[1]赵花蕊,曹仰杰.数据和规则混合驱动的全自动代客泊车轨迹规划[J].计算机技术与发展,2025,(07):148-155.[doi:10.20165/j.cnki.ISSN1673-629X.2025.0048]
　ZHAO Hua-rui,CAO Yang-jie.Data and Rule Hybrid-driven Motion Planning for Fully Automated Valet Parking[J].,2025,(07):148-155.[doi:10.20165/j.cnki.ISSN1673-629X.2025.0048]
点击复制

数据和规则混合驱动的全自动代客泊车轨迹规划()

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:
期数:
2025年07期
页码:
148-155
栏目:
人工智能
出版日期:
2025-07-10

文章信息/Info

Title:
Data and Rule Hybrid-driven Motion Planning for Fully Automated Valet Parking
文章编号:
1673-629X(2025)07-0148-08
作者:
赵花蕊1; 曹仰杰2
1. 河南省平台经济发展指导中心,河南 郑州 450008;
2. 郑州大学 网络空间安全学院,河南 郑州 450003
Author(s):
ZHAO Hua-rui1; CAO Yang-jie2
1. Platform Economy Development Guidance Center of Henan Province,Zhengzhou 450008,China;
2. School of Cyberspace Security,Zhengzhou University,Zhengzhou 450003,China
关键词:
全自动代客泊车系统; 深度强化学习; 混合A*; 课程学习; 轨迹规划
Keywords:
fully automated valet parking system; deep reinforcement learning; hybrid A*; curriculum learning; trajectory planning
分类号:
TP391
DOI:
10.20165/j.cnki.ISSN1673-629X.2025.0048
摘要:
该文研究了全自动代客泊车系统中的轨迹规划,并提出了一种创新的基于深度强化学习的方法。 当前的路径规划技术主要依赖几何算法,这些方法在复杂停车环境中面临诸多限制,尤其是在处理动态障碍物和环境变化不确定性时。此外,基于优化的策略虽然理论上有效,但其计算复杂度较高,难以实现实时响应,限制了其在实际应用中的可行性。 该文提出了一种数据与规则混合驱动泊车轨迹规划方法。 该方法通过利用历史数据和经验规则,显著提高了系统的可扩展性和泛化能力。 值得注意的是,该方法不依赖于实时交互获取其他车辆的精确物理信息,使其更加适合当前实际应用场景中信息不完全或传感器受限的情况。 此外,采用课程学习和混合 A*算法来加速强化学习模型的收敛速度,通过逐步增加任务复杂度,提升模型对环境变化的适应能力。 实验结果显示,该方法在复杂自动泊车任务中的表现优异,能够有效实现高效、安全的泊车操作,充分展现了其在全自动代客泊车系统中的应用潜力。
Abstract:
We investigate trajectory planning in fully automated valet parking systems and propose an innovative approach based on deep reinforcement learning. Current path planning techniques primarily rely on geometric algorithms,which face numerous limitations in complex parking environments,especially when dealing with dynamic obstacles and uncertainties in environmental changes. Although op-timization-based strategies are theoretically effective,their high computational complexity hinders real-time responsiveness,limiting their feasibility in practical applications. We introduce a hybrid data and rule driven method for parking trajectory planning that significantly enhances the system’s scalability and generalization capabilities by leveraging historical data and heuristic rules. Notably,this approach does not depend on real-time interactions to obtain precise physical information about other vehicles,making it more suitable for current practical scenarios, particularly in situations with incomplete information or limited sensors. Furthermore,we employ curriculum learning and a hybrid A * algorithm to accelerate the convergence speed of the reinforcement learning model by gradually increasing task complexity, enhancing the model’s adaptability to environmental changes. Experimental results demonstrate that the proposed method performs exceptionally well in complex automated parking tasks, effectively achieving efficient and safe parking operations, thereby showcasing its potential application in fully automated valet parking systems.

