不同结构参数下的半U形地下空间烟气运动参数的机器学习预测  

Machine Learning Prediction of Smoke Motion Parameters in a Half-U-shaped Underground Space with Different Structural Parameters

作　　者：徐志胜[1] 殷耀龙 雷志强 陈诗仪 应后淋 XU Zhisheng;YIN Yaolong;LEI Zhiqiang;CHEN Shiyi;YING Houlin(School of Civil Engineering,Central South University,Changsha 410075,China)

机构地区：[1]中南大学土木工程学院,湖南长沙410075

出　　处：《灾害学》2025年第1期67-73,共7页Journal of Catastrophology

基　　金：国家自然科学基金项目“环保型泡沫灭火剂性能调控机制与理论模型研究”(52176146)。

摘　　要：结合火灾数值模拟(FDS)与机器学习方法,该文对半U形地下空间火灾时的烟气运动进行了深入分析。研究发现,在预测烟气回流长度及烟气最高温升方面,BP神经网络相比于支持向量机回归(SVR)展现了更高的精度,其决定系数超过了95%,而相对误差仅集中在20%以内,显著优于SVR方法。通过shap值解释机器学习模型,并结合FDS数值模拟的结果,揭示坡高是影响烟气回流长度的决定性因素,且坡高的增大、宽度的减小或热释放速率增大均会缩短烟气回流。同时,热释放速率是影响烟气最高温升的主要因素,受坡高影响较大,而宽度的减小虽能在一定幅度上降低最高烟气温升,但效果并不显著。该研究拓展了地下空间火灾烟气运动参数的预测方法,为地下空间火灾动力学行为预测及通风排烟系统的优化设计贡献了创新性的方法。A comprehensive analysis of smoke movement during a fire in a half—U-shaped underground space is conducted by using a combination of Fire Dynamics Simulator(FDS)and machine learning techniques.The findings indicate that tlie Backpropagation Neural Network(BPNN)outperforms Support Vector Regression(SVR)in forecasting the length of smoke backlayering and the maximum temberature increase,with a determination coefficient exceeding 95%and relative errors predominantly within the 20%range,marking a significant improvement over the SVR method.By elucidating the machine learning model through SHAP values and integrating the outcomes of FDS numerical simulations,it is determined that the slope height is the pivotal factor influencing the length of smoke backlayering.An increase in slope height,a reduction in width,or an escalation in heat release rate are all found to curtail the smoke reflux length.Concurrently,the heat release rate is identified as the primary factor affecting tire maximum temjrerature rise of the smoke,with the slope height exerting a substantial influence.Although a decrease in width can marginally diminish the maximum temjrerature rise of the smoke,its impact is not pronounced.This research not only broadens tire scope of predictive methods for fire smoke motion parameters in underground spaces but also presents an innovative approach to forecasting tire dynamics of fires in underground environments and to the optimization of ventilation and smoke exhaust systems.

关 键 词：半U形地下空间 机器学习 截面宽度 坡高 烟气回流长度 最高烟气温升 

分 类 号：X43[环境科学与工程—灾害防治] X915.5