QLQi Li

Qi Li

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Qi Li has a total of 3 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Online Calibration of Low-Cost IMU With Observability-Aware Variable-Dimension Variational Bayesian Adaptive Kalman Filter

    by Qi Li, Guiling Zhao

    Accurate online calibration of low-cost inertial measurement unit (IMU) is essential for improving the navigation performance of distributed inertial navigation systems (INSs), particularly in cost-sensitive or space-constrained applications. However, conventional calibration methods often suffer from incomplete error modeling and limited state observability, resulting in degraded estimation accuracy. To overcome these limitations, this paper presents a novel online calibration method based on a variable-dimension variational Bayesian adaptive Kalman filter (VD-VBAKF). Inspired by the transfer alignment framework, a dedicated state-space model is formulated using attitude and velocity errors as measurements. A real-time observability evaluation mechanism is proposed by analyzing the evolution of the diagonal elements of the estimation error covariance matrix. The VD-VBAKF further incorporates a variational Bayesian strategy to adaptively estimate the measurement noise covariance, eliminating the need for manual tuning. By dynamically adjusting the initial covariance to reduce unobservable states, the filter improves both estimation robustness and computational efficiency. Experimental validation demonstrates that the proposed method accurately estimates the calibration parameters of low-cost IMUs. Compared to the conventional RSTKF, the VD-VBAKF achieves 36.2%, 20.8%, and 33.8% improvements in attitude, velocity, and position accuracy, respectively.

  • A System-Level Calibration Method for SINS Based on Observability Analysis

    by Guiling Zhao, Qi Li, Weidong Liang, Hongxing Zhao

    In the strapdown inertial navigation system (SINS), which is based on a high-precision rotary table for system-level calibration, the complexity of the transposition scheme and the long calibration time will inevitably lead to low calibration accuracy of the error parameters. These calibration results will seriously affect the navigation accuracy of SINS. In addition, the small observability of the error parameters often leads to unreliable results of system-level calibration, thus affecting the navigation accuracy of SINS. Here, we propose an improved observability calculation method and design a system-level calibration scheme with 4-position continuous rotation. First, we establish a calibration error model containing 24 error parameters and a 30-dimensional Kalman filter model. Then, based on the observability analysis method, we innovatively demonstrate the feasibility of the 3-axis rotation scheme. Finally, an efficient system-level calibration scheme for 4-position continuous transposition is designed, which can calibrate all the state variables. At the same time, compared with the popular 19-position calibration method, the calibration scheme in this paper improves the positioning accuracy by 20.49% and the velocity accuracy by 13.77%.