11 December 2025 in IEEE Transactions on Instrumentation and Measurement
The deflection of the vertical (DOV) constitutes a critical error source that fundamentally limits the positioning accuracy of high-precision inertial navigation systems (INS) during long-duration navigation. Building upon our previous work that pioneered the relative estimation method for DOV using double alignment (RE-DA), this study conducts a systematic error analysis of the approach, with particular emphasis on the critical role of precise level attitude determination in DOV estimation. Through observability degree analysis, we propose two innovative multi-position alignment schemes tailored for dual-axis and tri-axis rotational INS (RINS) architectures to achieve enhanced precision in horizontal attitude determination. To address practical implementation challenges, we develop a hybrid compensation strategy integrating inertial alignment principles with a state-error extended Kalman filter. This framework enables real-time compensation for inertial measurement unit (IMU) systematic errors, sensor nonlinearities, and accelerometer lever arm effects. Furthermore, for practical applications characterized by vehicle sway motion, we implement an attitude-tracking-based dynamic error compensation algorithm during horizontal velocity computations. Extensive Monte-Carlo simulations and practical experiments are conducted to evaluate the performance of the proposed approach. Compared with conventional alignment methods, the experiment results indicate superior performance and the accuracy improvement of pitch and roll angles are about 65.8% and 59.1%, respectively.