The extended Kalman filter (EKF) is widely used for the integration of the global positioning system (GPS) and inertial navigation system (INS). It is well known that the EKF performance degrades when the system nonlinearity increases or the measurement covariance is not accurate. For the loosely coupled GPS/INS integration, accurate determination of the GPS measurement and its covariance is not a simple task. An extended state observer (ESO) is proposed for the first time to improve the navigation performance of the loosely coupled GPS/INS integration with accurate measurement modeling. The performance of the proposed method is comprehensively evaluated and analyzed, and comparisons were made with respect to the standard EKF method. Simulations and a field vehicular test were conducted to evaluate the performance of the integrated system using the proposed algorithm. The results demonstrate that the navigation performance of the proposed method can be improved significantly in terms of position and velocity when compared with the EKF method. In the simulation test, the RMS values of the positioning errors are reduced by 52.57%, 48.56%, and 34.16% in the north, east, and vertical directions, respectively. The corresponding percentage for the velocity errors are 48.20%, 43.17%, and 22.65%, respectively. In the field test, the RMS values of the positioning errors are decreased by 40.91%, 47.63%, and 12.21%, respectively. The corresponding percentage for the velocity errors are 42.13%, 31.38%, and 33.86%, respectively. The improvement of attitude accuracy is not obvious as it mainly relies on the quality of the inertial sensors.

扩展卡尔曼滤波器（EKF）被广泛用于全球定位系统（GPS）和惯性导航系统（INS）的集成。众所周知，当系统非线性增加或测量协方差不准确时，EKF性能会下降。对于松散耦合的GPS / INS集成，准确确定GPS测量值及其协方差不是一项简单的任务。首次提出了扩展状态观察器（ESO），以通过精确的测量建模来改善松耦合GPS / INS集成的导航性能。对该方法的性能进行了全面的评估和分析，并对标准EKF方法进行了比较。使用所提出的算法进行了仿真和野外车辆测试，以评估集成系统的性能。结果表明，与EKF方法相比，该方法的导航性能在位置和速度上都可以得到显着改善。在模拟测试中，定位误差的RMS值在北，东和垂直方向分别降低了52.57％，48.56％和34.16％。速度误差的相应百分比分别为48.20％，43.17％和22.65％。在现场测试中，定位误差的RMS值分别降低了40.91％，47.63％和12.21％。速度误差的相应百分比分别为42.13％，31.38％和33.86％。