The invention discloses an automatic driving speed planning method, electronic equipment, a storage medium and a computer program product. The method comprises the steps of initializing a Monte Carlo tree at each decision time in the automatic driving process of a vehicle; at each decision time, iteratively executing a Monte Carlo tree search algorithm to search from a root node of the Monte Carlo tree to a termination node of the Monte Carlo tree, the root node representing the state of the initial prediction moment of the prediction time, the termination node representing the state of the termination prediction moment of the prediction time, and the root node representing the state of the initial prediction moment of the prediction time; the Monte Carlo tree search algorithm selects a node to be selected based on a state action value of a connection line, and the state action value of the connection line is determined through reinforcement learning simulation; and taking the action of a connecting line connected with the node to be selected in the Monte Carlo tree at the end of iteration as an optimal action combination, and controlling the vehicle to sequentially execute the actions in the optimal action combination at the prediction moment. According to the method, through Monte Carlo tree search, a complex road traffic environment can be better dealt with, and an optimal strategy combination is selected. In the Monte Carlo tree search algorithm solving process, reinforcement learning is combined for simulation, and an accurate decision can be obtained.
本发明公开一种自动驾驶的速度规划方法、电子设备、存储介质及计算机程序产品。方法包括:在车辆自动驾驶过程中,在每个决策时间,初始化蒙特卡洛树;在每个决策时间,迭代执行蒙特卡洛树搜索算法从所述蒙特卡洛树的根节点搜索到达所述蒙特卡洛树的终止节点,所述根节点表示预测时间的起始预测时刻的状态,所述终止节点表示预测时间的终止预测时刻的状态,所述蒙特卡洛树搜索算法基于连线的状态动作价值值选择待选择节点,所述连线的状态动作价值值通过强化学习模拟确定;将迭代结束时所述蒙特卡洛树中连接待选择节点的连线的动作作为最优动作组合,控制车辆按照在所述预测时刻依次执行所述最优动作组合中的动作。本发明通过蒙特卡洛树搜索,能够更好地应对复杂的道路交通环境,选取最优策略组合。在蒙特卡洛树搜索算法求解过程中,结合强化学习进行模拟,能够得到准确的决策。
Automatic driving speed planning method, electronic equipment, storage medium and computer program product
自动驾驶的速度规划方法、电子设备、存储介质及计算机程序产品
2024-08-06
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 |
Automatic driving speed planning method, electronic equipment, vehicle and storage medium
European Patent Office | 2022
|Decision planning method for automatic driving, electronic equipment and computer storage medium
European Patent Office | 2022
|Automatic driving method and device, computer equipment, readable storage medium and program product
European Patent Office | 2024
|Driving planning method, electronic equipment and computer storage medium
European Patent Office | 2022
|Track planning method, automatic driving equipment and computer storage medium
European Patent Office | 2023
|