摘要
决策与规划是实现自动驾驶的关键技术,针对自动驾驶货车在紧急转向避障时需要考虑换道时目标车道的安全性和规划出最优避障轨迹的问题,采用将换道时左右车道上的车辆与自车的相对距离和换道的最小安全距离的差值分别建立模糊关系的方法,通过比较模糊规则推理设计的安全值,选择更安全的车道进行转向避障,为了迅速规划出最优避障轨迹,采用三阶贝塞尔曲线,通过设计4个控制点坐标形成避障曲线,为了防止转向时横向加速度过大导致货车发生侧翻和速度过快与前车发生碰撞,设计车辆的稳定性边界和碰撞边界对控制点进行约束,使用MATLAB中的函数求解出不同车速下的最优换道轨迹,最后使用仿真软件进行仿真验证,结果表明设计出的避障决策和轨迹规划可以安全、有效地避开障碍。
Decision and planning are key technologies for achieving autonomous driving.In response to the issue of considering the safety of the target lane and planning the optimal obstacle avoidance trajectory for autonomous trucks during emergency turning and obstacle avoidance,a fuzzy relationship was established based on the difference between the relative distance between vehicles on the left and right lanes during lane changing and the minimum safe distance between the vehicles during lane changing.The safety values designed by fuzzy rule reasoning were compared,a safer lane for turning was selected to avoid obstacles.In order to quickly plan the optimal obstacle avoidance trajectory,a third-order Bézier curve was used to form an obstacle avoidance curve by designing the coordinates of four control points.In order to prevent the truck from rollover and collision with the vehicle in front due to excessive lateral acceleration during turning,the stability boundary and collision boundary of the vehicle were designed to constrain the control points,the functions in MATLAB was used to solve the optimal lane changing trajectory at different speeds,and finally simulation software was used for simulation verification.The results show that the designed obstacle avoidance decision and trajectory planning can safely and effectively avoid obstacles.
作者
田国富
张森
TIAN Guo-fu;ZHANG Sen(School of Mechanical Engineering,Shenyang University of Technology,Shenyang 110870,China)
出处
《科学技术与工程》
北大核心
2024年第5期2083-2089,共7页
Science Technology and Engineering
基金
辽宁省教育厅科学研究基金(LFGD202006)。
关键词
自动驾驶
转向避障
模糊规则
轨迹规划
换道决策
autonomous driving
steering obstacle avoidance
fuzzy rules
trajectory planning
lane change decision
