Path Tracking with Dynamic Control Point Blending for Autonomous Vehicles: An Experimental Study

TL;DR

The paper tackles precise path tracking for autonomous vehicles under diverse driving contexts by blending lateral commands applied at different wheelbase locations. The core method is a barycentric blend $\delta(t) = \alpha \delta_{front}(t) + (1-\alpha) \delta_{rear}(t)$, with $\alpha \in [0,1]$, where $\delta_{front}$ is produced by a front-axle Stanley controller and $\delta_{rear}$ by a rear-axle curvature-based controller. A curvature-aware longitudinal policy uses virtual borders and ray-tracing to convert upcoming curvature into a virtual obstacle distance, regulating speed via $a = \min(a_r, a_{idm})$, where $a_r$ is obstacle-based acceleration and $a_{idm}$ is the IDM term. Experimental results in simulation and on a real vehicle demonstrate improved trajectory tracking accuracy and steering smoothness versus fixed-control-point baselines, with notable gains in backward maneuvers.

Abstract

This paper presents an experimental study of a path-tracking framework for autonomous vehicles in which the lateral control command is applied to a dynamic control point along the wheelbase. Instead of enforcing a fixed reference at either the front or rear axle, the proposed method continuously interpolates between both, enabling smooth adaptation across driving contexts, including low-speed maneuvers and reverse motion. The lateral steering command is obtained by barycentric blending of two complementary controllers: a front-axle Stanley formulation and a rear-axle curvature-based geometric controller, yielding continuous transitions in steering behavior and improved tracking stability. In addition, we introduce a curvature-aware longitudinal control strategy based on virtual track borders and ray-tracing, which converts upcoming geometric constraints into a virtual obstacle distance and regulates speed accordingly. The complete approach is implemented in a unified control stack and validated in simulation and on a real autonomous vehicle equipped with GPS-RTK, radar, odometry, and IMU. The results in closed-loop tracking and backward maneuvers show improved trajectory accuracy, smoother steering profiles, and increased adaptability compared to fixed control-point baselines.

Figures (27)

• Figure 1: Three different control point for lateral control, and their impact on the behavior of a vehicle during track following
• Figure 2: Example: for backward parking, a trajectory is defined according to obstacles, the selection of the control point makes it possible to park
• Figure 3: Kinematic bicycle model
• Figure 4: Lateral error of the control point for various command (distance in meters between the control point and its projection on the trajectory)
• Figure 5: Heading error in degrees for the control point (angle between the heading of the vehicle and the heading of the trajectory at the projection of the control point)