Research On Trajectory Tracking For Space Flexible Manipulator With Large Payload

Along with the development of space technology, plenty of human activities in the outer space both in and out the space capsule become more and more difficult and complex.As the space manipulator possessed the advantages like anti-jumping, precise reliability and so on, it could replace the astronauts to do most of the on-orbit services. But when it comes to the services of building, maintaining of the space station, the manipulator needs to capture and move the targets with large quality and big inertia which would increase the flexibility of it and then the end control accuracy also be impacted. Therefore, oriented on the on-orbit tasks with large payload, this research has the significant scientific meaning and theoretical value.The research contents of this paper derive from the project of National Natural Science Foundation "Research on Optimizing Path Planning of Space Manipulator for On-orbit Large Payload Operate Task". Aiming at the on-orbit space missions of large payload, this paper uses a7-DOF space robot as the research object, and focuses on its analysis of dynamic modeling method and trajectory tracking strategy. The main content of this paper includes the following aspects:Firstly, the flexible arm of the space manipulator would simplified into Euler-Bernoulli beam, Assumed-mode method is used to describe its vibration deformation and then the forward kinematics model is deduced; Based on the Lagrange equations and the effects of the end large payload, the dynamic equations of flexible space manipulator with large payload are established and influence of the end payload bring to the dynamic parameters is also analyzed. Secondly, based on the given dynamic model and the requirements of on-orbit mission, a series of optimization index are given and Riccati equation solved through backward recurrence. After that, a reasonable and effective weight matrix evaluation method is proposed. Then the dynamic planning and Hamilton-Jacobi equation are used to solve the problem of optimal control. Through this way, we realized the trajectory tracking research of space flexible manipulator with large payload under the given index and numerical simulation is applied to testify the effectiveness and feasibility of the method above.Finally, in order to realize the trajectory tracking of the space manipulator with unknown dynamic model, RBF neural network is used to approach the non-liner parameters of the dynamic manipulator and PD controller is used to compensate the error. Then a fuzzy PID controller which makes the use of the fuzzy reasoning ability to adjust the parameter of PID controller is designed to meet the requirements of the trajectory tracking.