火灾对人体生命安全和生存环境具有巨大威胁。烟雾作为火灾的前奏是消防管理十分重视的对象。然而，现行的消防报警系统存在着系统安装繁琐、施工时间长、位置变更困难、响应慢、误报、漏报等问题，无法满足社会的发展需求和智能安防系统对人身安全的保证。因此，本文构建了一种基于NB-IoT的智能烟雾报警系统，主要工作内容如下：

      （1）以NB-IoT智能烟雾报警系统为主线，分别分析了该系统各个模块的需求，并给出相应的设计方案，方案主要包含：传感器终端、云平台、客户端监控管理平台，手机APP端监控报警平台等功能模块的设计。

      （2）针对传感器终端的需求，以STM32F103系列单片机为核心，使用烟雾传感器和NB-IoT模块BC95，实现数据采集与上传功能，同时设计策略降低传感器终端的功耗；云平台选择中国移动云平台OneNET，实现传感器终端的数据汇总；客户端监控管理平台整体设计采用B/S架构，数据库选择MySQL，实现对所有区域传感器终端的设备管理与数据管理，同时引入K-means算法对区域火灾实现分析预测功能，并利用历史数据进行了仿真。仿真结果与实际情况拟合度较高，达到了预测目标，并对未来趋势进行了预测；手机APP端监控报警平台采用Android Studio开发完成，实现了实时报警、信息查询等功能。

      本文研究实现了各个模块的设计与测试工作，并经过上线运行，验证了各个模块均可正常工作，并取得了良好的运行效果。研究结果表明，本文所设计系统为实现提升火灾预警及报警效率提供了一种有效的参考。

关 键 字：NB-IoT；OneNET云平台；烟雾报警；火灾预测分析；手机APP

研究类型：应用研究

      Fire is a great threat to human life and the living environment. Smoke as a prelude to fire is an object of great importance to fire management. However, the existing fire alarm system suffers from cumbersome system installation, long construction time, difficulty in location change, slow response, false alarms and missed alarms, which cannot meet the development needs of society and the assurance of personal safety by intelligent security systems. Therefore, an intelligent smoke alarm system based on NB-IoT is constructed in this thesis.the main job responsibilities are as follows:

      (1) Taking the NB IoT intelligent smoke alarm system as the main line, the requirements of each module of the system were analyzed and corresponding design plans were provided. The plan mainly includes the design of functional modules such as sensor terminals, cloud platforms, client monitoring and management platforms, and mobile APP monitoring and alarm platforms.

      (2) In response to the needs of sensor terminals, the STM32F103 series microcontroller is used as the core, smoke sensors and NB IoT module BC95 are used to achieve data collection and upload functions, and strategies are designed to reduce the power consumption of sensor terminals; The cloud platform chooses China Mobile's cloud platform OneNET to achieve data aggregation of sensor terminals; The overall design of the client monitoring and management platform adopts a B/S architecture and MySQL is selected as the database to realize the equipment management and data management of all regional sensor terminals, while K-means algorithm is introduced to realize the analysis and prediction function of regional fires, and the historical data is used for simulation. The simulation results fit well with the actual situation, achieve the prediction target, and predict the future trend; the cell phone APP terminal monitoring and alarm platform is developed using Android Studio, realizing real-time alarm, information query and other functions.

      The research of this thesis realized the design and testing of each module, and after on-line operation, verified that each module can work normally and achieved good operation effect. The research results show that the system designed in this thesis provides an effective reference for realizing the improvement of fire warning and alarm efficiency.

Keywords: NB-IoT; OneNET cloud platform; Smoke alarm; Fire predictive analytics; Mobile APP

Thesis: Application research

[1]Ahrens M, Evarts B. Fire Loss in the United States during 2020[M]. National Fire Protection Association, 2021.

[2]Kuligowski E. Predicting human behavior during fires[J]. Fire technology, 2013, 49: 101-120.

[3]Liu X, Sun B, Xu Z D, et al. An adaptive Particle Swarm Optimization algorithm for fire source identification of the utility tunnel fire[J]. Fire Safety Journal, 2021, 126: 103486.

[4]Fonollosa J, Solórzano A, Marco S. Chemical sensor systems and associated algorithms for fire detection: A review[J]. Sensors, 2018, 18(2): 553-592.

[5]马沙沙. 基于Arduino & Android的烟雾报警系统[D]. 长安大学, 2017.

[6]Erdal Cayirci, Chunming Rong. Security in Wireless Ad Hoc and Sensor Networks[J]. 2019, (7): 53-58.

[7]Guo S, Yang R, Zhang H, et al. New inverse model for detecting fire-source location and intensity[J]. Journal of thermophysics and heat transfer, 2010, 24(4): 745-755.

[8]Akyildiz I F, Su W, Sankarasubramaniam Y, et al. A survey on sensor networks[J]. IEEE Communications magazine, 2002, 40(8): 102-114.

[9]胡岳. 基于ZigBee网络的室内定位技术的研究[D]. 东南大学, 2019.

[10]梁振兴, 范加新. 短波发射台早期烟雾探测报警系统的应用[J]. 数字传媒研究, 2022, 39(02): 73-75.

[11]胡喻杰, 杨龙成. 基于单片机的家庭烟雾报警系统[J]. 电子技术与软件工程, 2019(22): 243-244.

