Research On Motion Planning Method Of Space Super-redundant Manipulator

In order to improve the dexterity of space manipulator motion,research on space super-redundant manipulator becomes an important branch of the space robot field.The computational complexity of manipulator motion planning algorithm is in polynomial relation with the geometry complexity of the surrounding environment,and in exponential relation with the degree of freedom(DOF)of robot manipulator.Abovementioned relations cause the difficulty of motion planning algorithm of space superredundant manipulator extremely The research in this dissertation is focused on on-orbit assembly task of space large antenna,and researches two motion planning methods of space super-redundant manipulator.Firstly,the kinematic model of space super-redundant manipulator is established using classical D-H modeling method.Then,two motion planning methods are studied and compared according to different application scenarios of space on-orbit assembly.In order to improve the ability of space manipulator to travel through narrow space,the tip-following planning method based on A * algorithm is studied.The A * algorithm is used to plan an optimal path curve of the manipulator end from starting point to end point,then the joint-nodes of manipulator are matched with the path points to realize tipfollowing motion.The tip-following motion makes the whole manipulator move forward along the end trajectory,which makes the links of manipulator approaching the path curve to the maximum extent.Thus this method is suitable for motion in narrow space.On the contrary,to ensure a large range of movement for super-redundant manipulator,an effective RRT algorithm in high-dimensional search space is studied.In order to solve the problem of low efficiency of motion planning algorithm in high dimensional search space,two improved strategies are introduced,namely target bias RRT and incremental subspace dimension RRT.The target bias guides the expansion of RRT to the target region.The incremental subspace dimension essentially reduces the search dimension of the algorithm and finds planning path in the lower dimension subspace faster.Finally,simulation and comparison of the two motion planning algorithms are carried out,and their application feasibility is shown: two motion planning methods can effectively complete the motion planning task of space super redundant manipulator.The comparative analysis of the two methods shows that,the tip-following method based on A* algorithm causes some motion limitations in macroscopically,and the running efficiency of the algorithm is high.On the other hand,the RRT algorithm based on incremental subspace dimension has obvious advantages in the high dimensional search space of super redundant arm.And the motion of manipulator is less limited in macroscopically.The two planning methods have different scope of application,and both of them are feasible in super-redundant space manipulator.