Data-driven predictive control for smart HVAC system in IoT-integrated buildings with time-series forecasting and reinforcement learning

Applied Energy, Год журнала: 2023, Номер 338, С. 120936 - 120936

Опубликована: Март 8, 2023

Язык: Английский

---

BMAE-Net: A Data-Driven Weather Prediction Network for Smart Agriculture

Agronomy, Год журнала: 2023, Номер 13(3), С. 625 - 625

Опубликована: Фев. 22, 2023

Weather is an essential component of natural resources that affects agricultural production and plays a decisive role in deciding the type production, planting structure, crop quality, etc. In field agriculture, medium- long-term predictions temperature humidity are vital for guiding activities improving yield quality. However, existing intelligent models still have difficulties dealing with big weather data predicting applications, such as striking balance between prediction accuracy learning efficiency. Therefore, multi-head attention encoder-decoder neural network optimized via Bayesian inference strategy (BMAE-Net) proposed herein to predict time series changes accurately. Firstly, we incorporate into gated recurrent unit construct Bayesian-gated units (Bayesian-GRU) module. Then, mechanism introduced design structure each layer, applicability time-length changes. Subsequently, framework hyperparameter optimization designed infer intrinsic relationships among time-series high accuracy. For example, R-evaluation metrics three locations 0.9, 0.804, 0.892, respectively, while RMSE reduced 2.899, 3.011, 1.476, seen Case 1 data. Extensive experiments subsequently demonstrated BMAE-Net has overperformed on location datasets, which provides effective solution applications smart agriculture system.

Язык: Английский

---

Dynamic material parameter inversion of high arch dam under discharge excitation based on the modal parameters and Bayesian optimised deep learning

Advanced Engineering Informatics, Год журнала: 2023, Номер 56, С. 102016 - 102016

Опубликована: Апрель 1, 2023

Язык: Английский

---

BO-STA-LSTM: Building energy prediction based on a Bayesian Optimized Spatial-Temporal Attention enhanced LSTM method

Developments in the Built Environment, Год журнала: 2024, Номер 18, С. 100465 - 100465

Опубликована: Апрель 1, 2024

In predicting building energy (affected by seasons), there are issues like inefficient hyperparameter optimization and inaccurate predictions, it is unclear whether spatial temporal attention improves performance. This study proposes a method based on Bayesian Optimization (BO), Spatial-Temporal Attention (STA), Long Short-Term Memory (LSTM). Seven improved LSTM models (BO-LSTM, SA-LSTM, TA-LSTM, STA-LSTM, BO-SA-LSTM, BO-TA-LSTM, BO-STA-LSTM) compared with the impacts of seasonal variations BO-STA-LSTM analysed using different sample types time domain analysis. To further demonstrate efficiency proposed method, comparisons convolutional neural network (CNN) (TCN) performed, followed validation new datasets. The findings indicate that adding STA BO to enhances average prediction performance 0.0885. alone contributes 0.0717, while 0.0560. achieves higher accuracy for similar test training samples or size 14016, effectively capturing seasonal, trend, peak patterns. Additionally, outperforms CNN TCN, demonstrating superior accuracy.

Язык: Английский

---

Towards a Data-Driven Predictive Framework

Practice, progress, and proficiency in sustainability, Год журнала: 2025, Номер unknown, С. 219 - 296

Опубликована: Янв. 3, 2025

There's a pressing need to democratise DL algorithms while leveraging their performance. This chapter proposes customisable and efficient Automated Machine Learning (AutoML) forecasting framework deal with volatile complex time series using Hyperparameter Optimization (HPO) techniques in combination ANN, LSTM, GRU, Bi-LSTM Bi-GRU. The uses hyperband random search high-dimensional hyperparameter space demonstrate the models' performance without requiring sophisticated pre-processing steps, thereby providing milestone design models after comparative analysis of specific recurrent models. After finding optimal combinations for each model, we study correlation variance between statistical tests, data visualisation tools, SHAP. results discussed improvements elaborating on relationship performance, dataset's size, its inherent noise selection.

Язык: Английский

---

A Novel Hybrid Model to Predict Dissolved Oxygen for Efficient Water Quality in Intensive Aquaculture

IEEE Access, Год журнала: 2023, Номер 11, С. 29162 - 29174

Опубликована: Янв. 1, 2023

Dissolved oxygen content is a key indicator of water quality in aquaculture environment. Because its nonlinearity, dynamics, and complexity, which makes traditional methods face challenges the accuracy speed dissolved prediction. As solution to these issues, this study introduces hybrid model consisting Light Gradient Boosting Machine (LightGBM) Bidirectional Simple Recurrent Unit (BiSRU). Firstly, Linear interpolation smoothing were used identify significant parameters. LightGBM algorithm then determines significance by eliminating irrelevant variables predicting intensive aquaculture. Finally, attention method was implemented map weighting learning parameter matrices, so enabling BiSRU's hidden states be assigned different weights. The findings shown that presented prediction can accurately anticipate fluctuating trend over 10-day period just 122 seconds, rate reached 96.28%. Comparing effects -BiSRU, - GRU, LightGBM-LSTM, BiSRU Attention takes least time. Its higher provide an essential reference for regulation.

Язык: Английский

---

Applicability of Deep Learning Algorithms for Predicting Indoor Temperatures: Towards the Development of Digital Twin HVAC Systems

Buildings, Год журнала: 2023, Номер 13(6), С. 1542 - 1542

Опубликована: Июнь 16, 2023

The development of digital twins leads to the pathway toward intelligent buildings. Today, overwhelming rate data in buildings carries a high amount information that can provide an opportunity for representation and energy optimization strategies Heating, Ventilation, Air Conditioning (HVAC) systems. To implement successful management strategy building, data-driven approach should accurately forecast HVAC features, particular indoor temperatures. Accurate predictions not only increase thermal comfort levels, but also play crucial role saving consumption. This study aims investigate capabilities approaches model predicting A case educational building is considered temperatures using machine learning deep algorithms. algorithms’ performance evaluated compared. important parameters are sorted out before choosing best architecture. Considering real data, prediction models created results reveal all investigated Hence, proposed neural network obtained highest accuracy with average RMSE 0.16 °C, which renders it candidate twin.

Язык: Английский

---

Energy modeling and predictive control of environmental quality for building energy management using machine learning

Energy Sustainable Development/Energy for sustainable development, Год журнала: 2023, Номер 74, С. 381 - 395

Опубликована: Май 11, 2023

Язык: Английский

---

Real-time energy consumption prediction method for air-conditioning system based on long short-term memory neural network

Energy and Buildings, Год журнала: 2023, Номер 298, С. 113527 - 113527

Опубликована: Сен. 7, 2023

Язык: Английский

---

A novel method of fuel consumption prediction for wing-diesel hybrid ships based on high-dimensional feature selection and improved blending ensemble learning method

Ocean Engineering, Год журнала: 2024, Номер 307, С. 118156 - 118156

Опубликована: Май 20, 2024

Язык: Английский

---