A Segmented Geometry Method for Kinematics and Configuration Planning of Spatial Hyper-Redundant Manipulators.
Zonggao Mu,Han Yuan,Wenfu Xu,Tianliang Liu,Bin Liang
DOI: https://doi.org/10.1109/tsmc.2017.2784828
2020-01-01
Abstract:With many degrees of freedom (DOFs), a hyper-redundant manipulator has superior dexterity and flexible manipulation ability. However, its inverse kinematics and configuration planning are very challenging. With the increase in the number of DOFs, the corresponding computation load or training set will be much larger for traditional methods (such as the generalized inverse method and the artificial neural network method). In this paper, a segmented geometry method is proposed for a spatial hyper-redundant manipulator to solve the above problems. Similar to the human arm, the hyper-redundant manipulator is segmented into three sections from geometry, i.e., shoulder, elbow, and wrist. Then, its kinematics can be solved separately according to the segmentation, which reduces the complexity of the solution and simplifies the computation of the inverse kinematics. Furthermore, the configuration is parameterized by several parameters, i.e., the arm-angle, space arc parameters, and desired direction vector. The shoulder has proximal four DOFs, which is redundant for positioning the elbow and avoiding the joint limit. The arm-angle parameter is defined to solve the redundancy. The wrist consists of the distal two DOFs, and its joints are determined to match the desired direction vector of the end-effector. All the other joints (except for the joints belonging to shoulder and wrist) compose the elbow. These joint angles are solved by using space arc-based method. The configuration planning for avoiding joint limit, obstacles, and inspecting narrow pipeline are detailed for practical applications. Finally, circular trajectory tracking and pipeline inspection are, respectively, simulated and experimented on a 20-DOFs hyper-redundant manipulator. The results show that the proposed method can give solutions of the three-dimensional-pose-determining problem and the configuration-planning problem. The computation of the inverse kinematics is simplified for real-time control. It can also be applied to other spatial hyper-redundant manipulators with similar serial configurations.