Citation

Li C, Deng X, Shi Y, Sun D, Jiao Y. China Saf. Sci. J. 2022; 32(3): 90-97.

Vernacular Title

社区户内燃气泄漏动态预警模型

Copyright

(Copyright © 2022, China Occupational Safety and Health Association, Publisher Gai Xue bao)

DOI

10.16265/j.cnki.issn1003-3033.2022.03.012

PMID

unavailable

Abstract

In order to improve gas leakage warning performance for community safety, a dynamic early warning model for household gas leakage was proposed. Firstly, indoor gas data of each home in the community were collected by using wireless sensor network, and uploaded to the cloud by smart gateway. Secondly, inputs of random forest algorithm were optimized by utilizing fuzzy control algorithm to reduce interference of features with lower importance on the cloud platform, based on which a fuzzy-random forest model was established with optimized data as input of random forest algorithm and leakage grade as output. Then, a visual module was developed to present gas leakage grade of each home. Finally, the model's effectiveness was verified through simulation test under lab conditions based on historical gas data collected form a certain community in Beijing. The results show that this model can effectively improve ability of online monitoring and dynamic early warning of gas leaks in the community. Compared with other algorithms, the fuzzy-random forest algorithm shows better performance in detecting early small leaks.

===

为提升社区户内燃气泄漏动态预警的能力，构建一种社区户内燃气泄漏动态预警模型。首先，利用无线传感器网络采集社区户内燃气数据，通过智能网关实现数据上云的功能；其次，在云平台中利用模糊控制算法优化随机森林算法的输入，减少重要度较低特征的干扰，并将优化后的数据作为随机森林算法的输入，以泄漏等级作为输出，建立模糊-随机森林模型；然后，开发可视化模块，以显示社区户内燃气泄漏的等级；最后，根据北京某社区燃气历史数据，在实验室条件下进行仿真模拟，验证该模型的有效性。结果表明：该模型可有效提升社区户内燃气泄漏的在线监测和动态预警的能力，与其他算法相比，模糊-随机森林算法在发现早期微小泄漏方面表现更优。

Language: en