| As the civil aviation industry has developed rapidly,safety and travel comfort become two factors of the cabin environment detection.Once there is a fire in the cabin,the environment will be greatly polluted and seriously threaten the passengers and flight safety.At present,the monitoring of the fire parameters in cabin mainly uses single sensor detection technology,which has a serious problem of false alarm.With the development of wireless sensor network,using various of types of sensors to monitor the fire parameters can effectively solve the problem of false positives.However,due to the limited energy of the sensor nodes used for information collection in the network,the life of the network is an element which influence the property of the detection system.With the purpose to figure out the related issues,there are solutions from three aspects:Firstly,the sensor nodes can be divided into clusters through ameliorating the K-means,which provides a network foundation for the subsequent distributed estimation of the fire parameters and the extension of network service life.Then,a new Event-Triggered Kalman Consensus Filter Algorithm Based on the Information Freshness was designed to estimate the fire parameters by calculating the Age of Information(Ao I)of the sampled data.This method lessens the transmission of outdated information and decreases the energy expenditure through comparing the Ao I of each packet with the minimum average Ao I of the network,the information is determined to be transmitted,saved or discarded.Finally,aiming at the problem of high false alarm rate when a single sensor is used to detect fire event in cabin,a multi-sensor data fusion method based on ameliorated Analytic Hierarchy Process is proposed to make a one-time decision on the fire situation in the cabin,and then fuse each decision value to calculate the probability of fire event by constructing the augmented support degree matrix.Finally,by comparing the fusion probability with the fire alarm threshold probability to judge whether there is a fire event. |
