A dimensionless model and ant colony optimization fusion temperature prediction in tunnel fires

Prior studies have noted the importance of ceiling temperature distribution prediction of tunnel fires. To overcome the limitations of precise modeling difficulties for model-driven methods and biased toward the convergence for the data-driven methods due to loss of physical interpretations, a hybri...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Applied soft computing Jg. 145; S. 110564
1. Verfasser: Sun, Bin
Format: Journal Article
Sprache:Englisch
Veröffentlicht: Elsevier B.V 01.09.2023
Schlagworte:
Ant colony optimization
Dimensional analysis method
Model-driven and data-driven fusion algorithm
Tunnel fires
Ventilation speed
Ant colony optimization
Tunnel fires
Ventilation speed
Dimensional analysis method
Model-driven and data-driven fusion algorithm
ISSN:1568-4946, 1872-9681
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Prior studies have noted the importance of ceiling temperature distribution prediction of tunnel fires. To overcome the limitations of precise modeling difficulties for model-driven methods and biased toward the convergence for the data-driven methods due to loss of physical interpretations, a hybrid model-driven and data-driven fusion algorithm is established to predict ceiling temperature distribution in tunnel fires based on dimensional analysis method and ant colony optimization algorithm. Thus, full-scale tunnel fire experiments with different ventilation speeds are conducted to support the algorithm. In contrast to the experimental data as well as the known theoretical model, data-driven back propagation neural network algorithm and hybrid ant colony optimization and back propagation neural network algorithm, the effectiveness and ability of the algorithm are verified, which can be used for the tunnel fire prediction. [Display omitted] •A hybrid model-driven and data-driven fusion algorithm is developed.•The algorithm have both advantages of the model-driven and data-driven methods.•Ceiling temperature distribution with different ventilations can be predicted.•Full-scale tunnel fire experiments were conducted to verify the algorithm.
ISSN:1568-4946
1872-9681
DOI:10.1016/j.asoc.2023.110564