Research And Implementation Of Fire Evacuation System Based On IoT

Fire is one of the most frequent disasters,which causes huge property losses and casualties every year.With the development of economy,there are more and more high-rise buildings and shopping malls,which bring pressure to the fire evacuation of public places.The traditional fire emergency system can not adapt to large-scale crowd evacuation,Therefore,it is necessary to build an intelligent fire evacuation system.This thesis studies the application of Internet of things and path algorithm in fire evacuation system,designs a system which use Internet of things and can guide the evacuation of people intelligently according to the location of fire point.The system is mainly used for fire evacuation in complex public places such as shopping malls,office buildings and teaching buildings.In the aspect of communication,Lora communication is used between the terminals of the system.A Lora gateway is designed for data exchange between Lora LAN and remote computer.The system control program runs on the computer,when a fire occurs,the operator can know the fire information,evacuation path information and the connection between each terminal and gateway in a remote and safe environment.LoRa gateway is developed on Raspberry Pi platform,which runs Linux operating system,and the gateway uses TCP/IP protocol to communicate with computer.In the aspect of evacuation path planning,this thesis analyzes the characteristics of several commonly used algorithms combined with the actual application scenarios of the subject,and simulates some algorithms.It is concluded that Dijkstra algorithm is more suitable for the scenario of the subject.Aiming at the problem that Dijkstra algorithm needs to recalculate all paths for fire spread,a scheme integrating Dijkstra algorithm and A*algorithm is proposed,When a fire occurs,Dijkstra algorithm is used to complete the whole area evacuation path planning.When the fire spreads,A*algorithm is used to complete the path planning of the affected area.There are multiple exits in the actual scene.In view of the fact that there are a lot of repeated calculations in the path planning of Dijkstra algorithm to multiple exits,Dijkstra algorithm is optimized to improve the efficiency of path planning.Considering the influence of temperature,smoke concentration and the number of people on the evacuation,the algorithm is further optimized by changing the corresponding path weight to increase the evacuation effect.After the completion of the path planning,the system will send the corresponding direction to the sign according to the calculated path,guide the crowd to evacuate through the sign,and send and receive information through the Lora module.The fire detection device,indicator board are controlled by MSP430F149 single chip microcomputer,and a 1500 mAh lithium battery is used as the backup power supply.When the power is off,it will automatically switch to lithium battery power supply,and the system can work normally for more than 2 hours after the power is off.The fire detection device uses smoke sensor,temperature sensor and humidity sensor to detect,so as to avoid false alarm caused by environmental factors.The decision tree algorithm is used to train the data collected by the sensor to get the rules for judging the occurrence of fire,and the system judges whether to alarm according to the rules.The LED indicator board is composed of several dot matrix,and the size of the indicator board can be adjusted according to the actual situation.The system is tested,and all parts of the system can maintain stable communication.When a fire occurs,the system can quickly detect the fire and complete the evacuation path calculation within one second.