Open AccessConstructing a Gas Explosion Inversion Model in a Straight Roadway Using the GA–BP Neural Network
Author(s) -
Bo Tan,
Heyu Zhang,
Gang Cheng,
Yanling Liu,
Xuedong Zhang
Publication year - 2021
Publication title -
acs omega
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.779
H-Index - 40
ISSN - 2470-1343
DOI - 10.1021/acsomega.1c03926
Subject(s) - overpressure , gas explosion , inversion (geology) , matlab , artificial neural network , intensity (physics) , computer science , simulation , engineering , geology , artificial intelligence , forensic engineering , seismology , physics , tectonics , quantum mechanics , thermodynamics , operating system
When the location and intensity of the source of an explosion are determined, the severity and impact of the explosion can be analyzed and predicted, such as the overpressure, temperature, and toxic gas propagation. Determining the location and intensity of the explosion source can also provide a theory for emergency rescue work, improve rescue efficiency, and ensure the safety of rescue personnel. Therefore, the location and intensity of the source of the explosion through field data inversion are of great significance. Based on a genetic algorithm (hereinafter GA) to improve back propagation (BP) neural network theory, the location and intensity of the roadway gas explosion source were inverted through a gas explosion experiment and simulated overpressure data. When all parameters reached the optimal iteration, MATLAB was used to realize the final inversion model of the roadway gas explosion disaster. Compared with the real results, this model has high precision in determining the location of the explosion source and has a high reference value. The overall accuracy of the roadway gas explosion disaster inversion model results is high and reliable, and the inversion model of the roadway gas explosion disaster is established to provide data support for emergency rescue and accident investigation.