基于差分GPS的局部区域无人车路径跟踪算法

为确保无人车在弯曲度变化较大的路径上运行的平稳性及速度变化的连续性,基于差分GPS定位导航系统,综合考虑前视距离、速度和弯曲度之间的关系,提出了一种改进的Pure Pursuit算法。首先,通过差分GPS对局部区域内的路径信息进行采集,基于NMEA协议对GPS接收机输出数据进行坐标转换;其次,结合车辆运动学自行车模型与Pure Pursuit改进算法进行路径跟踪。试验结果表明,无人车以30 km/h的最高车速与5 km/h的最低车速跟踪既定路径时,最大横向跟踪偏差相比既有算法降低55. 4%,平均横向跟踪误差相比既有算法降低23. 98%,达到了较好的跟踪效果。

unmanned vehicle path tracking algorithm based on differential GPS in local area

In order to make sure the unmanned vehicle run more stable and the velocity change more continuous on the path with a large change in curvature, An improved Pure Pursuit algorithm based on differential GPS positioning navigation system was proposed. It took the look-ahead distance, velocity and bending degree into consideration. Firstly, the path information in local area was collected by using differential GPS, and then went on the coordinate transformation for the output data from GPS receiver based on the NMEA protocol. Secondly, Path tracking was carried out by combining the kinematic bicycle model of the vehicle and improved Pure Pursuit algorithm. The experimental results show that the improved algorithm is better. In contrast to the original algorithm, the maximum tracking error in lateral direction can reduce 55.4 percent, And the mean tracking error in lateral direction can reduce 23.98 percent.