Research And Practice On Fire Alarm Misinformation Optimization Based On BP Neural Network

False alarms occur in fire alarm system because of fhuman activities and external environmental factors and so on,which results in not only the waste of manpower and material resources but also the reduction the sense of urgency and crisis of people in fire.Therefore,it becomes very important to avoid fire alarms in the duty management system.The shortcomings of single sensor are uncertainty and limitation,and multi-sensor based information fusion technology processes information at very low efficiency because its algorithms are too complex.To solve the problem,this thesis proposes a BP neural network based algorithm to determine spatial distribution of detectors.The main contributions of this thesis are as follows:Firstly,this thesis analyzes various physical signals when fire occurs and proposes to use the fusion of smoke signal,temperature signal and flame signal to determine fire alarm,so as to reduce false alarm.Secondly,this thesis verifies the spatial positive correlation of different types of fire detectors.By using the Apriori algorithm to excavate the spatial rules of fire detectors,this thesis discovers the type association rules,the quantity association rules and the location association rules of fire detectorsThirdly,according to the spatial association rules of fire detectors,this thesis proposes a BP neural network based detector spatial layout optimization algorithm to reduce the complexity of the false alarm algorithm as well as the time of false alarm judgement.Based on the above research,this thesis conducts some effective practices:Firstly,this thesis constructs a fire alarm information database,which provides encoded addresses,spatial locations,alarm times,etc.of fire detectors for false alarm analysis and research.Secondly,this thesis applies the intelligent fire analysis algorithm to the duty management system and analyzes the operation results.In summary,this thesis solves the false alarm problem by theoretical analysis and practice.There are some works to be done in the future,such as more precise modelling is required when buildings crowd together,because the spatial location of detectors is relatively complex.