1. Analysis of Gravity Disturbance Compensation for Initial Alignment of INS
- Author
-
Zhili Zhang, Xianyi Liu, Shiwen Hao, Zhaofa Zhou, and Chang Zhenjun
- Subjects
Gravity (chemistry) ,Inertial frame of reference ,inertial navigation system ,initial alignment ,General Computer Science ,02 engineering and technology ,Gravity disturbance compensation ,01 natural sciences ,Prime (order theory) ,spherical harmonic model ,0202 electrical engineering, electronic engineering, information engineering ,General Materials Science ,Pitch angle ,Inertial navigation system ,Physics ,Observational error ,020208 electrical & electronic engineering ,010401 analytical chemistry ,Mathematical analysis ,General Engineering ,Spherical harmonics ,0104 chemical sciences ,Gravity of Earth ,lcsh:Electrical engineering. Electronics. Nuclear engineering ,Kalman filtering ,lcsh:TK1-9971 - Abstract
To address the accuracy requirements of initial alignment of high-precision inertial navigation systems (INSs), gravity disturbance compensation for INSs based on a spherical harmonic model is investigated herein. First, the horizontal component of gravity disturbance at an alignment point is calculated using the high-resolution Earth Gravity Model EIGEN-6C4 and then compensated to the initial alignment. Subsequently, the self-alignment algorithm of solidified coordinate frame is used to derive the misalignment angle equation of gravity disturbance affecting the initial alignment. Meanwhile, the coupling relationship between the measurement error of an inertial unit and the gravity disturbance is simulated and analyzed. Finally, a laser strapdown inertial navigation system experiment is performed. The simulation result shows that the pitch angle, roll angle, and heading angle errors reduced by $27.41^{\prime \prime }$ , $- 0.37^{\prime \prime }$ , and $6.72^{\prime \prime }$ , respectively, after the gravity disturbance compensation. Experiment result shows that the alignment performance after compensation has been improved and the heading angle error is reduced by $6.76^{\prime \prime }$ . The simulations and experiments results validate the theoretical analysis.
- Published
- 2020