Hybrid hierarchical bounding volume collision detection in teleoperation manipulator system

It is well known that the reliability and precision of long-distance teleoperation may deteriorate owing to the time delay. Fortunately, virtual reality technique has its unique characteristics in overcoming the problem. But it is easy to overlap in the virtual environment, especially, in a grasping process. In order to prevent the manipulator to penetration into an object, a collision detection method named hybrid hierarchical bounding volume (HBV) is introduced. Due to the requirements of real-time detection, an improved collision detection algorithm named optimized hierarchy bounding volume (OHBV) based on hybrid HBV is proposed. In this paper, the hybrid hierarchical bounding volume consists of Sphere and oriental bounding box (OBB). compared with the similar algorithms, We select the best of the two bounding boxes by minimizing the volume for each level except the top level. Further, we first introduce predictive control base on the distance between the object and the end tool of the manipulator and optimize the traversal order under different circumstance in OHBV. The teleoperation process can be divided into three steps, approaching,grasping and dragging. The collision detection method is used in the first two steps. The bounding Sphere is used in the algorithm to execute the process of approaching and the OBB tree is applied to the grasping process. We have verified the effectiveness of our approach by experiments with a real teleoperation manipulator system. The experimental results show OHBV can be used to improve the performance of the process of grasping objects by teleoperated manipulators .