Research On Key Technologies Of A Rotorcraft-based Rendezvous And Docking Simulation Platform

With the continuous development of the space industry,increasing space missions have put forward higher requirements on space rendezvous and docking(Rv D)technology.In order to ensure a safe and reliable Rv D operation in space,Rv D operation must be carefully simu-lated and verified before the real space mission can be launched.Therefore,it’s significant to study the on-ground Rv D simulation platform.The existing simulators are mainly based on multi-degree-of-freedom manipulators,Stewart mechanisms,and air bearing vehicles,which are expensive and difficult for maintenance.Considering the shortcomings of the existing sim-ulators and the fact that there has been a rapid development of multirotor UAVs and multirotor UAVs have the characteristics of relatively low cost and multiple degrees of freedom,this thesis proposes a rotorcraft-based Rv D simulation platform as a new solution to support the research and verification of related technologies in the approaching and alignment phase of Rv D mis-sions.In this thesis,a detailed kinematic analysis of the proposed simulation platform is carried out.We get the conclusion that the inverse kinematics of the chasing spacecraft simulator has an analytical solution,and the speed Jacobian matrix is full rank.So that the chasing spacecraft simulator can satisfy requirements of any pose and space velocity.Therefore,It’s convenient for the chasing spacecraft simulator to simulate the position and attitude changes of the real spacecraft in Rv D missions.This thesis studies the vision-based measurement of the cooperative target attached on the target spacecraft.The April Tag,a kind of two-dimensional 2D barcode tag,is selected as the cooperative target of the simulation platform.Based on the April Tag algorithm,we combine kalman filtering algorithm with the concept of ROI(region of interest)to speed up the recog-nition.This thesis also studies the pose planning technology applied to chasing spacecraft.Specifically,this thesis studies the modeling method of pose planning problem with visual con-straints.A novel modeling method is proposed,that is to model the pose planning problem with visual constraints as a convex optimization problem with waypoint range constraints.And the experiment shows that this method greatly improves the speed of finding a solution without damaging the quality of the resulting trajectory.This thesis studies the trajectory tracking control method applied to chasing spacecraft sim-ulator.Specifically,the trajectory tracking control law in joint space and the trajectory tracking control law in task space are discussed separately,and experiments are carried out in Gazebo simulation environment.The task-space trajectory tracking control law utilizes real-time pose estimation of the target,which realizes the compensation of the originally planned trajectory.When the cooperative target is moving or the cooperative target pose estimation is not accurate at the initial moment,there is no need to re-plan the trajectory,which saves the time of trajectory re-planning.Finally,this thesis uses the proposed simulation platform to conduct a comprehensive ex-periment to simulate the Rv D approaching and alignment process.This experiment verifies the feasibility of the proposed simulation platform.