recent decades and ever-increasing energy consumption building sector, research is widely focused on green revolution ecological transition of buildings. In this regard, artificial intelligence can be a precious tool to simulate optimize performance, as shown by plethora studies. Accordingly, paper provides review more than 70 articles from years, i.e., mostly 2018 2023, about applications machine/deep learning (ML/DL) forecasting performance buildings their simulation/control/optimization. This was conducted using SCOPUS database with keywords “buildings”, “energy”, “machine learning” “deep selecting papers addressing following applications: design/retrofit optimization, prediction, control/management heating/cooling systems renewable source systems, and/or fault detection. Notably, discusses main differences between ML DL techniques, showing examples use The aim group most frequent ML/DL techniques used field highlighting potentiality limitations each one, both fundamental aspects for future approaches considered are decision trees/random forest, naive Bayes, support vector machines, Kriging method neural networks. investigated convolutional recursive networks, long short-term memory gated recurrent units. Firstly, various explained divided based methodology. Secondly, grouping aforementioned occurs. It emerges that efficiency issues while management systems.

Язык: Английский

---

Detection and analysis of deteriorated areas in solar PV modules using unsupervised sensing algorithms and 3D augmented reality

Heliyon, Год журнала: 2024, Номер 10(6), С. e27973 - e27973

Опубликована: Март 1, 2024

Solar Photovoltaic (PV) systems are increasingly vital for enhancing energy security worldwide. However, their efficiency and power output can be significantly reduced by hotspots snail trails, predominantly caused cracks in PV modules. This article introduces a novel methodology the automatic segmentation analysis of such anomalies, utilizing unsupervised sensing algorithms coupled with 3D Augmented Reality (AR) enhanced visualization. The outperforms existing techniques, including Weka Meta Segment Anything Model (SAM), as demonstrated through computer simulations. These simulations were conducted using Cali-Thermal Panels Panel Infrared Image Datasets, evaluation metrics Jaccard Index, Dice Coefficient, Precision, Recall, achieving scores 0.76, 0.82, 0.90, 0.99, respectively. By integrating drone technology, proposed approach aims to revolutionize maintenance facilitating real-time, automated solar panel detection. advancement promises substantial cost reductions, heightened production, improved performance installations. Furthermore, innovative integration AR visualization opens new avenues future research development field maintenance.

Язык: Английский

---

Uncovering the Potential of Indoor Localization: Role of Deep and Transfer Learning

IEEE Access, Год журнала: 2024, Номер 12, С. 73980 - 74010

Опубликована: Янв. 1, 2024

Indoor localization (IL) is a significant topic of study with several practical applications, particularly in the context Internet Things (IoT) and smart cities. The area IL has evolved greatly recent years due to introduction numerous technologies such as WiFi, Bluetooth, cameras, other sensors. Despite growing interest this field, there are challenges drawbacks that must be addressed develop more accurate sustainable systems for IL. This review gives an in-depth look into IL, covering most promising artificial intelligence-based hybrid strategies have shown excellent potential overcoming some limitations classic methods within IoT environments. In addition, paper investigates significance high-quality datasets evaluation metrics design assessment algorithms. Furthermore, overview emphasizes crucial role machine learning techniques, deep transfer learning, play advancement A focus on importance various technologies, methods, techniques being used improve it. Finally, survey highlights need continued research development create scalable can applied across range IoT-related industries, evacuation-egress routes, hazard-crime detection, occupancy-driven energy reduction asset tracking management.

Язык: Английский

---

Continual learning for energy management systems: A review of methods and applications, and a case study

Applied Energy, Год журнала: 2025, Номер 384, С. 125458 - 125458

Опубликована: Фев. 10, 2025

Язык: Английский

---

Digital Twins and Control Theory: A Critical Review on Revolutionizing Quadrotor UAVs

IEEE Access, Год журнала: 2024, Номер 12, С. 43291 - 43307

Опубликована: Янв. 1, 2024

This work explores the crucial roles that control theory and digital twins play in enhancing performance of underactuated quadrotor unmanned aerial vehicles (QUAVs). It describes how novel idea combined with could alter operations. Some basic ideas, such as UAV model, various schemes, innovative techniques to improve autonomy QUAV missions dynamic circumstances, may also be interest readers. highlights recent developments presents a game-changing combining twin computer vision, amalgamating artificial intelligence internet things like elements sensing perception better for autonomous flight control, human-UAV interaction, energy-efficient flight, swarming UAVs. The reader finally find suggestions applying understandings incorporating technology boost its revolutionary potential.

Язык: Английский

---

Anomaly detection in time-series data using evolutionary neural architecture search with non-differentiable functions

Applied Soft Computing, Год журнала: 2024, Номер 155, С. 111442 - 111442

Опубликована: Март 1, 2024

Deep neural networks have become the benchmark in diverse fields such as energy consumption forecasting, speech recognition, and anomaly detection, owing to their ability efficiently process analyze data. However, they face challenges managing complexity variability time series data, often leading increased model prolonged search duration during parameter tuning. This paper proposes a novel detection approach through evolutionary architecture (AD-ENAS), which is specifically designed for The proposed focuses on optimal minimal network architecture. AD-ENAS method consists of two main phases: evolution weight adjustment. phase highlights importance by evaluating fitness each agent using shared values. Subsequently, convolutional matrix adaptation technique used next adjustment network. operates without relying differentiable functions, thus expanding scope design beyond traditional backpropagation-based approaches. Various non-differentiable loss functions are explored facilitate effective Comparative experiments conducted with five baseline methods three well-known datasets from reputable sources NASA SMAP, MSL Yahoo S5-A1. results demonstrate that effectively evolves architectures, outperforming F1 scores across (MSL: 0.942, SMAP: 0.961, S5-A1: 0.988) showcasing its efficacy detecting anomalies

Язык: Английский

---

An intelligent geographic information system-based framework for energy efficient street lighting

Signal Image and Video Processing, Год журнала: 2025, Номер 19(4)

Опубликована: Фев. 19, 2025

Язык: Английский

---

Integration of IoT Devices with UAV Swarms

Apress eBooks, Год журнала: 2025, Номер unknown, С. 591 - 631

Опубликована: Янв. 1, 2025

Язык: Английский

---

Secure and transparent energy management using blockchain and machine learning anomaly detection: A case study of the Ausgrid dataset

Computers & Industrial Engineering, Год журнала: 2025, Номер unknown, С. 111040 - 111040

Опубликована: Март 1, 2025

Язык: Английский

---

Machine learning-based inertia estimation in power systems: a review of methods and challenges

Energy Informatics, Год журнала: 2025, Номер 8(1)

Опубликована: Апрель 30, 2025

Язык: Английский

---