| 舒舜,张羽杨,陆梓萍,王董,姜楠,杨立中.基于机器学习的热释放速率实时预测方法[J].火灾科学,2022,31(1):8-14. |
| 基于机器学习的热释放速率实时预测方法 |
| Real-time prediction of heat release rate based on machine learning |
| |
| 查看全文 查看/发表评论 下载PDF阅读器 |
| DOI：10.3969/j.issn.1004-5309.2022.01.02 |
| 基金项目:安徽省自然科学基金项目(2008085ME153);上海市黄浦区消防救援支队项目(EF2320000061) |
| 作者 | 单位 | | 舒舜 | 上海市消防救援总队黄浦区支队,上海,200011 | | 张羽杨 | 上海市消防救援总队黄浦区支队,上海,200011 | | 陆梓萍 | 中国科学技术大学火灾科学国家重点实验室,合肥,230026 | | 王董 | 中国科学技术大学火灾科学国家重点实验室,合肥,230026 | | 姜楠 | 中国科学技术大学火灾科学国家重点实验室,合肥,230026 | | 杨立中* | 中国科学技术大学火灾科学国家重点实验室,合肥,230026 |
|
| 中文关键词: 支持向量机 热释放速率 火焰体积 火焰高度 实时预测 |
| 英文关键词:SVM Heat release rate Flame volume Flame height Real-time prediction |
| 摘要点击次数: 256 |
| 全文下载次数: 2670 |
| 中文摘要: |
| 火灾态势的实时感知对提升消防救援指挥与决策水平有非常重要的意义。为实现火源热释放速率的实时预测,通过丙烷气体燃烧实验获取火焰体积和高度的实时数据,选取MATLAB中精细高斯SVM做分类训练,建立了基于图像处理的热释放速率预测模型。结果表明,训练完成后的模型可以根据3 s内的火焰体积/高度数据预测得到实时的燃料流量,获取相应的火源热释放速率。该预测模型在已知数据集和未知数据集上都表现出了较高的准确性,为火灾态势发展预测提供了理论和技术基础,有很好的应用前景。 |
| 英文摘要: |
| The real-time perception of fire situation is significant for fire rescue. In this paper, a prediction model based on image processing was established, which realized the real-time prediction of heat release rate of fire source. The refined Gaussian SVM in MATLAB was selected to conduct classification training on the continuous flame volume and height data obtained from propane combustion experiment. The results show that the model after training could predict the real-time fuel flow according to the flame volume/height data in 3 s, and obtain the corresponding heat release rate of fire source. This model has high accuracy in the validation set. Even though the accuracy of the prediction model for unfamiliar data is relatively low, it still proves that the model has certain scalability. The prediction model presented in this work provides a theoretical and technical basis for the prediction of fire situation. |
| 关闭 |
|
|
|
