Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Guess the Drift with LOP-UKF: LiDAR Odometry and Pacejka Model for Real-Time Racecar Sideslip Estimation

Alessandro Toschi, Nicola Musiu, Francesco Gatti, Ayoub Raji, Francesco Amerotti, Micaela Verucchi, Marko Bertogna

TL;DR

This work addresses the challenge of estimating vehicle lateral velocity and sideslip in autonomous racing by fusing LiDAR odometry with a Pacejka tire model inside an Unscented Kalman Filter (LOP-UKF). The method uses a non-linear prediction with a Pacejka-based dynamics model and LiDAR/IMU measurements in a two-stage UKF update, achieving real-time performance at 125 Hz and leveraging a SLAM-augmented localization framework. Experimental results on the Dallara AV-21 across oval and road-course tracks demonstrate robustness to varying grip and track conditions, with the method offering improved lateral-velocity estimates over model-only approaches, particularly in non-linear tire regimes. The work highlights the practical impact of combining physics-based tire modeling with LiDAR-based state estimation for enhanced safety and performance in autonomous racing, and suggests avenues for real-time adaptive tire modeling and multi-sensor fusion enhancements.

Abstract

The sideslip angle, crucial for vehicle safety and stability, is determined using both longitudinal and lateral velocities. However, measuring the lateral component often necessitates costly sensors, leading to its common estimation, a topic thoroughly explored in existing literature. This paper introduces LOP-UKF, a novel method for estimating vehicle lateral velocity by integrating Lidar Odometry with the Pacejka tire model predictions, resulting in a robust estimation via an Unscendent Kalman Filter (UKF). This combination represents a distinct alternative to more traditional methodologies, resulting in a reliable solution also in edge cases. We present experimental results obtained using the Dallara AV-21 across diverse circuits and track conditions, demonstrating the effectiveness of our method.

Guess the Drift with LOP-UKF: LiDAR Odometry and Pacejka Model for Real-Time Racecar Sideslip Estimation

TL;DR

This work addresses the challenge of estimating vehicle lateral velocity and sideslip in autonomous racing by fusing LiDAR odometry with a Pacejka tire model inside an Unscented Kalman Filter (LOP-UKF). The method uses a non-linear prediction with a Pacejka-based dynamics model and LiDAR/IMU measurements in a two-stage UKF update, achieving real-time performance at 125 Hz and leveraging a SLAM-augmented localization framework. Experimental results on the Dallara AV-21 across oval and road-course tracks demonstrate robustness to varying grip and track conditions, with the method offering improved lateral-velocity estimates over model-only approaches, particularly in non-linear tire regimes. The work highlights the practical impact of combining physics-based tire modeling with LiDAR-based state estimation for enhanced safety and performance in autonomous racing, and suggests avenues for real-time adaptive tire modeling and multi-sensor fusion enhancements.

Abstract

The sideslip angle, crucial for vehicle safety and stability, is determined using both longitudinal and lateral velocities. However, measuring the lateral component often necessitates costly sensors, leading to its common estimation, a topic thoroughly explored in existing literature. This paper introduces LOP-UKF, a novel method for estimating vehicle lateral velocity by integrating Lidar Odometry with the Pacejka tire model predictions, resulting in a robust estimation via an Unscendent Kalman Filter (UKF). This combination represents a distinct alternative to more traditional methodologies, resulting in a reliable solution also in edge cases. We present experimental results obtained using the Dallara AV-21 across diverse circuits and track conditions, demonstrating the effectiveness of our method.
Paper Structure (17 sections, 13 equations, 10 figures)

This paper contains 17 sections, 13 equations, 10 figures.

Table of Contents

  1. INTRODUCTION
  2. RELATED WORK
  3. APPLICATION
  4. UNSCENTED KALMAN FILTER
  5. Prediction Phase
  6. Sigma Points Generation
  7. Sigma Points Transformation
  8. Predicted State and Covariance Calculation
  9. Update Phase
  10. Measurement Incorporation
  11. Implementation
  12. PACEJKA TIRE MODEL
  13. LIDAR SLAM
  14. EXPERIMENTAL RESULTS
  15. Oval Speedway
  16. ...and 2 more sections

Figures (10)

  • Figure 1: Lateral and heading tracking error of the Model Predictive Controller in the simulator. Here, Prima Variante turn of the Monza circuit is shown.
  • Figure 2: Diagram showing LOP-UKF interaction with other key parts of the autonomous driving system
  • Figure 3: The Pacejka Tire Model (orange line) optimized based on the experimental data (blue scatter) from Texas Motor Speedway, during the event of the Indy Autonomous Challenge.
  • Figure 4: Geo-referenced LiDAR map of Monza F1 track. The zoomed portion shows Variante Ascari more in details
  • Figure 5: The velocity field of a hot lap in the Oval track of Texas Motor Speedway. The optical sensor serves as a reference.
  • ...and 5 more figures