Intelligent fire detector compound sensor neural network
data description
[Pages ]36[ Words] 23 1 19
[content]
abstract
abstract
1 Introduction 1
2 System Architecture Design 4
3 805 1 Introduction of Single Chip Microcomputer 5
4 System Hardware Design 9
5 system software design 25
6 Conclusion 28
refer to
Express gratitude/gratitude
[Original]
1 Introduction
1. 1 fire alarm overview
With the rapid development of sensor technology, microprocessor technology and signal processing technology, compound fire detection has become the development direction of automatic fire detection technology. At present, composite fire detectors mainly include photoelectric smoke sensing, ion smoke sensing, photoelectric smoke sensing and ion smoke sensing. The main purpose of using composite detection method is to make the detector detect all kinds of fires in a balanced way. Especially, the scattered light smoke detector overcomes its insensitivity to black smoke rising with temperature through temperature compensation, which effectively promotes the application of photoelectric smoke detector. However, the composite detector of photoelectric smoke sensor and temperature sensor is relatively poor in detecting black smoke rising at low temperature, and ion smoke is increasingly difficult to be accepted by the market because of the possibility of radioactive pollution. Moreover, both photoelectric and ion smoke detection methods are essentially ion detection, and the interference of dust, water vapor, oil mist and other particles also affects it. Although signal processing can be used to suppress these interferences, it is difficult to completely eliminate them, so it is necessary to find a new fire detection method, which can detect fires more effectively and reduce false positives.
As we all know, most fires will produce carbon monoxide (CO) gas, especially in the early stage of fire with insufficient combustion. CO gas is lighter than air and more diffuse than smoke, especially the false alarm sources of many commonly used smoke detection methods do not produce CO gas. Therefore, it is an ideal early fire detection method to introduce CO sensor into fire detection.
Since 1990s, the characteristics of self-learning, self-adaptation and self-organization of neural network have attracted great attention from the fire fighting and engineering circles in various countries. Japan's Y. Okayama proposed a three-layer feedforward BP neural network for fire detection, which has certain self-learning habits and adaptability. However, it does not fully consider the characteristics of sensor signals, and only uses a simple threshold to judge directly, which is not conducive to reducing the false alarm rate of fire. S. Nakanishi and others used fuzzy logic method to process the composite signals of smoke, temperature and CO in the smoke concentration signal, and used neural network algorithm to adjust the system. The actual results show that the false alarm rate is reduced by 50%, and the fire alarm time is advanced, but only the fuzzy logic method is used to adjust the alarm delay time, which does not give full play to the advantages of neural network. ......
[Abstract]
At present, the networking and automation technology in the field of automatic fire alarm system is improving day by day, but there are still some problems such as false alarm and missing report in fire detectors. The accuracy of fire detector detecting fire will directly affect the performance of the whole automatic alarm system. Therefore, fire detector technology has become the main development direction in this field.
The latest information in the information bar of the help home page.
The hottest information
Best selling materials
Open the label
Authoritative member
Circle search circle
Open circle
member enrollment
All the information
Circle name and circle introduction
Post in the circle
Design of compound intelligent fire alarm
Design of compound intelligent fire alarm
525.0 Crawl data files and search the collected shopping carts.
Design of compound intelligent fire alarm. doc (773KB)
label
Intelligent fire detector compound sensor neural network
data description
[Pages ]36[ Words] 23 1 19
[content]
abstract
abstract
1 Introduction 1
2 System Architecture Design 4
3 805 1 Introduction of Single Chip Microcomputer 5
4 System Hardware Design 9
5 system software design 25
6 Conclusion 28
refer to
Express gratitude/gratitude
[Original]
1 Introduction
1. 1 fire alarm overview
With the rapid development of sensor technology, microprocessor technology and signal processing technology, compound fire detection has become the development direction of automatic fire detection technology. At present, composite fire detectors mainly include photoelectric smoke sensing, ion smoke sensing, photoelectric smoke sensing and ion smoke sensing. The main purpose of using composite detection method is to make the detector detect all kinds of fires in a balanced way. Especially, the scattered light smoke detector overcomes its insensitivity to black smoke rising with temperature through temperature compensation, which effectively promotes the application of photoelectric smoke detector. However, the composite detector of photoelectric smoke sensor and temperature sensor is relatively poor in detecting black smoke rising at low temperature, and ion smoke is increasingly difficult to be accepted by the market because of the possibility of radioactive pollution. Moreover, both photoelectric and ion smoke detection methods are essentially ion detection, and the interference of dust, water vapor, oil mist and other particles also affects it. Although signal processing can be used to suppress these interferences, it is difficult to completely eliminate them, so it is necessary to find a new fire detection method, which can detect fires more effectively and reduce false positives.
