Research On Some Key Problems Of Aviation Emergency Rescue System

In the future,Low-altitude emergency rescue,as a kind of rescue and transportation mode with small space limitation and fast maneuvering speed,will gradually become the mainstream of disaster relief means in China,and emergency material demand prediction and low-altitude safe flight path planning are important links of low-altitude emergency rescue activities,which directly affect the effectiveness and safety of rescue operations.Firstly,this paper puts forward the whitenization weight function method and the improvement of the core GKNN fill algorithm used in pretreatment of disaster information,on the basis of perfecting the information after,by improving the GABP neural network model to predict demand of emergency relief supplies,the experimental results show that the improved GABP down 11.34% compared with the average prediction error of traditional GABP,after information pretreatment,up 44.7% average prediction accuracy of the model and the stability of 56.2% increase.Secondly,the wind field interpolation method based on slope is proposed.Compared with the traditional method based on elevation inverse distance interpolation,the wind field prediction of gentle slope is more accurate.Considering low-level wind and terrain obstacles influence on aircraft trajectories,put forward the safe path planning method based on fixed route node,and in wenchuan county,sichuan province,for example,to establish a low rescue simulation environment,analyses the complete path planning validation,the results show that the method in this paper,it is concluded that the planning track is not only meet the safety requirements,and have more advantages in the aspect of flight time and sailing journey.Finally,on the basis of the system developed by the research group,this paper added disaster information preprocessing function,low-altitude wind prediction function based on slope and low-altitude track planning function.Java-MATLAB hybrid programming technology is adopted to make the system combine the openness of Java with MATLAB's powerful numerical calculation and drawing ability,and have the advantages of cross-platform and visual image interface.