Adaptive feature extraction of general aviation forest fire rescue data
The flight of aircraft in general aviation forest fire rescue is affected by factors such as weather and terrain.In order to ensure rescue efficiency and flight safety,it is necessary to refer to objective data from fire scenes for rescue dispatch decisions.In order to address the problem of poor adaptability between objective data of fire scenes and rescue plans in general aviation forest fire rescue,the influence factors of the adaptability were analyzed based on the characteristics of general aviation forest fire rescue with multi-aircraft types.A"distance-radius-velocity"general aviation fire rescue dispatch data extraction model with multiple fire scenes and multi-aircraft types was established,the input of which was composed of training data and test data obtained by the Monte Carlo method.In order to reduce the number of distance features,a method for comprehensively characterizing the distance of the fire scenes was proposed.The feature extraction accuracy of combinations of different fire scenes was explored based on naive Bayes estimation and Gaussian Bayes estimation.The results show that the proposed calculation method of distance parameters can effectively express multiple distance features,the combinations of fire scene features composed of which have a small impact on the results under the unbalanced classification of the dataset.Gaussian Bayes method is used to extract features from 3 combinations of fire scene characteristics,and the prediction accuracy reaches 73.3％,70.2％,and 72.4％,respectively,with good results.
general aviation forest fire rescueadaptabilityfeature extractionmulti-aircraft typesrescue dispatch decisionBayesian estimation
万方数据
维普
