Vented gas explosion overpressure prediction of obstructed cubic chamber by Bayesian Regularization Artificial Neuron Network

Shi, J.Li, J.Hao, H.Pham, T.M.Zhu, Y.Chen, G.2025-12-182025-12-182018Journal of Loss Prevention in the Process Industries, 2018; 56:209-2160950-42301873-3352https://hdl.handle.net/11541.2/35334Data source: Supplementary data, https://doi.org/10.1016/j.jlp.2018.05.016This study aims to develop an integrated model, namely Bauwens-BRANN model, to estimate the maximum overpressure of vented gas explosion. A series of experiments designed for cubic enclosures with and without obstacles are used in the development of Bauwens-BRANN model. Two important parameters are modified to address the pre-existing issues of Bauwens model. By incorporating the Bayesian Regularization Artificial Neuron Network (BRANN) algorithm into the Bauwens model, the Bauwens-BRANN model is developed. Improved pressure estimation accuracy is seen for the Bauwens-BRANN model in comparison with the NFPA-68 2013 model.enCopyright 2018 Elsevier Ltd. All rights reserved.NFPA-68 2013bayesian regularization artificial neuron networkbauwens analytical modelbetter generalizationvented explosion of obstructed cubic chamberVented gas explosion overpressure prediction of obstructed cubic chamber by Bayesian Regularization Artificial Neuron Network - Bauwens modelJournal article10.1016/j.jlp.2018.05.0162-s2.0-85053399723Pham, T.M. [0000-0003-4901-7113]