[1]李 岩,孔冬冬.基于消防物联网的配电柜早期火灾智能预警研究[J].工业仪表与自动化装置,2026,(01):103-108.[doi:10.19950/j.cnki.CN61-1121/TH.2026.01.019]
　LI Yan,KONG Dongdong.Research on early fire intelligent warning of distribution cabinets based on fire internet of things[J].Industrial Instrumentation & Automation,2026,(01):103-108.[doi:10.19950/j.cnki.CN61-1121/TH.2026.01.019]
点击复制

基于消防物联网的配电柜早期火灾智能预警研究()

《工业仪表与自动化装置》[ISSN:1000-0682/CN:61-1121/TH]

卷:
期数:
2026年01期
页码:
103-108
栏目:
出版日期:
2026-02-15

文章信息/Info

Title:
Research on early fire intelligent warning of distribution cabinets based on fire internet of things
文章编号:
1000-0682（2026）01-0103-06
作者:
李 岩; 孔冬冬
中国核电工程有限公司郑州分公司 电气仪控所，河南 郑州 450052
Author(s):
LI Yan; KONG Dongdong
China Nuclear Power Engineering Co., Ltd. Zhengzhou Branch, Henan Zhengzhou 450052, China
关键词:
智慧消防物联网; 火灾早期预警; KPCA; 改进KFCM
Keywords:
smart fire internet of things; early fire warning; KPCA algorithm; improved KFCM algorithm
分类号:
X934；TP391
DOI:
10.19950/j.cnki.CN61-1121/TH.2026.01.019
文献标志码:
A
摘要:
核燃料元件厂电气火灾威胁核安全及关键系统运行。为提高配电柜早期火灾预警的准确率，降低误报漏报，实现消防安全管理的提质增效，对某厂房高低压变配电室新增电气火灾监控设备、柜内灭火及视频AI分析装置，构建智慧消防物联网系统。针对多参数融合火灾检测非线性关联及噪声强等特点，采用KPCA算法对数据进行了特征提取和降维降噪。针对密集簇和稀疏簇混合的数据分布特点，提出多参数自适应调整的改进KFCM算法。仿真实验中，利用感知层剩余电流探测器采集的线路电压和电流、剩余电流及线路温度参数，热解粒子探测器检测的PM1.0及CO浓度值作为特征输入，测试结果表明，KPCA-改进KFCM方法的准确率均优于单一KFCM及KPCA-KFCM算法，且鲁棒性、收敛速度及适用性更优异。
Abstract:
In nuclear fuel element factories the electrical fire threat nuclear safety and operation of key systems. To improve the accuracy of early fire warning for distribution cabinets, reduce false alarms and missed alarms, and achieve quality and efficiency improvement in fire safety management, new electrical fire monitoring equipment, in-cabinet fire extinguishing, and video AI analysis devices were added to the high and low voltage distribution rooms of a certain factory to build a smart fire Internet of Things system. Aiming at the characteristics of nonlinear correlation and strong noise in multi-parameter fusion fire detection, the KPCA algorithm is used for feature extraction, dimensionality reduction, and denoising of data. Aiming at the data distribution characteristics of mixed dense clusters and sparse clusters, an improved KFCM algorithm with multi-parameter adaptive adjustment is proposed. In the simulation experiment, the line voltage and current, residual current, and line temperature parameters collected by residual current detectors in the perception layer, and the PM1.0 and CO concentration values detected by pyrolysis particle detectors are used as feature inputs. The test results show that the accuracy of the KPCA-improved KFCM method is better than single KFCM and KPCA-KFCM algorithms, and its robustness, convergence speed, and applicability are more excellent.

参考文献/References:

[1] 冯子毅,马恒瑞,王红霞,等.基于多类型传感器的变电站火灾预警技术研究综述[J].智慧电力,2024,52(10): 103-111.

[2] 汪嘉俊,倪顺江.电力设备火灾危险性分析与防控技术[J].中国安全生产科学技术,2022,18(9):189-194.
[3] 何勇军,易欣,王伟峰,等.煤矿井下电气火灾智能监控与灭火技术综述[J].煤矿安全,2022,53(9):55-64.
[4] 邢国新,赵海龙,吴志强.热解粒子式电气火灾探测器在地铁中的应用[J].消防科学与技术,2021,40(11): 1695-1698.
[5] 谢丹,李越,洪伟艺,等.变电站电气柜火灾探测及分散式气体灭火装置设计[J].消防科学与技术,2023, 42(8):1126-1130.
[6] 贺胜,疏学明,胡俊,等.基于消防大数据的电气火灾风险预测预警方法[J].清华大学学报(自然科学版), 2024,64(3):478-491.
[7] 吕新东,单强,刘辉,等.三相不平衡对剩余电流保护器误动的影响分析[J].工业仪表与自动化装置, 2022(2):81-85.
[8] 严亚波.城市轨道交通车站剩余电流式电气火灾监控探测器误报原因分析[J].城市轨道交通研究, 2022(6):200-202.
[9] 张潜,徐少红,薛宏佺.降低地铁列车火灾报警系统误报率策略研究[J].城市轨道交通研究,2025(5):63-66.
[10] 谭启鹏,李勇琦,陈满,等.基于KPCA-MTCN 的锂离子电池故障诊断方法[J].工程科学学报,2024,46(12): 2297-2306.
[11] 赵文虎,蔡生宏,王文.基于KPCA融合AdaBoost-IBOA-ELM模型的TE过程故障诊断[J].工业仪表与自动化装置,2024(4):102-109.
[12] 张希望,朱前坤,王宪玉,等.基于VMD-KPCA-LSTM的桥梁监测应变数据预测[J].应用基础与工程科学学报,2025,33(1):76-86.
[13] XU Yanhui, GAO Yihao, CHENG Yundan, et al. Substation clustering based on improved KFCM algorithm with adaptive optimal clustering number selection[J].Global Energy Interconnection,2023,6(4): 505-516.
[14] 张友鹏,张迪,杨妮,等.基于CEEMDAN与KFCM聚类的转辙机退化状态识别方法[J].中国铁道科学, 2023,44(1):194-201.
[15] 马莉,霍耀佳,吴杨,等.基于VMD和KFCM-SVM的高压断路器声振联合故障诊断方法[J].高压电器, 2024,8(16):53-62.

相似文献/References:

备注/Memo

备注/Memo:
收稿日期：2025-09-19第一作者：李岩（1982—），男，河南许昌人，硕士，高级工程师，注册电气工程师，主要研究方向为智慧消防、火灾自动报警系统设计等。E-mail：ncepu_ly@126.com
更新日期/Last Update: 1900-01-01