Learning to Tune Pure Pursuit in Autonomous Racing: Joint Lookahead and Steering-Gain Control with PPO

TL;DR

This paper introduces a reinforcement learning approach using PPO to jointly optimize lookahead distance and steering gain in Pure Pursuit for autonomous racing, improving performance without per-track tuning.

Contribution

It presents a novel RL-based method that adaptively tunes key Pure Pursuit parameters online, outperforming traditional fixed or scheduled tuning methods in racing scenarios.

Findings

RL-PP outperforms fixed-lookahead PP in lap times and accuracy

The approach generalizes across different tracks and speeds

It reduces the need for manual retuning in autonomous racing

Abstract

Pure Pursuit (PP) is widely used in autonomous racing for real-time path tracking due to its efficiency and geometric clarity, yet performance is highly sensitive to how key parameters-lookahead distance and steering gain-are chosen. Standard velocity-based schedules adjust these only approximately and often fail to transfer across tracks and speed profiles. We propose a reinforcement-learning (RL) approach that jointly chooses the lookahead Ld and a steering gain g online using Proximal Policy Optimization (PPO). The policy observes compact state features (speed and curvature taps) and outputs (Ld, g) at each control step. Trained in F1TENTH Gym and deployed in a ROS 2 stack, the policy drives PP directly (with light smoothing) and requires no per-map retuning. Across simulation and real-car tests, the proposed RL-PP controller that jointly selects (Ld, g) consistently outperforms fixed-lookahead PP, velocity-scheduled adaptive PP, and an RL lookahead-only variant, and it also exceeds a kinematic MPC raceline tracker under our evaluated settings in lap time, path-tracking accuracy, and steering smoothness, demonstrating that policy-guided parameter tuning can reliably improve classical geometry-based control.