首页|面向自动驾驶的公交车辆进出港湾式停靠站安全轨迹模型

面向自动驾驶的公交车辆进出港湾式停靠站安全轨迹模型

扫码查看
为研究公交车辆安全平稳进出港湾式停靠站的轨迹,提出适用于自动驾驶场景的进出站安全轨迹规划方法.基于阿克曼转向原理,构建车身4个顶点运动轨迹方程,针对站内无车和有车两种工况,建立避障约束、运动学约束及起终点位姿约束,选择五次多项式曲线作为公交车辆进出港湾式停靠站安全轨迹模型,采用线性规划和遗传算法标定模型参数,采用数值模拟得到车辆安全进出站轨迹,并利用实车道路实验验证模型的可行性.研究结果表明:五次多项式曲线适用于公交车辆进出港湾式停靠站安全轨迹,规划轨迹与实际运行轨迹平均绝对百分比误差(MAPE)不超过10%,提出的轨迹规划方法可为自动驾驶公交车辆安全可靠运行提供一定的理论依据.
Safety Trajectory Model for Bus Entering and Leaving Harbor-shaped Stops for Autonomous Driving
To study a safe and smooth trajectory for buses during entering and leaving the harbor-shaped stops,a safe trajectory planning method for autonomous driving scenarios was proposed.The trajectory equations for four vertexes of the vehicle body were constructed based on Ackermann's steering principle.The obstacle avoidance constraints,kinematic constraints,and pose constraints of the starting and ending points were established for the two working conditions,including the vacant and occupied bus stops.The quintic polynomial curve was used to model the safety trajectory for buses when entering and leaving the harbor-shaped stops.The linear programming and genetic algorithm were used to calibrate model parameters.The entering and leaving trajectories were achieved by using numerical simulation.The real-vehicle road experiments were carried out to verify the feasibility of the model.The results show that the quintic polynomial curve is suitable for the safe trajectory of buses entering and leaving the harbor-shaped stops.The mean absolute percent error(MAPE)between the planned and experimental trajectories is no more than 10%.The proposed trajectory planning method can provide a theoretical basis for self-driving buses'safe and reliable operation.

traffic engineeringbusessafety trajectoryharbor-shaped stopautonomous drivingquintic polynomialgenetic algorithm

张文会、张攀、伊静

展开 >

东北林业大学土木与交通学院,黑龙江哈尔滨 150040

交通工程 公交车辆 安全轨迹 港湾式停靠站 自动驾驶 五次多项式 遗传算法

国家重点研发计划黑龙江省重点研发计划

2017YFC0803901JD22A014

2024

应用基础与工程科学学报
中国自然资源学会

应用基础与工程科学学报

CSTPCD北大核心
影响因子:0.895
ISSN:1005-0930
年,卷(期):2024.32(3)
  • 26