Dynamic Performance Comparison of Two Kalman Filters for Rate Signal Direct Modeling and Differencing Modeling for Combining a MEMS Gyroscope Array to Improve Accuracy.

摘要: In this paper, the performance of two Kalman filter (KF) schemes based on direct estimated model and differencing for input rate signal was thoroughly analyzed compared combining measurements a sensor array to improve accuracy microelectromechanical system (MEMS) gyroscopes. The principles noise reduction were presented KF algorithms designed obtain optimal estimates. in modeled with random walk process treated as state. model, operation established between outputs gyroscope array, then estimation achieved by compensating estimations bias drifts component Finally, dynamic simulations experiments six-gyroscope implemented compare models. 1σ error gyroscopes reduced from 1.4558°/s 0.1203°/s constant test 0.5974°/s model. filtered both models could reflect amplitude variation swing displayed factor about three noise. Results illustrate that is much higher than or lower variation. A similarity KFs' observed if has high