Authors

Abstract

To address the issue that single-satellite navigation systems are prone to signal occlusion and weakening, resulting in insufficient positioning accuracy when power chassis equipment operates in hilly and mountainous areas, a combined navigation scheme integrating Precise Point Positioning (PPP), Real-Time Kinematic (RTK), and an Inertial Measurement Unit (IMU) was proposed. A simulation model was developed using the PSINS (Precision Strapdown Inertial Navigation System) toolbox within the MATLAB environment. A straight-line trajectory was designed to simulate weakly nonlinear operating conditions, while a circular trajectory was used to represent the strong nonlinear continuous-turning conditions. The performance of the Extended Kalman Filter (EKF) was compared with that of the Unscented Kalman Filter (UKF). A combined PPP-RTK-IMU navigation test system was constructed, and both field experiments and verification tests were conducted in hilly and mountainous regions. The results showed that, in the circular trajectory simulation, the standard deviations of eastward, northward, and vertical position errors obtained using the UKF were reduced by 22%, 19%, and 18%, respectively, compared with those of the EKF. In field tests, the UKF demonstrated significantly better consistency with the reference values than the EKF. Results from five repeated field verification tests showed that the average maximum absolute lateral position deviation was 10.96 cm, the mean absolute deviation averaged 3.08 cm, and the average standard deviation was 3.02 cm, all meeting operational requirements. Overall, the findings indicate that the UKF is more suitable for strongly nonlinear operating scenarios encountered in hilly and mountainous terrain, and that the proposed combined navigation system effectively mitigates satellite signal occlusion, thereby meeting the precision requirements of modern agricultural machinery operations.

Abstract in Chinese

针对丘陵山区动力底盘作业时单一卫星导航易受遮挡、信号弱导致定位精度不足的问题，提出精密单点定位（PPP）-实时动态差分（RTK）与惯性测量单元（IMU）深度融合的组合导航方案。基于MATLAB 平台的 PSINS 惯性导航工具箱构建仿真模型，设计直线轨迹模拟弱非线性工况、圆形轨迹模拟强非线性连续转弯工况，对比扩展卡尔曼滤波（EKF）与无迹卡尔曼滤波（UKF）性能；搭建PPP-RTK与IMU组合导航测试系统，开展场地试验和丘陵山区验证试验。结果显示，圆形轨迹仿真中，UKF东向、北向、高程位置误差标准差较EKF分别降低 22%、19%、18%；场地试验中UKF与真实值重合度显著优于EKF；田间试验的横向位置最大绝对偏差均值 10.96cm、平均绝对偏差均值 3.08cm、标准差均值 3.02cm，符合作业要求。结果表明，UKF更适配山区强非线性场景，组合导航系统有助于解决卫星信号遮挡问题，满足农机精细化作业需求。