Environment Modeling And Experimental Research On Space Robot Based On Multi-sense Information

Space exploration activities in which space robots replace astronauts have always been a hotspot of research by experts and scholars at home and abroad.The semi-autonomous space robot remote operation human-computer interaction technology with human participation is still the main method in the field of space exploration.However,There is a large delay between the signal transmission between space robots and operators,which greatly affects the teleoperation task.According to the operational requirements of space robots and the research status of teleoperation robots,a ground teleoperation verification system for space robots based on multi-perception information is designed and built.Firstly,the kinematics analysis of the master-slave robot realizes the forward and inverse solutions of the master-slave robot and establishes a kinematics model based on the Schunk manipulator and seven degree freedom force feedback hand controller.It provides a theoretical basis and experimental platform for the performance development and experimental verification of each module of the space teleoperation robot system.Then,the geometric modeling of the virtual environment is carried out for the predefined environment and the unstructured environment of the space robot,and the corresponding geometric correction method is proposed for experimental research.For the geometric modeling of the virtual manipulator under a predefined environment,the geometric modeling of the predefined environment is realized by importing OBJ files in Open GL language,and a DAASD algorithm for collision detection of virtual robots and virtual objects is proposed.The prediction model correction algorithm completes the correction.For the virtual environment modeling of unstructured environment,a method based on RGB-D real-time point cloud modeling is proposed,and the geometric characteristics of the target object are extracted using the RANSAC algorithm,which is carried out through the point cloud information collected by the Barrett smart hand touch sensor Calibration completes the geometric modeling of the virtual environment.Focusing on the dynamic modeling of the virtual robot arm of the space robot,a dynamic model CSCM model suitable for Schunk robot arms and different contact environments is proposed,and a suitable parameter identification method is proposed for the model parameter estimation for the model.The nonlinear terms of the model are linearized,and for the linear terms of the dynamic model,a least squares identification algorithm based on adaptive forgetting operator is proposed,and the dynamic parameters identified at each step are fed back to the detection and correction module primary virtual environment.In addition,for the correction of dynamic parameters,a correction algorithm based on the minimum perceivable difference is proposed to ensure the virtual force feedback modeling reality by ensuring that the virtual force deviation is within the minimum perceivable difference.For the study of the stability of virtual environment systems,the relationship between environment and damping is fully considered,and stability experiments are designed and the system stability is verified.Aiming at the negative impact of human random input,which is inevitable in the remote operation system of the loop,on the operation of the system,a virtual environment based on RGB-D point cloud data is proposed,and an artificial force field is constructed through the point cloud data,By constructing repulsive force and attractive potential field models and generating guided virtual fixtures in real time,constructing forbidden virtual fixtures for the prohibited operation area of the robotic arm to restrict the operator's behavioral norms to ensure the safety of the system,and modeling the virtual constraint force The earth improves the efficiency of the operator moving the robot arm to perform the task,and the effectiveness is verified through experiments.Finally,design and launch a comprehensive system experiment,linking the geometric modeling and dynamic modeling of the virtual environment,and verifying the force feedback effect of the system when the operator performs the teleoperation task in the virtual environment under different delay conditions Integrity.