Szczegóły publikacji
Opis bibliograficzny
Safe lane keeping with MPPI and control barrier functions / Maciej RÓŻEWICZ, Krzysztof KOGUT // IFAC-PapersOnLine [Dokument elektroniczny]. — Czasopismo elektroniczne ; ISSN 2405-8963 . — 2025 — vol. 59 iss. 36, s. 1-6. — Wymagania systemowe: Adobe Reader. — Bibliogr. s. 6, Abstr. — 19th IFAC workshop on Control Applications of Optimisation CAO 2025 : Kraków, Poland, December 10-13, 2025 : proceedings
Autorzy (2)
Słowa kluczowe
Dane bibliometryczne
| ID BaDAP | 166935 |
|---|---|
| Data dodania do BaDAP | 2026-04-08 |
| Tekst źródłowy | URL |
| DOI | 10.1016/j.ifacol.2026.03.001 |
| Rok publikacji | 2025 |
| Typ publikacji | referat w czasopiśmie |
| Otwarty dostęp |  |
| Creative Commons | |
| Czasopismo/seria | IFAC-PapersOnLine |
Abstract
Ensuring both safety and performance in autonomous driving remains a major challenge. Model Predictive Path Integral (MPPI) control is a sampling-based stochastic optimal control method well-suited for nonlinear vehicle dynamics, but it does not inherently guarantee safety. Control Barrier Functions (CBFs) provide formal safety guarantees by defining forward-invariant safe sets. In this paper, we propose a framework for lane keeping that integrates MPPI with CBF constraints applied to polynomial representations of lane boundaries. We derive higher-order barrier formulations suitable for the bicycle model and demonstrate, through simulations, that the proposed method improves safety while maintaining competitive performance compared to baseline Model Predictive Control (MPC) and unconstrained MPPI approaches.
