基于Frenet坐标下改进人工势场法的无人车局部路径规划

Local Path Planning for Unmanned Ground Vehicles Based on Improved Artificial Potential Field Method in Frenet Coordinate System

人工势场法由于运算量小、精度高,广泛应用于无人车的局部路径规划。针对传统人工势场法存在目标不可达、局部最小值及陷入U型障碍物的问题,提出一种基于Frenet坐标系下改进人工势场法的路径规划算法。构建Frenet坐标系来表述车辆避障运动,简化规划模型,解决路径规划中车辆与所在道路相对位置不易表述的问题。提出安全椭圆模型和预测距离的概念来调整势场影响区域,加入基于Frenet坐标系下的参考线势场及动态速度势场改进斥力场函数,解决车辆在静态和动态下的避障问题。利用数学仿真软件进行仿真,以不同车速在直道和弯道场景中对所提出的路径规划方法进行静态和动态避障仿真实验。研究结果表明:不同车速下的前轮转角、横摆角速度均控制在较小范围内,改进算法可以有效解决传统人工势场法的缺陷,同时与快速搜索随机树(Rapidly-exploring Random Tree,RRT)算法相比,其在避障过程中路径规划计算效率提高了42.8%,改进算法优势明显。

The artificial potential field method is widely used in the local path planning for unmanned ground vehicle(UGV)due to its small amount of computation and high accuracy.For the problems of target unreachability,local minimum and falling into U-shaped obstacles existing in the conventional artificial potential field method,a local path planning algorithm based on the improved artificial potential field method in Frenet coordinate system is proposed.In this paper,the Frenet coordinate system is used to describe the UGV's obstacle avoidance movement,which simplifies the planning model and addresses the difficulty in expressing the relative position of UGV and the road during path planning.A safety ellipse model and the concept of prediction distance are proposed to adjust the influence area of the potential field.Additionally,the repulsive field function is improved by adding the reference line potential field and the dynamic velocity potential field based on the Frenet coordinate system.These modifications enable the UGVs to avoid obstacles under both static and dynamic conditions.The path planning methods are proposed to launch the static and dynamic obstacle avoidance simulation experiments with different vehicle speeds in straight and curved road scenarios using mathematical simulation software.The results demonstrate that the front wheel turning angle and traverse angular velocity at different vehicle speeds are controlled within a small range,and the improved algorithm can effectively solve the defects of the conventional artificial potential field method.Besides,compared with the rapidly-exploring random tree(RRT)algorithm,the computational efficiency of path planning of the improved algorithm in the obstacle avoidance process is improved by 42.8%,and achieves better computational performance.