Research On Fire Detection Algorithm And Integration Technology Fused With Typical Feature Parameters In Civil Aircraft Cargo Compartment

With the continuous development of the civil aviation transportation industry,the number of fire accidents in the cargo compartment of civil aviation aircraft is also increasing.In order to ensure the safe and stable operation of civil aviation aircraft,it is important to identify the fire quickly and accurately in the cargo compartment of the aircraft and issue an alarm in time,which can greatly reduce the loss to the safety of people’s lives and property.Therefore,in view of the high false alarm rate of fire detectors,this paper conducts research on the key technologies of typical fire detectors,combines the improved algorithm training model,and designs a composite fire detection system based on smoke,temperature and CO.It aims to realize accurate identification and quick alarm when a fire breaks out in the cargo compartment of the aircraft.The main tasks are as follows:(1)Experiments were designed for two single-characteristic parameter fire sensors,smoke and CO,to analyze their advantages and disadvantages in fire detection in confined space environments.For the smoke characteristics of the fire,this paper uses the principle that different wavelengths of light have different scattering signals on smoke particles,selects a dual-wavelength photoelectric smoke sensor for data collection and uses a genetic algorithm to optimize the random forest for fire classification detection;for the CO concentration characteristics of fire,In this paper,the commonly used electrochemical CO sensor is tested for fire measurement accuracy,compared with the high-precision flue gas analyzer,and combined with PSO-LSTM to train the CO concentration compensation model to reduce the measurement error of the sensor.(2)Based on the research and analysis of two typical fire characteristic parameters,the shortcomings of single-parameter fire detection are summarized,and dual-wavelength photoelectric smoke sensors,electrochemical CO sensors and temperature sensors are selected as characteristic data acquisition terminals to study multi-sensor Integrating the fire detector hardware integration technology,a composite aircraft cargo cabin fire detection system is designed,which has the functions of real-time monitoring,data query and playback,etc.(3)Using multi-sensing fusion fire detectors to carry out fire detection experiments under simulated aircraft cargo compartment pressure changes,research and analysis of typical combustibles CO concentration,infrared light and blue light PTR,and the change of combustion characteristics of Sauter average particle size,and The improved BP-Adaboost algorithm is used to train the fire detection model to achieve high-precision discrimination of complex environments and fire conditions.