Research On Autonomous Followup Tracking Of Noncooperative Spatial Target

Space-based spacecraft tracking is one of the essential issues on aerospace information processing and control, and important for successful implementation of space surveillance, autonomous rendezvous, satellite formation, on-orbit service and so on. The object of space target tracking is to provide dynamic state of target real-timely and sustainably, so as to support the decision made on target recognition, classification, catalogue, et al. However, space target tracking sometimes becomes difficult because of target increasing capability of non-cooperative or active anti-surveillance maneuvering, and long-term tracking of spacecraft is much harder to be realized. Although people have made wide researches and gained many useful results, it is still very difficult to develop an effective autonomous tracking method to achieve the accuracy of long-term tracking, especially for those non-cooperative targets with orbital maneuver capability.In this dissertation, to acquire the long-term accuracy, spatial target follow-up tracking approach is presented by introducing the chaser following-up scheme. The research is mainly focused on the space-based tracking and control method under the condition that space target has noncooperative maneuver capability. Based on that target maneuvers uncooperatively and that chaser satellite is follow-up controlled autonomously, the research is emphasized at three main parts: relative motion model, robust tracking algorithm and chaser follow-up control law. The primary contributions of the work are summarized as follows.1. A type of relative motion model based on orbital osculation is presented as the math description of follow-up tracking problem. First, based on the thought of orbital osculation, a transient relative motion model is built for considering the uncooperative target maneuver. It describes the relative dynamics between two maneuvering spacecrafts in enough small time spans. Such can avoid the unknown term of target maneuver and simplify the relative motion description with orbital maneuvers. Then, for solving relative motion uncertainty in follow-up tracking, we combined the osculation characteristics and reference orbit, built an osculating reference orbit (ORO) based transient relative motion model. Besides properties of transient model, the reference base of this model can timely vary with the chaser maneuver. Such can effectively restrain the magnification of linearized model error caused by motion uncertainty, and long-termly sustain the model accuracy in the environment of follow-up tracking. This performance can not be achieved by traditional models with static reference orbit.2. Considering the disturbance environment of tracking problem, especially the case that target probably maneuvers non-cooperatively, we developed two different tracking algorithms: uncoupled extended Kalman filter (UEKF) and redundant adaptive robust extended Kalman filter (RAREKF). In the UEKF, estimation results can quickly follow the disturbed system states by decoupling covariance of prediction error vector. Because disturbance information is not fully eliminated, if used in space target tracking, the algorithm can identify the variable of target maneuver. Boundedness of the tracking error is proved by the error structure analysis. RAREKF can adjust the redundancy to disturbance by the introduced redundancy factor. It recovers the adaptive switching function of filtering status, which is disabled in the traditional AREKF when modeling error and outer disturbance exists simultaneously. RAREKF not only normalizes the switching between robust and optimal filtering status, enhances the optimality of tracking result, but makes disturbance restraint be controllable so that only the disturbances exceed the redundancy range are eliminated. Sufficient condition of convergence is provided, and the algorithm stability is proved. Also, a different method of tracking error evaluation is presented. It considers effects of tracking accuracy from both tracking algorithm and model, can evaluate tracking methods with different relative motion models and filtering algorithms.3. A dynamic optimal sliding-mode control (DOSMC) is provided as follow-up control law of follow-up tracking problem. Sliding surface of this method has optimality because it is produced by a nominal optimal controller. Also, the sliding mode reaching law is self-tunable so that it can be guaranteed that the system state can reach the sliding surface within limited time. DOSMC combines the advantages of both optimal control and sliding-mode control. It has global robustness. Meanwhile, the fuel consumption is optimized. Besides that the chaser can be orbital maneuvered optimally and robustly, it can help to build a reasonable dynamic optimal sliding surface for the chaser attitude control. Such makes six-DOF follow-up control realized, and fully guarantees the measurement condition for sustainable space target tracking.4. Based on aforementioned ORO relative motion model, RAREKF tracking algorithm and DOSMC follow-up control law, a fundamental scheme of autonomous follow-up tracking of non-cooperative space target is presented. ORO relative motion model satisfies the model requirement of follow-up tracking. It can accurately and long-termly describe relative orbital motion with target maneuvered and chaser controlled. RAREKF algorithm provides accurate tracking capability for chaser to non-cooperative maneuvering target. The tracking results can help chaser to judge and implement the orbital follow-up control. Also, by transformation relation between relative motion state and chaser attitude, it can be applied to make decision on attitude control. DOSMC drives chaser to maintain the required measurement status and to guarantee the measurable tracking condition with reasonable price and control performances, such that the integrated function of follow-up tracking can be finally realized and work stably.