Font Size: a A A

Research Of High-Speed Train Fire-Alarm System Based On Fuzzy Neural Network

Posted on:2019-02-12Degree:MasterType:Thesis
Country:ChinaCandidate:C K NieFull Text:PDF
GTID:2371330563992254Subject:Software engineering
Abstract/Summary:PDF Full Text Request
With the development of society and science and technology,high-speed train has become more and more popular because of its convenience and rapidity.Due to the environment of high-speed train is closed and crowded,which is not conducive to the safety when fire,the fire will pose a serious threat to the safety of life and property of the occupants.In addition,in order to reduce the harm of fire as much as possible,we need a series of control to train when the fire happens.If there is a false report,it will cause unnecessary losses.Therefore,high speed trains have high requirements for the fire alarm system,which not only has high sensitivity to fire,but also ensures low false alarm rate.By analyzing the characteristic of fire,the principle of fire detection and the environmental characteristics of high-speed trains,a fire alarm system based on fuzzy neural network fire detection algorithm is proposed.The system uses 2-bus as the connection bus of the fire detector.It detects three kinds of fire characteristics including CO,temperature,smoke concentration,and input those characteristics to the fuzzy neural network through serial port communication.The fuzzy neural network consist of three parts which are fuzzy controller,neural network and decision layer,the fuzzy controller and the neural network processing those three fire characteristics signal respectively,and both of them will input its result to decision layer for making final decision.The high sensitivity and low false alarm rate of fire detection are realized by using the learning and generalization ability of fuzzy neural network.The neural network is verified by 32 groups of data after it been trained,and its final MSE is 0.005773.Then I chose 5 groups of data from fire,no fire and smoldering data sets respectively to verify the decision layer.The validation result shows that for three fire situations including no fire,smoldering fire and open fire,the fire detection algorithm proposed in this paper can identify them goodly without false fire alarm.Therefore,in general,this fire alarm system can realize the fire protection demand of high speed train and ensure the safety of passengers.
Keywords/Search Tags:high speed train, fire alarm system, fuzzy neural network, fire detection
PDF Full Text Request
Related items