基于模糊神经网络的复合火灾探测系统的研究

发布时间：2018-04-11 23:32

本文选题：模糊神经网络 + 火灾探测　；参考：《中国民用航空飞行学院》2017年硕士论文

【摘要】：火灾严重威胁人民的生命和财产安全,但传统火灾探测器多采用单一传感器,且火灾探测算法使用简单的阈值法,在现实火灾环境下容易出现火灾误报等现象。因此提高火灾探测系统的性能,降低火灾误报率是当前亟待解决的问题。本文主要研究基于模糊神经网络的复合火灾探测系统。以CO气体、烟雾、温度作为火灾特征参量,设计了复合火灾探测器并以RS-485总线构建多点火灾探测系统。利用CO传感器MQ-7、温度传感器TC1047、红外发光管PD243B和硅光电池IR5123C设计火灾探测电路;利用MAX1348芯片构建了RS-485总线电路;利用CATIA软件设计探测器外壳,并用3D打印完成外壳制作。设计单片机控制程序完成下位机对三路火灾信号的采集、A/D转换和传输;利用模糊系统和神经网设计火灾探测算法;利用Labview设计上位机用户界面,实现对火灾特征参量、通信自检和实时报警的显示、存储、再现。应用新研制的复合火灾探测器进行功能性验证实验、火灾响应时间实验、抗干扰实验、分布式模拟火灾探测实验。实验结果表明:新研制的复合火灾探测器能够对木材阴燃火、棉绳阴燃火、聚氨酯明火、乙醇明火进行火灾探测,火灾平均响应时间19.8s,比市售的海湾型探测器响应时间快2.1s,满足国家标准中规定的火灾响应时间≤30s的要求;在800次香烟火、打火机火、暖风机、粉尘抗干扰实验中误报率为0.3%;由三路复合火灾探测器组建实验系统能进行单点和多点火灾探测。
[Abstract]:Fire safety seriously threaten the people's life and property, but the traditional fire detectors using a single sensor, and fire detection algorithm using simple threshold method, false alarm phenomenon is easy to appear in the real fire environment. So to improve the performance of the fire detection system, reduce the false alarm rate is the current problems to be solved in this paper. Based on the fuzzy neural network composite fire detection system. With CO gas, smoke, fire temperature as characteristic parameters, design and construction of multi point composite fire detector fire detection system based on RS-485 bus. Using CO MQ-7 sensor, TC1047 temperature sensor, infrared light emitting diode PD243B and silicon photocell IR5123C fire detection circuit design; construction of RS-485 bus circuit using MAX1348 chip; using CATIA software design and 3D detector shell, shell print production. Control design of monolithic machine In order to complete MCU of three fire signal, A/D conversion and transmission; detection algorithm using fuzzy system and neural network design of fire; using Labview to design PC user interface, realize the fire characteristics parameters, communication self-test and real-time alarm display, storage and reproduction. Combined fire detector by means of a new function validation experiments, the response time of fire experiment, interference experiment, distributed simulation of fire detection experiments. The experimental results show that the newly developed composite fire detector to smoldering fire wood, cotton rope smoldering fire, polyurethane fire, fire detection of ethanol flame, fire average response time 19.8s, time response of 2.1S bay type detector than the commercially available, meet the national standards in the fire response time is less than 30s; in the 800 cigarette lighter fire, fire, air blower, dust in the experiments of anti disturbance false positive rate is 0. 3%, an experimental system constructed by a three road composite fire detector can be used for single point and multi point fire detection.

【学位授予单位】：中国民用航空飞行学院
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP183;X932

【相似文献】