Research And Application Of High Confidence Fusion Technology For Multi-source Target Information Features

The right to control air space is the most critical battlefield in modern high-tech wars.The party who has the right to control air means that he has the initiative to fight.In order to achieve the super information advantage of the latest generation of military aircraft,the author participates in and develops some functions in the high confidence system of multi-source target information characteristics of airborne platforms.Based on JDL data fusion framework,this system fuses measurement information from distributed heterogeneous sensors,combines linear regression,fuzzy mathematics,evidence theory and other technologies,and studies and applies the key function points involved in multisource target information feature high confidence system based on the performance of airborne platform.In this thesis,the importance of high confidence system of multi-source target information characteristics,the research history and current situation at home and abroad are summarized,the research objectives and contents is put forward.Then,the related theory and technology in the system are introduced.Conduct module-based functional and non-functional requirements analysis of the system.Conduct a general design of the system,including the overall framework,system architecture design,module design,data flow design;Then,the key functions in the system are designed and implemented in detail,with emphasis on single-machine passive positioning module,track association module,and intention estimation module.Finally,the test results of each module are shown,and the superiority of the design is explained.The main work of this thesis is as follows:1.A single-machine positioning method based on angle sensor is presented.Passive angle sensors can be used for reconnaissance of special airborne platforms.Using least squares method,sliding time window,Gaussian filter and other technical means,locate the position of the air target,and output the latitude and longitude information of the target.Due to the need for real-time fast positioning in real tasks,a new function to determine convergence is added to accelerate the convergence of target location information.2.A comprehensive track association method based on type information and motion information is presented.Most of the traditional track association methods only apply to motion information,but ignore the type information.In this thesis,the characteristics of type information and motion information is combined.And further extracts the degree of association of motion information using grey correlation,describes the degree of association of type information using Wasserstein distance,and finally combines the above two degree of association with D-S evidence theory to propose a comprehensive degree of association.3.An intention estimation method based on target location is presented.With the development of sensor technology and the performance improvement of airborne platforms,airborne platforms can use location information to provide pilots with more flight aids.The intent estimation method proposed in this thesis can calculate the intent of a target based on its short-time location information and give the corresponding confidence level.