Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Robust Lateral Control of a Convoy of Autonomous & Connected Vehicles with Limited Preview

Mengke Liu, Neelkamal Somisetty, Sivakumar Rathinam, Swaroop Darbha

TL;DR

This paper advances convoy-level lateral control for autonomous and connected vehicles during Emergency Lane Change maneuvers by leveraging GPS/IMU connectivity to synthesize a composite ELC trajectory from lead and preceding-vehicle previews. The controller combines feedforward curvature-based steering with error-driven feedback, and uses a D-decomposition approach to identify a stabilizing fixed-structure gain set that remains robust under slow variations in longitudinal speed and mass distribution. The authors prove stability for frozen-parameter gains and demonstrate, through numerical simulations and Lincoln MKZ experiments, that incorporating lead-vehicle preview data eliminates lateral string instability that arises when only predecessor data is used. The results show the method achieves sub-decimeter lateral tracking errors (under typical speeds) and robust string stability across diverse mass and speed scenarios, highlighting practical viability for infrastructure-free convoy control in real-world settings.

Abstract

This paper addresses the lateral control of Autonomous & Connected Vehicles (ACVs) convoys during Emergency Lane Change (ELC) maneuvers. These maneuvers are initiated in response to emergency cues from either the front or rear of the convoy, responding to the need to avoid obstacles or facilitate the passage of other vehicles. The primary objective of this study is to develop a lateral control scheme for ACVs based on the available information. The foundational assumption in this study is the existence of reliable connectivity among ACVs, wherein each subsequent ACV possesses information concerning the GPS position traces of both the lead and immediately preceding vehicles within the convoy. This connectivity facilitates the construction of a composite ELC trajectory that synthesizes this information, serving as a "discretized" preview of the trajectory to be tracked. The procedural steps include constructing this composite trajectory, determining cross-track error, heading, and yaw rate errors relative to it, and subsequently formulating a lateral control strategy. Furthermore, the paper presents findings on the lateral string stability of ACV convoys across various scenarios, encompassing changes in longitudinal velocity and scenarios where lead vehicle information is unavailable. Numerical and experimental results validate the efficacy of the proposed lateral control scheme for ACV convoys.

Robust Lateral Control of a Convoy of Autonomous & Connected Vehicles with Limited Preview

TL;DR

This paper advances convoy-level lateral control for autonomous and connected vehicles during Emergency Lane Change maneuvers by leveraging GPS/IMU connectivity to synthesize a composite ELC trajectory from lead and preceding-vehicle previews. The controller combines feedforward curvature-based steering with error-driven feedback, and uses a D-decomposition approach to identify a stabilizing fixed-structure gain set that remains robust under slow variations in longitudinal speed and mass distribution. The authors prove stability for frozen-parameter gains and demonstrate, through numerical simulations and Lincoln MKZ experiments, that incorporating lead-vehicle preview data eliminates lateral string instability that arises when only predecessor data is used. The results show the method achieves sub-decimeter lateral tracking errors (under typical speeds) and robust string stability across diverse mass and speed scenarios, highlighting practical viability for infrastructure-free convoy control in real-world settings.

Abstract

This paper addresses the lateral control of Autonomous & Connected Vehicles (ACVs) convoys during Emergency Lane Change (ELC) maneuvers. These maneuvers are initiated in response to emergency cues from either the front or rear of the convoy, responding to the need to avoid obstacles or facilitate the passage of other vehicles. The primary objective of this study is to develop a lateral control scheme for ACVs based on the available information. The foundational assumption in this study is the existence of reliable connectivity among ACVs, wherein each subsequent ACV possesses information concerning the GPS position traces of both the lead and immediately preceding vehicles within the convoy. This connectivity facilitates the construction of a composite ELC trajectory that synthesizes this information, serving as a "discretized" preview of the trajectory to be tracked. The procedural steps include constructing this composite trajectory, determining cross-track error, heading, and yaw rate errors relative to it, and subsequently formulating a lateral control strategy. Furthermore, the paper presents findings on the lateral string stability of ACV convoys across various scenarios, encompassing changes in longitudinal velocity and scenarios where lead vehicle information is unavailable. Numerical and experimental results validate the efficacy of the proposed lateral control scheme for ACV convoys.

Paper Structure

This paper contains 26 sections, 2 theorems, 48 equations, 14 figures, 1 table.

Table of Contents

  1. Introduction
  2. Relation to Literature & Novelty of the Proposed Work
  3. Organization
  4. Mathematical Model for Lateral Dynamics of an ACV
  5. Lateral Controller Design
  6. Target Trajectory Construction
  7. Fitting a circular arc based on preview data
  8. Computation of Feedback Errors
  9. Lateral Controller Synthesis
  10. Feedforward Controller Design
  11. Feedback Controller Design
  12. Construction of the Set of Stabilizing Structured Feedback Controllers
  13. Robustness in Stability
  14. Closed-Loop Stability with Varying Longitudinal Velocity
  15. Lateral Instability with Predecessor's Information Only
  16. ...and 11 more sections

Key Result

Lemma 1

Suppose the linear system: satisfies the following conditions: Then, the solution ${\bf z} = 0$ is exponentially stable.

Figures (14)

  • Figure 1: Structure of the proposed lateral controller.
  • Figure 2: Illustration of the bicycle model.
  • Figure 3: Illustration of the sampled trajectories of ACVs in the convoy and the information available to the ego ACV.
  • Figure 4: Heading and lateral errors for a straight line segment.
  • Figure 5: Heading and lateral errors for a circular arc.
  • ...and 9 more figures

Theorems & Definitions (5)

  • Lemma 1
  • proof
  • Proposition 1
  • proof
  • Definition 1: Lateral String Stability