The invention provides a safety trajectory planning method and system based on reinforcement learning and a reachable set, and belongs to the field of automatic driving. According to the method, the track rough solution is generated by utilizing reinforcement learning, and compared with a method based on search and sampling, the calculation efficiency is higher, complex and tedious rule design is not needed, and the ability of learning evolution is achieved; the reachable set is used for correcting the track rough solution, constraint conditions are provided for the track optimization problem, and the efficiency and safety of vehicle driving are guaranteed; a constraint iteration linear quadratic regulator is used for trajectory optimization, and compared with a quadratic optimization algorithm, the generated trajectory curvature is smaller, and the trajectory comfort level is higher.
本发明提供一种基于强化学习和可达集的安全轨迹规划方法与系统及系统,属于自动驾驶领域。本发明利用强化学习进行轨迹粗解生成,与基于搜索和采样的方法相比,计算效率更高,无需复杂繁琐的规则设计,具备学习进化的能力;利用可达集对轨迹粗解进行修正,并为轨迹优化问题提供约束条件,保障了车辆行驶的效率以及安全性;利用约束迭代线性二次型调节器进行轨迹优化,与二次优化算法相比,生成的轨迹曲率更小,轨迹舒适度更高。
Safety trajectory planning method and system based on reinforcement learning and reachable set
一种基于强化学习和可达集的安全轨迹规划方法与系统
22.11.2024
Patent
Elektronische Ressource
Chinesisch
IPC: | G06F ELECTRIC DIGITAL DATA PROCESSING , Elektrische digitale Datenverarbeitung / B60W CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION , Gemeinsame Steuerung oder Regelung von Fahrzeug-Unteraggregaten verschiedenen Typs oder verschiedener Funktion / G06N COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS , Rechnersysteme, basierend auf spezifischen Rechenmodellen |
Boost-Phase Trajectory Planning with the Nonregular Reachable Area Constraints
DOAJ | 2022
|Reachable domain of spacecraft with a single tangent impulse considering trajectory safety
Online Contents | 2013
|