Performance prediction of a ground source heat pump system using denoised long short-term memory neural network optimised by fast non-dominated sorting genetic algorithm-II

Geothermics, Journal Year: 2024, Volume and Issue: 120, P. 103002 - 103002

Published: March 22, 2024

Language: Английский

---

An ultra-short-term wind power prediction method based on spatial-temporal attention graph convolutional model

Energy, Journal Year: 2024, Volume and Issue: 293, P. 130751 - 130751

Published: Feb. 19, 2024

Language: Английский

---

Modal decomposition integrated model for ultra-supercritical coal-fired power plant reheater tube temperature multi-step prediction

Energy, Journal Year: 2024, Volume and Issue: 292, P. 130521 - 130521

Published: Jan. 31, 2024

Language: Английский

---

Multi-scale feature enhanced spatio-temporal learning for traffic flow forecasting

Knowledge-Based Systems, Journal Year: 2024, Volume and Issue: 294, P. 111787 - 111787

Published: April 10, 2024

Language: Английский

---

Operational strategy optimization of an existing ground source heat pump (GSHP) system using an XGBoost surrogate model

Energy and Buildings, Journal Year: 2024, Volume and Issue: 318, P. 114444 - 114444

Published: June 24, 2024

Language: Английский

---

Ensemble learning for predicting average thermal extraction load of a hydrothermal geothermal field: A case study in Guanzhong Basin, China

Energy, Journal Year: 2024, Volume and Issue: 296, P. 131146 - 131146

Published: April 3, 2024

Language: Английский

---

CNN-LSTM Hybrid Model to Promote Signal Processing of Ultrasonic Guided Lamb Waves for Damage Detection of Metallic Pipelines

Published: July 14, 2023

Ultrasonic-guided lamb wave approach is an effective non-destructive testing (NDT) method used for detecting localized mechanical damage, corrosion, and welding defects experienced in metallic pipelines. Signal processing of the guided waves often challenged due to complexity operational conditions environments Machine learning approaches recent years, including convolutional neural networks (CNN) long short-term memory (LSTM), have exhibited their advantages overcome these challenges signal process data classification complex systems, thus great potential damage detection critical oil/gas pipeline structures. In this study, a CNN-LSTM hybrid model utilized decoding ultrasonic pipelines, twenty-nine features are extracted as input classify different types from pipes. The prediction capacity assessed by comparing it CNN LSTM. results demonstrate that exhibits much higher accuracy, with 94.8%, compared those Interestingly, also reveal predetermined features, time-, frequency-, time-frequency domains, could significantly improve robustness deep approaches, even though believed they include automated feature extraction, without hand-crafted steps shallow do. Furthermore, displays performance when noise level relatively low (e.g., SNR=9 or higher), other two models, but its drops gradually increase noise.

Language: Английский

---

CNN-LSTM Hybrid Model to Promote Signal Processing of Ultrasonic Guided Lamb Waves for Damage Detection in Metallic Pipelines

Sensors, Journal Year: 2023, Volume and Issue: 23(16), P. 7059 - 7059

Published: Aug. 9, 2023

The ultrasonic guided lamb wave approach is an effective non-destructive testing (NDT) method used for detecting localized mechanical damage, corrosion, and welding defects in metallic pipelines. signal processing of waves often challenging due to the complexity operational conditions environment Machine learning approaches recent years, including convolutional neural networks (CNN) long short-term memory (LSTM), have exhibited their advantages overcome these challenges data classification complex systems, thus showing great potential damage detection critical oil/gas pipeline structures. In this study, a CNN-LSTM hybrid model was utilized decoding pipelines, twenty-nine features were extracted as input classify different types pipes. prediction capacity assessed by comparing it those CNN LSTM. results demonstrated that much higher accuracy, reaching 94.8%, compared Interestingly, also revealed predetermined features, time, frequency, time-frequency domains, could significantly improve robustness deep approaches, even though are believed include automated feature extraction, without hand-crafted steps shallow learning. Furthermore, displayed performance when noise level relatively low (e.g., SNR = 9 or higher), other two models, but its dropped gradually with increase noise.

Language: Английский

---

Evaluation and machine learning prediction on thermal performance of energy walls in underground spaces as part of ground source heat pump systems

Tunnelling and Underground Space Technology, Journal Year: 2024, Volume and Issue: 148, P. 105750 - 105750

Published: April 6, 2024

Language: Английский

---

Explainable machine learning models to predict outlet water temperature of pipe-type energy pile

Renewable Energy, Journal Year: 2025, Volume and Issue: unknown, P. 122972 - 122972

Published: March 1, 2025

Language: Английский

---