Study On State Estimation Techniques Of Space Targets Based On ISAR Sequential Imaging

Space target state monitoring technology aims to obtain accurate and real-time attituede,three-dimensional structure and motion information of on-orbit targets,so as to analyze their intentions,predict their situation development,and avoid potential space collisions.It is the core technology in the field of current space situation awareness.Ground-based Inverse Synthetic Aperture Radar(ISAR)is an important way to obtain observation data of space targets as it works in all-weather and all-weather conditions.Beside tracking on-orbit targets by transmitting narrow-band signals,ISAR devices can also transmit wide-band signals to obtain high-resolution two-dimensional images,which achieves the integration of target tracking and imaging and support the analysis of its state.With the intensification of space confrontation among countries,research on space target attitude measurement technology for complex space environment is of great significance to the development of national space situation awareness system and maintenance of the healthy operation of space assets.The existing research on ISAR imaging and state monitoring of space targets focuses on high-resolution imaging and general motion state evaluation.Most exsiting methods of target state estimation depend on the feature matching between the target observation data and accumulated observation database.This sort of methods lacks the support from the observation physical model,leading the limition of the performance in practical application scenarios.It is an urgency to explore the internal connection between the on-orbit state and ISAR observation characteristics to achieve the state estimation of space targets with the support from the physical mechanism of radar imaging.To satisfy the growing requirements on numerous applications,this dissertation studies new techniques for the state estimation of on-orbit space targets from three aspects,i.e.radar signal processing,ISAR image interpretation,and optical-and-radar information fusion.The relevant work is supported by the several Chinese national programs,such as National Science Foundation of China,Defense Advance Research Program of Scienceand Technology of China and National Programs for High Technology Research and Development of China.According to the complexity of the problem,this paper studies the state estimation of on-orbit space target in the stable-attitude and spin conditions successively.The specific research content can be summarized into the following four parts:1.Target attitude estimation based on projection features of ISAR sequence imagesThe first part studies the geometric characteristics of ISAR high-resolution sequence imaging observations of space targets.From the perspective of the inversion of ISAR projection characteristics,a new target attitude estimation method based on ISAR sequence images is proposed.By analyzing the continuous changes of the target projection characteristics in the ISAR observation image sequence,the accurate parameters of the target on-orbit attitude is estimated.In this way,it voids the difficulty of associating the scattered point features in the long-term ISAR observation image sequence in the traditional methods.2.Target attitude estimation based on the quadratic phase features of ISAR imagesIn the second part,a new target attitude estimation method based on the quadratic phase feature of the ISAR image is proposed.A signal model is established to describe the defocusing of the target in the observation images,and an explicit expression is derived between the quadratic phase feature in ISAR image and the target attitude parameters.Quadratic phase terms are usually regared as the negative factor and compensated within the phase adjustment for the sake of improving the focal performance of ISAR imagery.While it is developed for space target attitudeestimation in this work,which provides a new perspective of ISAR image interpretation in related studies of radar imaging and image interpretation.3.Dynamic estimation of spin targets based on Multi-station ISAR imagesIn the third part,on the basis of the research for attitude-stable space targets in previous parts,the dynamic estimation of spin space targets is further discussied.The core of this problem lies in how to solve the coupling between target spin motion and ISAR imaging geometric changes.By introducing the multi-station ISAR joint observation,the angle resources of radar observation are supplemented,the decoupling between target spin motion and ISAR imaging geometric changes is achieved by solving a two-stage independent optimization.The range dimension information of multiple-station ISAR images is utilized to determine the target instantaneous attitude while the Doppler dimension information of multiple-station ISAR images is employed to estimate the spin parameters of the target.An explicit expression of the relative motion between the spin in-orbit satellite and radar is also built with the usage of the quadratic phase feature of multi-station ISAR images.It avoids the influence of the ISAR image projection characteristics and the observation imaging plane changes during target dynamic estimation,which also verifies the advantages of multi-station ISAR joint observation in space applications.4.Optical-and-radar image fusion for dynamic estimation of space targetsThe fourth part draws on the multi-view image scene reconstruction technology in the field of computer vision,and compares the geometries of optics and radar imaging to explore target three-dimensional information interpretation based on the optical-and-radar feature fusion under the same place.At first,it catches the complementary information of the optical and radar image under the same observation angle,and a dynamic estimation method for spin targets is proposed based on the feature fusion in optical-and-radar images.In this way,the core problem of ISAR imaging plane uncertainty is carried out.Then,the similarities of the optical and radar multi-view imaging observation scenes are compared,and a new target 3D reconstruction method based on the multi-view ISAR image sequence is proposed.It is extended to the optical-radar joint observation mode to solves the limition of viewing angle in single-station single-turn ISAR observation for space targets.The optical-and-radar observation mode in this part breaks through the limitation of the single-sensor viewing angle in space target observation,and provides a new perspective for the studies of multi-sensor cooperative mode in space situation awareness network.