GPU-Accelerated Optimization-Based Collision Avoidance

TL;DR

This paper introduces a GPU-accelerated optimization framework for collision avoidance that decomposes complex problems into parallelizable low-dimensional QPs, enabling faster computations for real-time applications.

Contribution

It presents a novel collision avoidance constraint based on convex duality and scale detection, allowing high-dimensional problems to be efficiently solved using GPU parallelization.

Findings

Significantly reduced computational time with GPU acceleration.

Effective collision avoidance demonstrated in high-fidelity simulations.

Decomposition into low-dimensional QPs enables real-time performance.

Abstract

This paper proposes a GPU-accelerated optimization framework for collision avoidance problems where the controlled objects and the obstacles can be modeled as the finite union of convex polyhedra. A novel collision avoidance constraint is proposed based on scale-based collision detection and the strong duality of convex optimization. Under this constraint, the high-dimensional non-convex optimization problems of collision avoidance can be decomposed into several low-dimensional quadratic programmings (QPs) following the paradigm of alternating direction method of multipliers (ADMM). Furthermore, these low-dimensional QPs can be solved parallel with GPUs, significantly reducing computational time. High-fidelity simulations are conducted to validate the proposed method's effectiveness and practicality.