| In recent years,the autonomous computing capability of space-based resources and the inter-satellite real-time communication network have been continuously developed,which has greatly enhanced the feasibility of low-middle orbit satellite constellation for tracking space moving targets.However,this also brings new challenges for satellite resource scheduling:(1)The high-speed moving characteristic of space-based moving targets makes real-time tracking of satellite necessary;(2)The three-dimensional characteristic of space-moving targets requires to stereo positioning,which requiring at least two satellites resources cooperate to complete the tracking;(3)The uncertainty of the state of the space moving target(such as the increase or decrease of the target)requires the fast response and real-time adjustment to targets.Aiming at the above-mentioned space-oriented moving target tracking problem,this paper conducts research from the following aspects:First,the space moving targets studied in this problem are defined,and then the characteristics of the space-oriented moving target tracking problem are analyzed,which are mainly reflected in the need for stereo positioning,resource relay collaborative,and rapid response.According to the description of the problem,a single-satellite autonomous online scheduling model is established by using Markov Decision Modeling(MDP),and the idea of Decentralized-Markov Decision Modeling(DEC-MDP)is introduced to realize the modeling of multi-satellite autonomous collaborative scheduling.On this basis,proposed the autonomous cooperative scheduling of satellites resources framework based on a closed-loop feedback structure and designed a corresponding scheduling process.Secondly,because of the high timeliness requirements of space moving target tracking in the uncertain and complex environment of onboard dynamics,an online hybrid trigger scheduling mechanism based on periodic rolling method and event dynamic triggering is proposed to serve the online real-time response of space moving targets onboard track.At the multi-satellite level,by comparing the advantages and disadvantages of the previous centralized and distributed multi-satellite autonomous collaboration architectures,a multi-satellite autonomous collaboration architecture based on the blackboard model is constructed.As a multi-satellite negotiation and communication mode,the blackboard model can efficiently coordinate inter-satellite tasks and command communication.Thirdly,to timely respond to the uncertainty of the tracking target,the multi-target patrol algorithm based on the dynamic priority is used to realize the real-time decision of the space moving target tracking,which guides the improvement of satellites resource utilization efficiency.For large-scale target scenarios,a group target aggregation algorithm is designed to aggregate and track targets in the same field of view to improve tracking efficiency.The sharing strategy is used to solve the tracking schedules,and the features that affect the tracking are extracted to estimate the score.This strategy solves the problem of inconsistency of information between satellites in online scheduling of satellites under distributed conditions,and realize fast scheduling of space moving target tracking.Finally,a simulation experiment is carried out on the research content,and the experimental results are evaluated from both qualitative and quantitative aspects.The results of simulation experiments show that the architecture,mechanism,and algorithm proposed in this study can effectively solve the problem of autonomous collaborative scheduling of satellite resources for space moving target tracking,meet the needs of emergency response,online and real-time tracking,and achieve continuous tracking of space moving targets. |