Analysis Of Dynamic Characteristics And Research On Precise Control Method For Disturbance Free Payload Spacecraft

The Disturbance-Free-Payload(DFP) spacecraft is a new class of spacecraft for payload vibration isolation.The spacecraft structure divides the entire spacecraft into a payload module(PM) and a support module(SM).PM and SM are connected through a DFP interface which includes non-contact actuators and non-contact sensors.Non-contact actuators and non-contact sensors can make PM and SM close to each other but have no mechanical contact,thereby realizing the micro-vibration isolation and precise pointing for PM.In engineering applications,the non-contact actuator is an electromagnetic actuator whose back electromotive force(back-EMF) will cause the coupling between PM and SM and transmit the influence of vibration on SM to PM.In addition,there may be connection cables between SM and PM used for communication and power supply which also can cause coupling between PM and SM.In this paper,the back-EMF of the non-contact actuator and the stiffness of cables are considered to research the DFP spacecraft.The coupling characteristics between PM and SM caused by them are analyzed,and the back-EMF and uncertainty of PM moment of inertia are considered to study the precise control problem of PM.Focused on the DFP spacecraft with a hexapod cubic DFP interface,the forces and torques acting on PM and SM by the DFP interface are analyzed without the back-EMF of non-contact actuators and the stiffness of cables.The dynamics model of the DFP spacecraft is established,including PM attitude dynamics model,the relative translation dynamics model of PM relative to SM and the relative attitude dynamics model of SM relative to PM.The PD control system of the DFP spacecraft with three-loop architecture is designed.The superior performance of the DFP spacecraft is verified by establishing simulation model.This part is the basis of the subsequent research.The hexapod cubic DFP interface is essentially a six-degree-of-freedom platform,and its dynamics model is established by combining the Lagrangian method and the Newton-Euler method.Considering the back-EMF,the drive force of the non-contact actuator is given.Considering the stiffness,the action forces of the connecting cables is given.The drive forces of non-contact actuators and the action forces of the connecting cables are introduced into the dynamics model of the DFP interface to obtain the coupled dynamics model between PM and SM which is used to analyze the coupling characteristics of the DFP spacecraft in the time domain and the frequency domain.Time domain analysis shows that there is cross-coupling between PM and SM.The coupling caused by the back-EMF increases with the increase of the back-EMF coefficient,while the coupling caused by cables stiffness is related to the frequency of the vibration on SM.Frequency domain analysis shows that the presence of coupling makes the low frequency vibration on SM affect PM more obviously.The PM attitude dynamics model considering the non-contact actuator back electro-motive force and PM moment of inertia uncertainty is an uncertain nonlinear system.To realize the attitude precise control of PM,a self-organizing HBF neural network adaptive control method is proposed which can be applied for the uncertain nonlinear system.The self-organizing HBF neural network is used to approximate the inverse model of the uncertain nonlinear system,and the output of this HBF neural network is used to design the control law.Aiming at the online learning of HBF neural network,an online self-organizing learning algorithm is proposed,which not only can adjust network parameters online,but also can add or remove hidden-layer nodes in the network according to the situation of the network.The proposed online learning algorithm can satisfies the high approximation accuracy and the simple network topological structures of the HBF neural network simultaneously.Considering the back-EMF of the non-contact actuator and the uncertainty of PM moment of inertia in engineering applications,a PM nonlinear attitude dynamics model with uncertainty and back-EMF disturbance is established.Based on the multi-output self-organizing HBF neural network,a self-organizing HBF neural network adaptive attitude controller is proposed,which is applied to PM attitude control to improve the DFP system.The performance of the self-organizing HBF neural network adaptive attitude controller is verified by simulation analysis,and the results show that the controller can realize high-precision attitude control for PM.