Research On Multi-sensor And Multi-target Track Association And Fusion Algorithm

With the diversification of the battlefield environment,the diversification of information jamming methods,the expansion of the operational range,and the enhancement of the precision strike capability of weapons and equipment,the monitoring performance of the battle command center and the viability of the battlefield monitoring system are put forward a severe test.In the current situation of complex electronic warfare battlefield environment,the single sensor monitoring system has unique data source and limited sensor tracking accuracy.If it is detected by the enemy’s electronic monitoring system and carries out signal jamming or force attack on our sensors,the performance of the monitoring system will decline rapidly and even be destroyed by the enemy.The problem of insufficient survivability will put us at a disadvantage.Therefore,by integrating the advantages of multiple sensors and the target information obtained from multiple platforms,we can obtain a more detailed and complete battlefield situation,solve the problem of poor reliability of single sensor monitoring system,and improve the performance of the command and control system.In the middle and late 20 th century,under the military requirements,the research on multi-sensor data fusion technology began to start.The research results were applied to the military fields such as battlefield target detection,target recognition and target track processing,providing a basis for the follow-up research and development of track association and fusion.In the complex battlefield environment,there are many types of fighters from both sides to form a multi-target battlefield environment.Multi-sensor system is to integrate the advantages of multiple sensors,such as laser inertial navigation,surveillance radar,Beidou satellite,ground command platform,etc.,to observe the status of multiple targets from multiple different positions,and then send the target information to the system center for processing,making full use of the advantages of each sensor to obtain more accurate targets More comprehensive information.Compared with single sensor,multi-sensor increases the reliability of track fusion system and extends the time and space range that the monitoring system can detect.In this paper,multi-sensor track association and track fusion are studied under the complex battlefield background of multiple targets.The data fusion algorithm generates detailed information such as target position,motion characteristics,attribute characteristics,etc.by associating and fusing the same or different multi-sensor track data.Firstly,the data preprocessing technology is briefly introduced,including time registration,space registration,outlier removal and other operations.The improved Lagrange difference method is used for time registration,and the Kalman filter method is introduced into the process of time registration.Secondly,in order to solve the problem of multiple sensors tracking the target in the coverage area in the same space,and determine whether the target measurement data from different data sources correspond to the same target entity,the method of combining gray correlation technology and Hausdorff distance is used to determine whether the target track is associated with the system track by comparing the similarity and similarity between tracks.Track correlation technology is the premise of track fusion.The fusion algorithm in this paper is based on the weighted track fusion technology.The determination of the weight in the fusion algorithm is very important,and the result of dynamic calculation of track weight using track quality is more reliable.A dynamic weight distribution method based on track quality is proposed for track fusion.Finally,use Matlab software to simulate the track association algorithm and track fusion algorithm.Through the analysis and comparison of the classical fusion algorithm and the algorithm in this paper,the results show that the algorithm in this paper has a higher accuracy.