Post-Impact Motion Planning and Tracking Control for Autonomous Vehicles

Wang, Cong; Wang, Zhenpo; Zhang, Lei; Yu, Huilong; Cao, Dongpu · 2022 · DOAJ

DOI: 10.1186/s10033-022-00745-w

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 critical safety challenge of post-impact vehicle stability, specifically aiming to prevent secondary collisions and stabilize autonomous vehicles following an initial crash. The motivation stems from statistics indicating that multi-impact crashes account for over 30% of fatal accidents, often because drivers lose control due to panic or severe vehicle dynamics like drifting and over-spinning. The authors propose a comprehensive control scheme that integrates motion planning with hierarchical tracking control to restore stability and avoid obstacles in post-impact scenarios. The methodology consists of two main components: motion planning and tracking control. For motion planning, the authors combine quintic polynomial curves with an artificial potential field (APF). Unlike traditional methods that rely on nonholonomic constraints, this approach uses separate polynomial curves for longitudinal, lateral, and yaw motions to fully describe vehicle dynamics without such constraints. The APF constructs an objective function using exponential potential functions to penalize proximity to obstacles and road boundaries, while a secondary objective minimizes the average vehicle sideslip angle to ensure stability. The planning problem is solved as a constrained optimization task using MATLAB’s `fmincon` toolbox, subject to linear terminal state constraints and nonlinear dynamics constraints, including road adhesion limits and tire force saturation. For tracking the planned trajectory, a hierarchical controller is employed. The upper layer utilizes a Time-Varying Linear Quadratic Regulator (TVLQR) to calculate desired generalized forces. This controller linearizes the nonlinear vehicle dynamics around the reference trajectory at each time step, allowing for efficient multi-objective optimization of error dynamics. The lower layer implements a nonlinear-optimization-based torque allocation algorithm to coordinate actuators and realize the desired forces, particularly when tire forces are saturated. The entire scheme was verified through comprehensive hardware-in-the-loop (HIL) tests under various driving scenarios. The significance of this work lies in its ability to simultaneously address trajectory safety and vehicle stability during the critical post-impact phase. By integrating polynomial-based motion planning with APF for obstacle avoidance and using TVLQR for robust tracking, the proposed method effectively mitigates the risks of secondary collisions. The study demonstrates that active safety systems can significantly enhance post-impact control, offering a viable solution for autonomous vehicles to maintain stability and navigate safely after high-intensity impacts, thereby potentially reducing fatalities and injuries associated with multi-impact crashes.

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.

• driver post crash behavior