Inertial navigation using low cost micro-electromechanical systems (MEMS) inertial measurement units (IMUs) loses its accuracy after a short time. Although integration with auxiliary sensors, such as Doppler Velocity Log (DVL), control the rate of INS drift error, Inertial Navigation System (INS)-DVL navigation using low-cost IMUs has not yet achieved sufficient accuracy, because of un-observability of gyro drift error (low-band frequency error). One of the ways to increase accuracy of this algorithm is pre-filtering the gyro. In the past works, stochastic models was used in the gyro pre-filter. But in this paper longitudinal and lateral deterministic dynamic model of vehicle are used in the gyro pre-filter process. These models include the coupling of linear and angular velocity components. This coupling allows the use of DVL to improve gyro accuracy by filtering the low-band frequency errors. It is shown that in the conventional INS-DVL algorithm, the mentioned error bias is not observable, but in the proposed method, it is observable. The proposed algorithm is evaluated in a simulation and two field tests. In the simulation, gyro pre-filering by first order Gauss Markov model reduces the Root Mean Square (RMS) position error by 25% compared to when the gyro was not pre-filtered. The gyro pre-filter using the proposed method reduces this error by 62%. In the field tests, these error reductions reaches 19% and 41%, respectively.
MEMS-Based Gyro De-Noising via AUV Dynamic for Enhancing Navigation
IEEE Transactions on Intelligent Vehicles ; 9 , 10 ; 6208-6216
01.10.2024
2578215 byte
Aufsatz (Zeitschrift)
Elektronische Ressource
Englisch
Research on signal de-noising technique for MEMS gyro
IEEE | 2010
|The Standing Calibration Method of MEMS Gyro Bias for Autonomous Pedestrian Navigation System
Online Contents | 2017
|Magnetically suspended MEMS spinning wheel gyro
IEEE | 2005
|Magnetically Suspended MEMS Spinning Wheel Gyro
Online Contents | 2005
|