[12]王千, 金光, 钮俊. 一种基于RSSI的混合定位算法[J]. 传感技术学报, 2015, 28(12): 1823-1829.

[13]沈庆阳. 集成烟雾报警系统的压力仪表控制系统设计概要[J]. 自动化应用, 2020(03): 44-45.

[14]鲍官军, 计时鸣, 张利, 等. CAN 总线技术, 系统实现及发展趋势[J]. 浙江工业大学学报, 2003, 31(1): 58-61.

[15]Chen H, Zhou J, Liu S, et al. Flame-retardant triboelectric generator with stable thermal-mechanical-electrical coupling performance for fire Bluetooth alarm system[J]. Nano Energy, 2022, 102: 107634.

[16]Zhang L, Huang Y, Dong H, et al. Flame-retardant shape memory polyurethane/MXene paper and the application for early fire alarm sensor[J]. Composites Part B: Engineering, 2021, 223: 109149.

[17]晏小庆, 阮云兰, 文志诚. 基于无线红外通信的烟雾自动报警系统设计[J]. 现代电子技术, 2021, 44(08): 24-28.

[18]Zhu H, Ji J, Nie J. Early fire evolution and alarm characteristics of cable fires in long and narrow spaces[J]. Fire safety journal, 2022: 103627.

[19]Jafari M J, Pouyakian M, Hanifi S M. Reliability evaluation of fire alarm systems using dynamic Bayesian networks and fuzzy fault tree analysis[J]. Journal of Loss Prevention in the Process Industries, 2020, 67: 104229.

[20]韩倩倩. 基于LoRa技术的无线火灾报警系统研究[D]. 北京建筑大学, 2020.

[21]罗万鑫. 极早期烟雾报警系统在博物馆的应用[J]. 设备管理与维修, 2021(06): 85-86.

[22]杜锋, 郑腾腾, 翟德超. 基于DTU模块的无线烟雾报警系统的设计与实现[J]. 电脑知识与技术, 2020, 16(34): 219-220.

[23]李红霞, 顾亭亭. 基于ST89C52的烟雾报警系统的设计[J]. 电子技术, 2020, 49(03): 76-77.

[24]肖家文, 茅胜荣, 陈小平. 低功耗远程烟雾报警系统设计[J]. 微型机与应用, 2017, 36(02): 96-99.

[25]张子良. 基于AHP多特征融合的矿用烟雾传感器设计[J]. 煤矿机械, 2023, 44(06): 6-10.

[26]王林, 薛国阳. 基于ViBe和目标检测的视频烟雾检测算法[J]. 单片机与嵌入式系统应用, 2023, 23(04): 33-37.

[27]丰颖. 基于双光源检测的烟雾多特征感知技术研究[J]. 电子设计工程, 2023, 31(05): 18-22.

[28]郑远攀, 许博阳, 王振宇. 改进的YOLOv5烟雾检测模型[J]. 计算机工程与应用, 2023, 59(07): 214-221.

[29]吕洪坤, 陈云, 毛智超.基于高强超声波的电力隧道火灾烟雾颗粒试验研究[J]. 浙江电力, 2022, 41(02): 47-52.

[30]孙浩东, 王超, 李深奥, 等. 基于Arduino的烟雾探测实验装置设计[J]. 机电信息, 2021(17): 42-43.

[31]白宝军. 基于图像增强与多特征的煤矿井下火灾烟雾检测方法研究[D]. 西安建筑科技大学, 2021.

[32]杨剑, 陈世娥. 基于图像处理技术的烟雾检测在家庭火灾防范中应用[J]. 电脑知识与技术, 2021, 17(06): 189-191.

[33]Zhang Z H, Zhang J W, Cao C F, et al. Temperature-responsive resistance sensitivity controlled by L-ascorbic acid and silane co-functionalization in flame-retardant GO network for efficient fire early-warning response[J]. Chemical Engineering Journal, 2020, 386: 123894.

[34]李澎林, 章军伟, 李伟. 基于光流改进与YOLOv3的烟雾检测方法[J]. 浙江工业大学学报, 2021, 49(01): 9-15.

[35]Deng T, Wang S, Zhu M. Dual-wavelength optical sensor for measuring the surface area concentration and the volume concentration of aerosols[J]. Sensors and Actuators B: Chemical, 2016, 236: 334-342.

[36]Wang S, Xiao X, Deng T, et al. A Sauter mean diameter sensor for fire smoke detection[J]. Sensors and Actuators B: Chemical, 2019, 281: 920-932.

[37]秦莉艳. 单片机的智能烟雾报警系统的设计[J]. 电子测试, 2020(21): 16-17+75.

[38]王晓薇, 张笑语, 孙可.基于改进深度信念网络的校园烟雾检测[J]. 沈阳师范大学学报(自然科学版), 2020, 38(04): 345-350

[39]Loepfe M, Nenkov M. Application of the Rayleigh-Debye-Gans theory on the scattering of light by standard test fire aerosols[C]//Proc. of the 15th International Conference on Automatic Fire Detection AUBE. 2014, 1: 45-52.

[40]彭建英, 刘雨丽, 郭杰荣. 一款单片机智能烟雾报警系统的设计[J]. 机械工程与自动化, 2018(06): 151-152.

[41]张清河, 单琰, 吴欣悦等. 基于K-means聚类的相控阵任意形状波束子阵综合方法[J/OL]. 微波学报: 1-8[2023-05-25].