Research On Planning Method Of Low-orbit Platform For Closed-loop Detection And Tracking

The space-based low-orbit optical detection platform has the advantages of wide coverage,high imaging resolution,strong maneuverability and low launch cost.It has become an important means for the detection and identification of weak and small targets in the air.On-board autonomous closed-loop detection and tracking of low-orbit satellites is a key technology to achieve early target detection,real-time detection,and identification and confirmation.In order to meet the needs of the on-board autonomous closed-loop detection and tracking task,it is necessary to combine the current position of the target,the satellite attitude and orbit status to make a reasonable scheduling for the satellite observation path.This article is to face the needs of this task,carry out the target position correction for closed-loop detection and tracking,the task path scheduling method under the satellite fixed-point staring imaging working mode,and carry out the corresponding simulation experiment verification.The main research contents are as follows:(1)Target position correction algorithm for detection and tracking.In the process of target detection under the low-orbit detection platform,the target,background and platform motion are coupled,resulting in changes of the target and background motion characteristics on the image plane,which affects the detection process.Therefore,it is necessary to study the target position correction algorithm in the task and One of the purposes of the path scheduling is also to reduce the target position correction error.This thesis first analyzes the target characteristics in the low-orbit fixed-point gaze imaging mode,and establishes a correction model based on the analysis results,designs an algorithm to correct the target position,and carries out experimental verification.Experimental results show that the algorithm designed in this thesis can effectively complete the correction task,and the correction error under a given platform does not exceed 2 pixels.(2)Path scheduling algorithm for multi-platform formation observation scenarios.In the multi-platform formation observation scenario,since the number of gaze points to be observed is more than the number of satellites,it is necessary to design a path scheduling algorithm according to mission requirements to allocate imaging area,imaging time and order.Based on the existing evolutionary algorithm,this thesis studies the model solving method by designing the objective function and improving the coding method,establishes a path scheduling algorithm for multi-platform formation observation scenarios,and carries out experimental verification.Experiments show that the algorithm can complete the observation path scheduling task within 10-20 seconds under the premise of reducing the correction error as much as possible and the coverage rate as high as possible.(3)Path scheduling algorithm for quick response tasks.Satellite marshalling may be used to perform rapid response tasks,and it is necessary to optimize the path planning algorithm for the problem of insufficient on-board processing resources and incapability of rapid response in orbit.From the perspective of limiting the scale of the solution space,this thesis designs a single-satellite path scheduling algorithm based on dynamic programming and an optimal staring point allocation algorithm,and then proposes a path scheduling algorithm for rapid response tasks.The experimental results show that the algorithm can greatly reduce the running time,and the time consumed under the same test platform is reduced by about 2 orders of magnitude.In addition,the observation path income obtained by the algorithm is lower than 10% compared with the evolutionary algorithm,which significantly improves the application capability of the algorithm on the satellite.