Reseach On Control For Autonomous Close-Range Rendezvous And Docking With A Spatial Tumbling Target

Over the last decades,with the rapid development of space science and technology,on-orbit servicing technology has become an important research direction for the new commanding heights of the future space,and has attracted widespread attention from various countries and organizations around the world.Autonomous rendezvous and docking with a spatial tumbling target,which is of vital significance in space science and application,has been regarded as one of the key technologyies for many on-orbit servicing mission,such as space debris removal,failed satellite maintenance,on-orbit refueling and so on.Different from the traditional cooperative spacecraft,the tumbling target is generally out of control,and can not cooperate with the active spacecraft for autonomous atttidue control,which puts forward higher requirements for the autonomous proximity operations.In this dissertation,several robust and reliable control strategies are investigated for autonous proximity and docking with a tumbling target with consideration of input saturation,parametric uncertainties,external disturbances,and motion constraints,et al.The achievements and contributions are summarized as follows:(1)Robust adaptive autonomous rendezvous and docking control for the close-range relative motion system under input saturation and effectiveness loss.It is well known that input saturation and effectiveness loss characteristics are typical input nonlinearities in practical space missions,which will seriously affect the execution capability of the active spacecraft and the reliability and safety of the proximity operations.In addition,in the process of space maneuvers and operations,the model parameters(i.e.,mass and inertia matrix of the active spacecraft)will change due to fuel consumption and payload variation,which poses a serious challenge to the design of control strategies.In order to overcome the influence of the above factors,a model-free autonomous rendezvous and docking control law based on sliding mode variable structure method and adaptive control technology is proposed.The proposed control law does not depend on accurate model parameters,and can realize the requirements of attitude synchronization and relative position tracking in the presence of parameteric uncertainties,external disturbances,and input saturation and effectiveness loss.(2)Robust output feedback finite-time prescribed performance control for autonomous proximity and docking with a tumbling target with consideration of lack of speed-level state information.In practical proximity operations,the speed-level states(i.e.,relative angular velocity and relative velocity)are usually difficult to obtain due to on-orbit sensor failure and/or heavy meansurement noise.To deal with this problem,a linear high-gain observer is utilized to estimate the unmeasured relative angular velocity and relative velocity.Based on the measured and estimated state information,the Gaussian radial basis function based neural networks are investigated to approximated the unknown nonlinear term containing the unknown model parameters,and the optimal weight matrix is estimated by the well designed adaptive law.In addition,with the aid of the finite time presecribed performance function,the transient and steady-state convergence performance of the relative attitude and relative position can be predesigned during the whole rendezvous and docking process.The proposed control law can realize the finite time attitude synchrnozaiton and relative position tracking requirements,and without using model parameter information.(3)Autonomous safety control for spacecraft rendezvous and docking by integrating the core idea of sliding mode variable structure method and artificial potential function technique with consideration of dynamic motion constraints.In practice,there always exists some motion constraints during the close-range rendezvous and docking process,such as collision avoidance constraint,line-of-sight constraint and so on.In this dissertation,a cardioid-based based three-dimensional nonconvex surface is used to describe the collision avoidance constraint,then a three-dimensional cone is used to describe the line-of-sight constraint.On this basis,two virtual artificial fields corresponding to the abovementioend motion constraints are constructed by utilizing artificial potential function.After that,a novel sliding manifold is designed by combining the sliding mode method and artificial potential technique.Meanwhile,for the close-range proximity operations,a nonlinear disturbance observer base safety control law is first proposed for the case of knowing model parameter,and an adaptive safety control law is then proposed for the case of existing model parameter uncertainties.The proposed control law can not only achieve the rendezvous and docking mission,but also ensures the aforementioned montion constraints are always satisfied during the whole process of rendezvous and docking.(4)Model predictive control based autonomous safety control strategy is proposed for spacecraft rendezvous and docking in the presence of multiple constraints.In the real rendezvous and docking process,there are several practical constraints,such as input saturation constraint,collision avoidance constraint,and approaching velocity constraint,which are difficult to slove by the traditional nonlinear control methods.After that,within the framework of model predictive control method,the abovementioned constraints are transformed into a unified linear inequality constraint about system input and states,and a quadratic performance index is constructed about the tracking error and fuel consumption.On this basis,the original model predictive control problem is transformed into the quadratic programming procedure,which can be sovled by the existing algorithms easily.The proposed control strategy can not only ensure the chaser spacecraft to track the desired docking point of the tumbling target,but also provides good robustness due to the finite receding horizon optimation.