- 本論文は、autonomous car racing のための Model Predictive Control (MPC) ベースの motion planning and control pipeline を提案する。
- 異なる driving contexts(overtaking, nominal driving, countersteering)に適応可能。
- Cost Blending state machine が driving context を識別し、Model Predictive Planning (MPP) と MPC モジュールに適用する predefined weights を選択する。
- 両最適化は同一の problem formulation と model を共有し、horizon length, rate, tuning, open-loop vs closed-loop approach のみが異なる。
- 完全自律型 open-wheel racecar Superformula EAV-25 で検証され、lap time が bes…
2. 先行研究と比べてどこがすごい?
- 先行研究と比べて、driving contexts に応じて Cost Blending state machine で重みを動的に切り替える点が新しい。
- MPP と MPC が同一の problem formulation と model を共有しつつ、horizon length や rate などの違いで役割分担する点が特徴的。
- 具体的な先行研究との比較は要旨からは不明。
3. 技術・手法の肝は?
- Model Predictive Control (MPC) ベースの motion planning and control pipeline。
- Cost Blending state machine が driving context(overtaking, nominal driving, countersteering)を識別し、predefined weights を選択。
- Model Predictive Planning (MPP) と MPC は同一の problem formulation と model を共有。
- 両者は horizon length, rate, tuning, open-loop vs closed-loop approach で異なる。
- これにより interaction を最大化。
4. どうやって有効だと検証した?
- 完全自律型 open-wheel racecar Superformula EAV-25 で検証。
- lap time が best human driver reference の 2% 以内を達成。
- driving at the limit of handling、smooth overtaking maneuvers、high oversteering conditions への迅速な反応と車両安定性回復を実証。
This paper presents a Model Predictive Control-based motion planning and control pipeline for autonomous car racing capable of adapting to different driving contexts, such as overtaking, nominal driving, and countersteering. A Cost Blending state machine manages the identification of different driving contexts and the selection of their predefined weights to be applied to the Model Predictive Planning (MPP) and Control (MPC) modules. The two optimization-based solutions share the same problem formulation and model, differing only in horizon length, rate, tuning, and in their open-loop versus closed-loop approach to maximize the effectiveness of their interaction. The work is validated on the fully autonomous open-wheel racecar Superformula EAV-25, with a lap time achieved that is within 2% of the best human driver reference. The results demonstrate the capability of the solution in driving at the limit of handling, smoothly executing overtaking maneuvers, and quickly reacting to high oversteering conditions to recover the vehicle stability.