Optimal Multi-Criteria Waypoint Selection for Autonomous Vehicle Navigation in Structured Environment

Vilca, José; Adouane, Lounis; Mezouar, Youcef · 2015 · OpenAlex-citations

DOI: 10.1007/s10846-015-0223-1

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This paper addresses the challenge of autonomous navigation for unmanned ground vehicles (UGVs) in structured environments, such as warehouses or urban settings. The authors identify that traditional trajectory planning methods, which rely on generating continuous reference paths, are often computationally expensive and lack flexibility in dynamic or cluttered environments. To overcome these limitations, the study proposes a navigation strategy based on sequential waypoint assignment rather than full trajectory tracking. This approach simplifies the navigation problem to reaching discrete key configurations, allowing for smoother, safer, and more flexible vehicle movement without the need for complex replanning. The methodology centers on an Optimal Multi-Criteria Waypoint Selection (OMWS) framework. The authors develop two distinct algorithms to generate these waypoints: one based on a grid map (OMWS-GM) and another based on an expanding tree (OMWS-ET). Both methods formulate waypoint selection as a multi-criteria optimization problem solved via dynamic programming. The optimization criteria include safety (distance to obstacles and road boundaries), feasibility (respecting vehicle kinematic constraints like non-holonomy, maximum velocity, and steering angles), and robustness against localization uncertainties. The OMWS-ET method specifically integrates the vehicle’s kinematic model and control law directly into the tree expansion, whereas the grid-based method treats the vehicle as a single cell, potentially ignoring specific kinematic constraints. The navigation system utilizes a Lyapunov-based target reaching controller to guide the vehicle between selected waypoints. The study evaluates the proposed methods through simulations and experiments using an urban electric vehicle. The results demonstrate that the expanding tree-based approach (OMWS-ET) is superior to the grid-based method (OMWS-GM) in terms of flexibility and efficiency. The OMWS-ET method successfully generates waypoints that account for the vehicle's specific kinematic constraints and control laws, resulting in smoother trajectories and better handling of localization uncertainties. In contrast, the grid-based method was found to be less flexible and less efficient, as it does not adequately consider the vehicle's orientation or kinematic limitations during the planning phase. The experiments confirm that the proposed waypoint-based navigation allows the UGV to reach its final configuration safely and smoothly, avoiding collisions with obstacles and route boundaries. The significance of this work lies in providing a robust and computationally efficient alternative to traditional trajectory planning for autonomous vehicles. By focusing on optimal waypoint selection rather than continuous path generation, the proposed method reduces computational load and enhances adaptability in structured environments. The integration of kinematic constraints and uncertainty models into the planning process ensures that the generated waypoints are not only geometrically safe but also dynamically feasible. This approach offers a practical solution for real-world applications where rapid, reliable, and flexible navigation is required, contributing to the advancement of autonomous vehicle technologies in complex operational settings.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• steering pattern