The invention relates to a driver individuation-based anthropomorphic lane changing trajectory optimization method, which comprises the following steps of: constructing a constraint of a lane changing trajectory according to a vehicle state of a starting point of a vehicle lane changing behavior decision, and generating a candidate lane changing trajectory by traversing vehicle states of an ending point of all possible lane changing behavior decisions; a maximum entropy inverse reinforcement learning model is constructed, a reward function of the maximum entropy inverse reinforcement learning model is composed of a group of primary functions, the primary functions are designed from two aspects of driving safety and personalized driving styles, the safety of vehicle lane changing and driving style personalization are ensured, and coefficients of the primary functions are determined through inverse reinforcement learning; carrying out personification training on the maximum entropy inverse reinforcement learning model based on a human driver demonstration trajectory; and evaluating the candidate lane changing trajectory based on a reward function of the trained maximum entropy inverse reinforcement learning model, and selecting an optimal lane changing trajectory. Compared with the prior art, the method has the advantages of personification, individuation, high safety and the like.
本发明涉及一种基于驾驶员个性化的拟人变道轨迹优化方法,包括:根据车辆变道行为决策的起点的车辆状态,构建变道轨迹的约束,通过遍历所有可能的变道行为决策的终点的车辆状态,生成候选变道轨迹;构建最大熵逆强化学习模型,所述最大熵逆强化学习模型的奖励函数由一组基函数组成,基函数从驾驶安全、个性化驾驶风格两个方面进行设计,保证车辆变道的安全性和驾驶风格个性化,同时基函数的系数通过逆强化学习进行确定;基于人类驾驶员示范轨迹对最大熵逆强化学习模型进行拟人化训练;基于训练完成的最大熵逆强化学习模型的奖励函数对候选变道轨迹进行评估,选择最优变道轨迹。与现有技术相比,本发明具有拟人化、个性化、安全性强等优点。
Personalized driver-based anthropomorphic lane changing trajectory optimization method
一种基于驾驶员个性化的拟人变道轨迹优化方法
2023-08-04
Patent
Electronic Resource
Chinese
IPC: | 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 |
Lane changing method for reflecting personalized behaviors of driver
European Patent Office | 2020
|Personalized lane changing trajectory planning method based on quintic polynomial
European Patent Office | 2024
|Personalized lane changing decision-making method considering driver heterogeneity
European Patent Office | 2023
|European Patent Office | 2024
|Driver Lane Keeping Characteristic Indices for Personalized Lane Keeping Assistance System
British Library Conference Proceedings | 2017
|