基于驾驶员经验的无人驾驶车辆平行泊车操作模型

doi:10.3969/j.issn.1674-8484.2023.02.006

摘要/Abstract

摘要：

提出了一种充分融合驾驶员经验的、低能耗的无人驾驶车辆平行泊车操作模型。将驾驶员平行泊车经验，描述为双梯形曲线图像的方向盘转角操作模型；对模型轨迹关键点位上的3个构型参数，建立平行泊车品质评价体系，得出参数对泊车品质的影响规律；考虑车辆运动学约束以及无碰撞约束，实现双梯形操作模型的优化与校正。在一辆6.6 m的无人驾驶客车上，进行了不同算法的对比验证，并在不同变量条件下进行了适应性验证。结果表明：相比传统的五次多项式轨迹规划和比例—积分—微分控制的泊车方法，用本操作模型，省去了循迹环节，使泊车过程更加流畅，且能耗降低达44.7%。

关键词: 无人驾驶车辆, 平行泊车, 驾驶员经验, 操作模型, 方向盘转角, 轨迹规划

Abstract:

A parallel parking operation model was proposed for driverless vehicles with fully integrated drivers' parking experience and with less parking energy consumption. The driver's parallel parking experiences were summarized and described as an operation model of the steering wheel rotation angle with double trapezoidal curve image. A parallel parking quality evaluation system was established for three configuration parameters at the key trajectory points of the model to find the influence law of the parameters about parking quality. The double trapezoidal operation model was optimized and rectified by considering the vehicle kinematic constraints and the no-collision constraints. The comparative validations were carried out under multiple algorithms with the adaptive validation under different variable situations on a 6.6 m driverless bus. The results show that the double-trapezoidal operation model has a more fluent parking process with a less energy consumption by 44.7% reducing for eliminating the tracking link, compared with the traditional parking methods of quantic-polynomial trajectory-planning and proportional integral-differential control.

Key words: autonomous vehicle, parallel parking, driver experience, operation model, steering wheel rotation angle, trajectory planning

中图分类号:

U461.8

修国涛, 谢辉, 宋康, 毕凤荣. 基于驾驶员经验的无人驾驶车辆平行泊车操作模型[J]. 汽车安全与节能学报, 2023, 14(2): 191-201.

XIU Guotao, XIE Hui, SONG Kang, BI Fengrong. Parallel parking operation model for driverless vehicles based on driver experiences[J]. Journal of Automotive Safety and Energy, 2023, 14(2): 191-201.

