Abstract: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.

Learning-Based Proxy Collision Detection for Robot Motion Planning Applications

Fastron: An Online Learning-Based Model and Active Learning Strategy for Proxy Collision Detection

Forward Kinematics Kernel for Improved Proxy Collision Checking

Efficient Collision Detection Framework for Enhancing Collision-Free Robot Motion

Predicting Sample Collision with Neural Networks

Neural Implicit Swept Volume Models for Fast Collision Detection

Object Rearrangement Using Learned Implicit Collision Functions

A Collision Predict Based Local Path Planning of Mobile Robots

Active learning of the collision distance function for high-DOF multi-arm robot systems

Configuration Space Decomposition for Learning-based Collision Checking in High-DOF Robots

Efficient Collision Detection Oriented Motion Primitives for Path Planning

Stochastic Modeling of Distance to Collision for Robot Manipulators

Perceptual Local Collision-Free Motion Generation for Manipulators

A Collision-Free Motion Planning Method by Integrating Complexity-Reduction SLAM and Learning-Based Artificial Force Design

A Fast Collision Detection Method Based on Point Clouds and Stretched Primitives for Manipulator Obstacle-Avoidance Motion Planning

Safe and Efficient Path Planning under Uncertainty via Deep Collision Probability Fields

Aggressive Collision-Inclusive Motion Planning

Preprocessing-based Kinodynamic Motion Planning Framework for Intercepting Projectiles using a Robot Manipulator

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

On the Synthesis of Reactive Collision-Free Whole-Body Robot Motions: A Complementarity-based Approach

Collision-free Motion Generation Based on Stochastic Optimization and Composite Signed Distance Field Networks of Articulated Robot