The Research And Implementation Of Fusion Technique For Aerial Surveillance Information Based On Random Forest Algorithm

With the great enrichment of material life,more and more people choose to travel by air,resulting in the continuous increase in the number of aircraft and the increasingly crowded airspace.Therefore,in the safe flight of more aircraft in a limited airspace,aviation surveillance information systems are becoming increasingly important.As an important part of the aerial surveillance information system,the aerial surveillance information fusion combines the track data obtained from multiple radars to track the same target to form a new,more accurate track.As a traditional aeronautical surveillance information fusion algorithm,the Kalman filtering has problems such as difficulty in obtaining accurate noise statistical characteristics and complex parameter adjustment process.The random forest algorithm has the advantages of simple construction,fast training speed,and high accuracy.Therefore,this paper proposes to use random forest algorithm to fuse aerial surveillance information.In this dissertation,the weighted random forest regression algorithm is proposed based on the application of random forest algorithm in aerial surveillance information fusion.By using the Kalman filtering algorithm and random forest algorithm for single-radar surveillance information fusion,and using random forest algorithm and weighted random forest algorithm for multi-radar surveillance information fusion,this paper designs and implements an aerial surveillance information fusion system.The system completes the functions of monitoring information preprocessing,fusion and fusion result display.On the current data set,this thesis carries out single-radar surveillance information fusion and multi-radar surveillance information fusion.In single radar surveillance information fusion,the prediction accuracy of the random forest algorithm is significantly better than the Kalman filter algorithm.In the fusion of multi-radar surveillance information,the fusion result of the weighted random forest algorithm is slightly better than the fusion result of the random forest algorithm.The error of both algorithms is about 80 meters,indicating that the random forest algorithm and the weighted random forest algorithm are effective.