| 摘要: |
| 结合车载全球导航卫星系统（GNSS）与惯性测量单元（IMU）组合定位系统,针对传统定位算法中实际路况变化带来的数据融合噪声误差的问题,以车载传感器收集的运动数据作为阈值条件进行自适应滤波动态切换,提高组合定位系统的稳健性与适应性. 提出一种改进的自适应动态组合定位方法,将扩展卡尔曼滤波（EKF）与无迹卡尔曼滤波（UKF）算法,基于实时运动模型结合,以此抑制在车辆运动模型变化时传统单一算法产生的误差干扰.改进后算法的均方根误差（RMSE）相较于传统的EKF滤波与UKF滤波算法分别提升了约75.26%和58.48%.在高架桥下的实际车载场景实验中,改进算法的平均距离误差为2.32 cm,相较于改进前的定位性能提升了约61.65%.在复杂的城市交通的环境下,能够实现精准定位. |
| 关键词: 定位系统 运动模型 卡尔曼滤波 自适应滤波 |
| DOI：10.3969/J.ISSN.1000-5137.2024.03.006 |
| 分类号:TN967.2 |
| 基金项目: |
|
| Improved positioning algorithm based on vehicle-mounted GNSS and IMU positioning system |
|
LI Tianchen, ZHU Sulei, WANG Yang, LYU Hailin
|
|
College of Information, Mechanical and Electrical Engineering, Shanghai Normal University, Shanghai 201418, China
|
| Abstract: |
| To solve the problem of data fusion noise errors caused by changes in actual road conditions in traditional positioning algorithms, the motion data collected by on-board sensors was adapted as a threshold condition for adaptive filtering and dynamic switching to enhance the robustness and adaptability of the integrated positioning system which combined vehicle-mounted global navigation satellite system （GNSS） with inertial measurement unit （IMU） positioning system. An improved adaptive dynamic positioning algorithm was proposed, combining the extended Kalman filter（EKF） with the unscented Kalman filter（UKF） based on real-time motion models, as a way to suppress the error disturbance generated by the traditional single algorithm when the vehicle motion model changed. Compared to the EKF and UKF, the root mean square error（RMSE） of the improved algorithm was reduced by 75.26% and 58.48% respectively. The average distance error of the improved algorithm in the actual in-vehicle scenario beneath an overpass was 2.32 cm, an improvement of roughly 61.65% over the performance before the change. These results demonstrated the efficacy of the improved algorithm in the complex urban traffic. |
| Key words: positioning system motion model Kalman filtering adaptive filtering |
