Skip to content

Constrained Global Path Optimization for Articulated Steering Vehicles

Authors: Ji-wung Choi, Kalevi Huhtala

Published: 2016 (Journal Paper)

Source: IEEE Transactions on Vehicular Technology

Algorithm: Quintic Bezier Splines

DOI: 10.1109/TVT.2015.2487964

Summary

Plans paths for articulated steering vehicles by pre-computing quintic Bezier motion primitives offline, then searching over them with A*, with online gradient-based smoothing. Combines the tractability of A* over discrete motion primitives with the geometric smoothness of Bezier curves for non-holonomic articulated vehicles.

Abstract

This paper proposes a new efficient path-planning algorithm for articulated steering vehicles operating in semi-structured environments, in which obstacles are detected online by the vehicle's sensors. The first step of the algorithm is offline and computes a finite set of feasible motions that connect discrete robot states to construct a search space. The motion primitives are parameterized using Bezier curves and optimized as a nonlinear programming problem (NLP) equivalent to the constrained path planning problem. Applying the A* search algorithm to the search space produces the shortest paths as a sequence of these primitives. Online path smoothing, which uses a gradient-based method, is applied to solve another NLP.

Primary

Paper ieeexplore.ieee.org

Standard

DOI 10.1109/TVT.2015.2487964

Tags

  • Path planning

  • Articulated vehicles

  • Bezier curves

  • Motion primitives

  • A* search

  • Gradient optimization

  • Nonlinear programming