The Research And Implementation Of Fusion Technique For Aerial Surveillance Information Based On SVR

Safe and reliable air traffic control relies on reliable aerial surveillance information,but aerial surveillance information is not always reliable and highly accurate.If aerial surveillance information obtained by the air traffic control center is not accurate,it may lead to serious consequences.Common sources of aviation surveillance information at this stage include secondary surveillance radars,ADS-B.The original aerial surveillance information cannot be used directly since it contains various system inherent errors and measurement errors,resulting in its accuracy not meeting the corresponding requirements.Taking the secondary surveillance radar as an example,its surveillance data must be processed by Kalman filter before it can be used.This paper proposes a multi-source radar data fusion method based on the SVR algorithm.This method uses multi-radar data to fuse to obtain higher confidence results to reduce radar measurement errors.After fusion,the average error can be reduced by about 40%,and the effect is better than the Kalman filtering algorithm.This paper uses the automatic parameter adjustment process of machine learning to replace the complex manual parameter adjustment process of traditional Kalman filter.And the entire process does not need to know the radar instrument measurement noise and other system inherent errors.This paper also proposes a single radar data filtering method based on SVR algorithm.The single radar filtering problem degenerates from the multi-source radar data fusion problem.The purpose is to filter out measurement errors in single radar measurement data.After filtering using the SVR algorithm,the error can be reduced by about 20%,which is better than the filtering effect of the Kalman filter algorithm.Based on the above two modeling methods,this paper establishes an aerial surveillance information fusion system,which integrates aerial surveillance information data processing,automated modeling,and simulation testing functions.It can process large quantities of original radar data and ADS-B data and generate corresponding models,and finally verify the validity of the model in a simulation environment.