Development Of On-Orbit Inspection Mission Planning System For Space Manipulator

The Chinese space station is currently the most promising space station project on Earth’s orbit.During the construction and operation of the station,it will face a variety of complex and dangerous on-orbit challenges.As a typical on-orbit service mission,extravehicular inspection is an important means to carry out health inspection of the external structure of the space station,identifying potential threats and ensuring the long-term safe and stable operation of the space station.Space manipulator plays a significant role in the extravehicular inspection mission,but it currently lacks autonomous planning abilities.Thus,the efficiency of completing the inspection mission largely depends on the experience and knowledge of ground personnel.Therefore,research on on-orbit inspection mission planning and aid decision-making method for space manipulator has important theoretical significance and engineering application value,which is helpful to efficiently complete on-orbit inspection mission and ensure the smooth implementation of the Chinese space station construction.The main work of this paper is as follows:Firstly,designing a system for space manipulator on-orbit inspection mission planning.Through system requirements analysis,specific functional and performance requirements are identified.Following the principle of "high cohesion and low coupling",the overall system architecture is designed,and the user’s workflow is defined.According to functional requirements,the system is divided into on-orbit inspection mission planning subsystem and on-orbit inspection aid decision-making subsystem,and the design of subsystem is completed respectively.The goal is to achieve the mission planning and motion planning processes that include the full process of space manipulator on-orbit inspection and to provide visualization and simulation capabilities for verification.Secondly,in order to optimize the efficiency of space manipulator onorbit inspection mission,a mission planning algorithm for space manipulator on-orbit inspection mission is studied.According to the characteristics of space manipulator on-orbit inspection mission,the research defines the mission constraints and evaluation criteria,and develops the on-orbit inspection mission planning mathematical model.The inverse kinematics of the manipulator based on arm angle and the collision detection technology based on FCL are used to solve the feasible configurations corresponding to the inspection points,and a feasible configuration set selection strategy is proposed.To address the safety issues during the space manipulator inspection process and to determine the cost of executing inspections between different points,the study presents a collision-free path planning algorithm based on PRM.Aiming at the multi-objective on-orbit inspection mission planning model,multiple ant colony optimization algorithm is used to solve the optimal space manipulator inspection action sequence,thereby achieving autonomous planning of space manipulator inspection mission.Thirdly,to address the issue of long-time consumption for dense inspection tasks and repeated work when determining the position of inspection points by ground personnel,the study proposes an aid decisionmaking method based on knowledge graph to assist decision-making for space manipulator inspection missions.The workflow of the aid decisionmaking method is designed,and the knowledge system is divided into the knowledge representation layer and the knowledge operation layer.To determine the inspection target that the camera at the end of the manipulator can see in the position of the inspection point,a method of inspection object determination based on point cloud model is proposed.The link prediction method is used to deduce the connectivity probability of the inspection targets from prior inspection knowledge.Then,considering the mission evaluation criteria,the deep search algorithm is used to retrieve the optimal space manipulator inspection action sequence.Simulation experiments show that the aid decision-making method reduces the calculation time by 92.4%compared to the mission planning method when completing the same inspection task,thus improving the efficiency and knowledge management capability of space manipulator on-orbit inspection mission planning.Finally,in order to verify the application value of the system in the space manipulator inspection mission,the development and experiment of the space manipulator on-orbit inspection mission planning system are conducted.According to the business requirements of the space manipulator on-orbit inspection mission,the experimental scenarios are designed for both mission planning algorithm and aid decision-making method.The experiment is carried out according to the experimental scheme,the results are analyzed,the experimental data is processed,and the effectiveness and efficiency of the proposed methods are validated.