一個以大規模人類數據為基礎的自主賽車模擬基準測試

A Simulation Benchmark for Autonomous Racing with Large-Scale Human Data

July 23, 2024

作者: Adrian Remonda, Nicklas Hansen, Ayoub Raji, Nicola Musiu, Marko Bertogna, Eduardo Veas, Xiaolong Wang

cs.AI

摘要

儘管有國際獎金競賽、規模化車輛和模擬環境可用，但自主賽車和控制運作接近極限的跑車的研究受到車輛取得和管理成本高昂，以及開源模擬器物理精度有限的限制。本文提出了一個基於模擬器Assetto Corsa的賽車模擬平台，用於測試、驗證和基準自主駕駛算法，包括強化學習（RL）和經典模型預測控制（MPC），在逼真且具挑戰性的情境中。我們的貢獻包括開發這個模擬平台、幾種針對賽車環境量身定制的最先進算法，以及從人類駕駛員收集的全面數據集。此外，我們在離線RL設置中評估算法。所有必要的代碼（包括環境和基準）、工作示例、數據集和視頻均已公開發布，可在以下網址找到：https://assetto-corsa-gym.github.io。

English

Despite the availability of international prize-money competitions, scaled vehicles, and simulation environments, research on autonomous racing and the control of sports cars operating close to the limit of handling has been limited by the high costs of vehicle acquisition and management, as well as the limited physics accuracy of open-source simulators. In this paper, we propose a racing simulation platform based on the simulator Assetto Corsa to test, validate, and benchmark autonomous driving algorithms, including reinforcement learning (RL) and classical Model Predictive Control (MPC), in realistic and challenging scenarios. Our contributions include the development of this simulation platform, several state-of-the-art algorithms tailored to the racing environment, and a comprehensive dataset collected from human drivers. Additionally, we evaluate algorithms in the offline RL setting. All the necessary code (including environment and benchmarks), working examples, datasets, and videos are publicly released and can be found at: https://assetto-corsa-gym.github.io.