Study On Angles-Only Relative Navigation And Maneuver Planning In Space Debris Removal

Angles-only relative navigation(AON)provides a low-mass,power saving,and low-cost solution for active space debris removal(ADR),on-orbit service,non-cooperative rendezvous,and formation flying,which are in the phase of conceptual research and in-orbit demonstration.In background of ADR,camera pointing control in in-orbit debris optical detection,AON,optimal maneuver planning for AON,as well as ground validation proposal design are systematically studied in this thesis.The main achievements are as follows.Covariance analysis based in-orbit camera pointing control strategy of space debris optical detection is studied.The basic in-orbit optical target detection model is built,and the quantitative indices of detection efficiency are defined.Based on the analysis of common pointing control strategies,linear covariance analysis method is employed for prior relative state error prediction,and the potential detection areas are obtained.Two camera pointing control methods are designed,which are stepping sweeping mode and double harmonic sweeping mode.High fidelity simulation shows that the stepping sweeping strategy can achieve 99.9% detection rate of the two-line elements(TLE)accuracy level debris in four minutes.The robustness of detection method against camera performance parameters and task constraints is analyzed.The improved initial relative state determination model and corresponding estimation error model are established for AON.To improve the initial relative state estimation accuracy for AON with increasing observation information,an extended vector including initial relative state is defined,and the least squares estimation model of the initial relative state is constructed.Monte Carlo simulation model is established and simulation test is carried out,the results show that the state estimation accuracy is improved by about 25%.Considering of the errors exists in camera installation,coordinate rotation and angle measuring,the state estimation error model is deduced.The validity of the estimation error model is verified by simulation experiments,and sensitivity analysis is carried out,and the relevant conclusions are helpful to guide the design of observation scheme.An AON method applicable for the whole process of relative navigation is established.Camera observation models are established according to relative distance,and the corresponding observation sensitive matrices are deduced.It is proved theoretically that camera offset in radial and normal can improve the observation degree of close range V-bar relative position keeping,but has no influence on far range navigation accuracy.The square root unscented Kalman filter is used instead of the traditional extended Kalman filter,simulation results show that the navigation error is reduced to half,and the computational load is reduced by about 10%.Through two analysis means,i.e.Cramér–Rao lower bounds(CRLB)and Monte Carlo simulation,the efficiency of navigation algorithm is verified,and the influence of illumination condition,camera parameters,and the relative motion trajectory on navigation algorithm accuracy is analyzed.The optimal maneuver assisted observation planning and closed-loop optimal robust planning methods are proposed.New observability criteria for maneuver assisted AON is established,based on the criteria,a quantitative index is defined to characterize the observability improvement.Considering the fuel consumption and task constraints,an optimal maneuver planning method taking both observability and fuel into account for AON is established.The real time control error is introduced into the maneuver planning through closed-loop control state covariance analysis,considering the fuel consumption and task constraints,in view of the shortage of traditional time triggered optimal maneuver planning,a two-layer optimal maneuver planning method based on time and covariance threshold triggered is established,simulation results show that the proposed method can significantly improve the trajectory robustness and the final state control accuracy is improved by one order.Hardware in the loop simulation system proposal of AON is designed.The requirement analysis of the hardware in the loop simulation system is carried out,and the functional requirements and the performance indexes of the hardware devices are defined.The overall design of hardware in the loop simulation system is carried out by using the existing nine degree of freedom motion platform,the function module,the interface and the interaction relation of the subsystem are designed,and the component selection is determined.According to the similarity principle,the similarity criteria of the system are deduced,and the similarity parameters of the ground hardware in the loop simulation system are determined.Simulation results show that all the data of the similar system simulation is proportional to the prototype system,which verifies the correctness of the similar system design.AON and maneuvering planning are studied in this thesis.The methods and conclusions proposed have important reference significance for the application of AON in future ADR,on-orbit service,non-cooperative rendezvous and formation flying.