相似文献/References:

[1]赵 纯,董小明.基于深度 Q-Learning 的信号灯配时优化研究[J].计算机技术与发展,2021,31(08):198.[doi:10. 3969 / j. issn. 1673-629X. 2021. 08. 034]
　ZHAO Chun,DONG Xiao-ming.Research on Signal Timing Optimization Based on Deep Q-Learning[J].,2021,31(07):198.[doi:10. 3969 / j. issn. 1673-629X. 2021. 08. 034]
[2]况立群,冯 利,韩 燮,等.基于双深度 Q 网络的智能决策系统研究[J].计算机技术与发展,2022,32(02):137.[doi:10. 3969 / j. issn. 1673-629X. 2022. 02. 022]
　KUANG Li-qun,FENG Li,HAN Xie,et al.Research on Intelligent Decision-making System Based on Double Deep Q-Network[J].,2022,32(07):137.[doi:10. 3969 / j. issn. 1673-629X. 2022. 02. 022]
[3]高文斌,王 睿,王田丰,等.基于深度强化学习的 QoS 感知 Web 服务组合[J].计算机技术与发展,2022,32(06):92.[doi:10. 3969 / j. issn. 1673-629X. 2022. 06. 016]
　GAO Wen-bin,WANG Rui,WANG Tian-feng,et al.QoS-aware Service Composition Based on Deep Reinforcement Learning[J].,2022,32(07):92.[doi:10. 3969 / j. issn. 1673-629X. 2022. 06. 016]
[4]詹 御,张郭健,彭麟杰,等.基于 DRL 的 MEC 卸载网络竞争窗口优化[J].计算机技术与发展,2022,32(06):99.[doi:10. 3969 / j. issn. 1673-629X. 2022. 06. 017]
　ZHAN Yu,ZHANG Guo-jian,PENG Lin-jie,et al.Optimization of Contention Window of MEC Offloading Network Based on DRL[J].,2022,32(07):99.[doi:10. 3969 / j. issn. 1673-629X. 2022. 06. 017]
[5]牟轩庭,张宏军,廖湘琳,等.规则引导的智能体决策框架[J].计算机技术与发展,2022,32(10):156.[doi:10. 3969 / j. issn. 1673-629X. 2022. 10. 026]
　MU Xuan-ting,ZHANG Hong-jun,LIAO Xiang-lin,et al.Rule-guided Agent Decision-Making Framework[J].,2022,32(07):156.[doi:10. 3969 / j. issn. 1673-629X. 2022. 10. 026]
[6]林泽阳,赖 俊,陈希亮.基于课程学习的深度强化学习研究综述[J].计算机技术与发展,2022,32(11):16.[doi:10. 3969 / j. issn. 1673-629X. 2022. 11. 003]
　LIN Ze-yang,LAI Jun,CHEN Xi-liang.An Overview of Deep Reinforcement Learning Based on Curriculum Learning[J].,2022,32(07):16.[doi:10. 3969 / j. issn. 1673-629X. 2022. 11. 003]
[7]吕相霖,臧兆祥,李思博,等.基于注意力的循环 PPO 算法及其应用[J].计算机技术与发展,2024,34(01):136.[doi:10. 3969 / j. issn. 1673-629X. 2024. 01. 020]
　LYU Xiang-lin,ZANG Zhao-xiang,LI Si-bo,et al.Attention-based Recurrent PPO Algorithm and Its Application[J].,2024,34(07):136.[doi:10. 3969 / j. issn. 1673-629X. 2024. 01. 020]
[8]龚亮亮,张 影,张俊尧,等.基于深度强化学习的任务卸载和资源分配优化[J].计算机技术与发展,2024,34(04):116.[doi:10. 3969 / j. issn. 1673-629X. 2024. 04. 018]
　GONG Liang-liang,ZHANG Ying,ZHANG Jun-yao,et al.Joint Optimization of Task Offloading and Resource Allocation Based on Deep Reinforcement Learning[J].,2024,34(07):116.[doi:10. 3969 / j. issn. 1673-629X. 2024. 04. 018]
[9]聂雷,张明萱,黄庆涵,等.基于Double DQN的双模式多目标信号配时方法[J].计算机技术与发展,2024,34(08):143.[doi:10.20165/j.cnki.ISSN1673-629X.2024.0124]
　NIE Lei,ZHANG Ming-xuan,HUANG Qing-han,et al.A Dual-mode Multi-objective Signal Timing Method Based on Double DQN[J].,2024,34(07):143.[doi:10.20165/j.cnki.ISSN1673-629X.2024.0124]
[10]王钰童,顾进广*.边缘场景下基于DDQN的容器组调度策略[J].计算机技术与发展,2024,34(09):16.[doi:10.20165/j.cnki.ISSN1673-629X.2024.0163]
　WANG Yu-tong,GU Jin-guang*.Container Group Scheduling Optimization Strategy Based on DDQN in Edge Scenarios[J].,2024,34(07):16.[doi:10.20165/j.cnki.ISSN1673-629X.2024.0163]

更新日期/Last Update: 2025-07-10