Fast Global Collision Detection Method Based on Feature-Point-Set for Robotic Machining of Large Complex Components

This paper presents a fast global collision detection method for robotic machining of large complex components, aiming to quickly determine whether there is a collision between the robot and the surrounding environment during the whole machining process. Geometric analysis shows that there are always some trajectory points on the motion path of the manipulator that are more likely to collide than the surrounding points during machining. These trajectory points with the highest collision probability within a certain range are defined as the feature points of global collision detection, and are used to replace all trajectory points to perform global collision detection, thus greatly improving the efficiency of related operations while ensuring accuracy. Compare to the traditional discrete collision detection method with computational complexity O( $\text{n}^{2}$ ), the computational complexity of the proposed method is only O(n). Numerical analysis and application experiments verify the effectiveness of the proposed method. Note to Practitioners—Motion planning in robotic machining of large complex components usually needs to perform a lot of global collision detection. Existing methods generally have the problems of large calculation and low efficiency, which seriously affects the efficiency of motion planning. This is mainly because a single global collision detection usually includes no less than $n$ times of static collision detection, where $n$ is the number of trajectory points. In order to solve this problem, we present a new global collision detection method based on feature-point-set. It does not need to traverse all trajectory points for static collision detection, but only needs to detect a few feature points, that is, the trajectory points most likely to collide within a certain range. On the premise of ensuring the collision detection accuracy, the proposed method greatly reduces the execution times of static collision detection, and significantly improves the computational efficiency of global collision detection. Numerical analysis and experiments show that this method effectively improves the efficiency of motion planning in robotic machining of large complex components.