IoT Based Fire Escaping Algorithm And System Design

In recent years,great efforts have been made to lucubrate the Internet of Things(IoT)for its ability to obtain and process information of the physical world and its potential application prospects.Benefited from the wide usage of smartphones,it can connect the physical world and human behavior and thus can be used in broader scenarios.This thesis leverages IoT's advanced sensing,computing and communication technologies and the wide usage of smartphones to design fire escaping algorithm and system.The work in this thesis,which combines the advantages of IoT and the unique function of smartphone in human behavior detection,is a novel research direction and has wide application prospects and research value.When fire occurs in a building,the inner structure of the building,fire dynamics and crowd behavior are the main impacts on fire escaping.Existing fire escaping approaches plan fire escaping route based on real-time monitoring of fire hazard and assume that crowd density is known.Therefore,they do not fully consider fire dynamics and crowd behavior.To tackle the problem,this thesis designs an IoT based fire escaping system.The main contributions of this thesis are as follows:(1)This thesis designs a fire escaping route planning algorithm based on fire spread prediction so as to fully consider the impact of fire dynamics.It designs a fire spread model based on fire data generated by Fire Dynamics Simulator(FDS)to predict Available Egress Duration(AED)of each escaping route in the building.The length and AED of each escaping route in the building are considered in fire escaping route planning so as to provide efficient and safe fire escaping routes for people in fire scene.Experiment results confirm that the proposed algorithm outperforms existing algorithms by utilizing more escaping time.(2)This thesis designs a crowd density detection approach based on smartphone,which enables the crowd density to be considered in fire escaping route planning.It designs an E-grid(k)Quorum system and a Plain(k)Quorum system.Based on the two quorum systems,two crowd detection algorithms,namely E-grid(k)and Plain(k),are designed.Experiment results show that the proposed algorithms can alleviate communication congestion among smartphones and improve the performance of crowd density detection.(3)This thesis designs a feasible system construction plan so that fire dynamics and crowd behavior can both be considered in fire escaping system.It also constructs a network containing TelosB nodes,Android smartphones and Arduino suite to provide network infrastructure for fire escaping system.