As we all know, most fires will produce carbon monoxide (CO) gas, especially in the early stage of fire with insufficient combustion. CO gas is lighter than air and more diffuse than smoke, especially the false alarm sources of many commonly used smoke detection methods do not produce CO gas. Therefore, it is an ideal early fire detection method to introduce CO sensor into fire detection.
Since 1990s, the characteristics of self-learning, self-adaptation and self-organization of neural network have attracted great attention from the fire fighting and engineering circles in various countries. Japan's Y. Okayama proposed a three-layer feedforward BP neural network for fire detection, which has certain self-learning habits and adaptability. However, it does not fully consider the characteristics of sensor signals, and only uses a simple threshold to judge directly, which is not conducive to reducing the false alarm rate of fire. S. Nakanishi and others used fuzzy logic method to process the composite signals of smoke, temperature and CO in the smoke concentration signal, and used neural network algorithm to adjust the system. The actual results show that the false alarm rate is reduced by 50%, and the fire alarm time is advanced, but only the fuzzy logic method is used to adjust the alarm delay time, which does not give full play to the advantages of neural network. ......
[Abstract]
At present, the networking and automation technology in the field of automatic fire alarm system is improving day by day, but there are still some problems such as false alarm and missing report in fire detectors. The accuracy of fire detector detecting fire will directly affect the performance of the whole automatic alarm system. Therefore, fire detector technology has become the main development direction in this field.
Based on the analysis of the research status and existing problems at home and abroad, this paper studies the fire detector technology from both hardware and software. In terms of hardware, a composite detector consisting of a temperature sensor, a CO sensor and a photoelectric smoke sensor is used to measure three main parameters when a fire occurs. In terms of software, neural network is used to process the signals collected by various sensors, and then an alarm signal is sent by fuzzy decision-maker. This makes the detector have a certain intelligent function when outputting alarm signals. The experimental results show that the fire detector with this structure not only greatly improves the alarm accuracy, but also can further judge the fire state. It can be seen that the application of composite technology and neural network technology can greatly improve the performance of fire detectors, greatly reduce the false alarm rate and false alarm rate, and provide favorable conditions for early warning.
[References]
[1] Wu, the prospect of fire protection technology in the new century. China fire control encyclopedia, 2002 (1 1): 4-5.
Xu Chunyan, development and application of fire detection technology. Angang Science and Technology, No.9, 2000: 60-62.
Performance comparison and application of three kinds of fire detectors in Wang Xinjun. Tunnel construction, 2005,25 (3) 74-78.
Zhang Jiawei, Song. Principle and application of sensor. Harbin: Northeast Forestry University Press, 2003.99- 102.
Zhang Jianhua, Wang Zhanlin. Research on condition monitoring and fault diagnosis method of control system based on neural network. 1996 Proceedings of the Annual Conference on Control and Decision-making in China.
The present situation and development of multi-component fire detection technology. Fire fighting technology and product information 1998 04: 2 1-24.
Artificial neural network method for fire detection. Ergonomics.1997,3 (2): 39-41.
Shang Feng, Jiang Guoping, the application of DS 18B20 in fire detector. Electronic measurement technology 2003.
[9] Fan, home wireless intelligent anti-theft alarm system. Security technology, 2003. 3: 70-7 1
[10] Yang Zhenting, a popular practical subroutine of single chip microcomputer and its application examples. Xidian university Publishing House, 2002.83~86.
[1 1] Zhang, Peng Xiyuan, Jiang Shouda, Qiao Liyan. Newly compiled MCS-5 1 single chip microcomputer application system. Harbin: Harbin Institute of Technology Press, 2004.8 ~ 27,142 ~146,65436.
[12] Zhou hangci, single chip microcomputer application programming technology (revised edition). Beijing University of Aeronautics and Astronautics Press, 2002.56~72.
[13] Jiao Licheng, Neural Network System Theory. Xidian university Publishing House, 1995.
[14] Sui, editor-in-chief, application technology and examples of single chip microcomputer. Beijing: Electronic Industry Press, 2005+03 ~ 25,81~ 85.
[15] Shine Wong, computer control system. Beijing: Teaching Material Publishing Center of Chemical Industry Press, 2005.53~62.
[16] Shang Feng, research on compound fire detector [master thesis]. Dalian University of Technology. 2003.8。
[17] Cao Jun, design of fire alarm system [master thesis]. Harbin university of science and technology. 2006.3。
[18] British Standard, Fire Detection and Fire Alarm System-Part 5: Thermal Detector-Point Detector. BS EN 54-5:20065438
[19] James A. Milke, monitoring multiple aspects of fire signals to identify fire detection and fire fighting technology. Volume 35. Number three. 1999:25-29
20 Holt, Mike. Fire alarm signal system. Electrical construction and maintenance, 2003, 102(8):40, 42-43