图/表 23

参考文献 16

[1]	Parmar J, Das P, Dave S M. Study on demand and characteristics of parking system in urban areas: A review[J]. J Traff Transp Eng, 2019, 7(1): 111-124.
[2]	ZHANG Jiaxu, SHI Zhengtang, YANG Xiong, et al. Trajectory planning and tracking control for autonomous parallel parking of a non-holonomic vehicle[J]. Meas Control, 2020, 53(9-10): 1800-1816. doi: 10.1177/0020294020944961 URL
[3]	ZHANG Bingzhan, LI Zhiyuan, NI Yaoyao, et al. Research on path planning and tracking control of automatic parking system[J]. World Elec Vehi J, 2022, 13(1): 14-26.
[4]	Jhang J H, Lian F L. An autonomous parking system of optimally integrating bidirectional rapidly-exploring random trees* and parking-oriented model predictive control[J]. IEEE Access, 2020, 8(1): 163502-163523. doi: 10.1109/ACCESS.2020.3020859 URL
[5]	LIU Wei, LI Zhiheng, LI Li, et al. Parking like a human: a direct trajectory planning solution[J]. IEEE Trans Intel Transp Syst, 2017, 18(12): 3388-3397. doi: 10.1109/TITS.2017.2687047 URL
[6]	ZHANG Jiren, CHEN Hui, SONG Shaoyu, et al. Reinforcement Learning-Based Motion Planning for Automatic Parking System[J]. IEEE Access, 2020, 8(1): 154485-154501. doi: 10.1109/Access.6287639 URL
[7]	Ján C. Automatic parking control using fuzzy logic[C]// 2019 6th Int’l Conf Adv Contr Circuits Syst (ACCS) & 2019 5th Int’l Conf New Paradigms in Electr Info Tech(PEIT), 2019: 203-208.
[8]	刘颂, 陈慧, 张继仁. 神经网络借鉴驾驶员经验的自动泊车运动规划[J]. 中国集成电路, 2019, 28(Z1): 57-64.
	LIU Song, CHEN Hui, ZHANG Jiren. Automatic parking motion planning based on neural network learning form drivers’ experience[J]. Chin Integ Circuit, 2019, 28(Z1): 57-64. (in Chinese)
[9]	Martin A, Lattarulo R, Zubizarreta A, et al. Trajectory planning for automated buses in parking areas[C]// 2021 25th Int’l Conf Syst Theory, Contr Comput(ICSTCC), 2021: 688-694.
[10]	LI Jing, WU Qingbin, WANG Junzheng, et al. Neural networks-based sliding mode tracking control for the four wheel-legged robot under uncertain interaction[J]. Int’l J Robust Nonli Control, 2021, 31(9): 4306-4323.
[11]	XIE Hui, ZHANG Ze, SONG Kang. A self-optimization algorithm of multi-style smart parking driven by experience, knowledge and data[C]// 2021 5th CAA Int’l Conf Vehi Contr Intell (CVCI), 2021: 1-7.
[12]	Furukawa T, Lavis B, Durrant-Whyte H. Parallel grid-based recursive Bayesian estimation using GPU for real-time autonomous navigation[C]// 2010 IEEE Int’l Conf Robot Automa, 2010: 316-321.
[13]	周扬, 谢辉, 肖蓬勃, 等. 基于主动优化的无人驾驶客车实时性运动规划算法[J]. 汽车安全与节能学报, 2020, 11(4): 476.
	ZHOU Yang, XIE Hui, XIAO Pengbo, et al. Real-time motion planning algorithm for autonomous bus based on initiative optimization[J]. J Automo Safet Energ, 2020, 11(4): 476. (in Chinese)
[14]	Maekawa T, Noda T, Tamura S, et al. Curvature continuous path generation for autonomous vehicle using B-spline curves[J]. Compu Aided Design, 2010, 42(4): 350-359. doi: 10.1016/j.cad.2009.12.007 URL
[15]	Philip A K, Sackey M, Owusu-Ansah P, et al. Automatic parking control algorithms and simulation research based on fuzzy controller[R]. SAE Tech Paper, 2020-01-0135.
[16]	DONG Hangrui, JIN Shangtai, HOU Zhongsheng. Model free adaptive control for automatic car parking systems[C]// Proc 11th World Cong Intell Contr Automa, 2014: 1769-1774.

整车质量，m	2.978 t
轴距，l	2.840 m
前悬，l_f	1.360 m
后悬，l_b	1.690 m
车宽，w	2.360 m
转向系数，α	25
最大方向盘转角，δ_max	485°

整车质量，m	2.978 t
轴距，l	2.840 m
前悬，l_f	1.360 m
后悬，l_b	1.690 m
车宽，w	2.360 m
转向系数，α	25
最大方向盘转角，δ_max	485°

分类	计算耗时 s	成功率 %	总评分 %	总耗时 s	总能耗 kJ
模拟退火法	0.663	86	87.96	14.03	1.108
最优边界干涉法	0.201	100	88.09	14.02	1.092

分类	计算耗时 s	成功率 %	总评分 %	总耗时 s	总能耗 kJ
模拟退火法	0.663	86	87.96	14.03	1.108
最优边界干涉法	0.201	100	88.09	14.02	1.092

长×宽×高	6.605 m ×2.320 m ×2.287 m
整车质量，m	5.8 t
轴距，l	4.400 m
前悬，l_f	1.000 m
后悬，l_b	1.205 m
转向系数，α	17.8
最大方向盘转角，δ_max	485°