Kalman augmented control system design of a coriolis vibratory gyroscope identified using peak picking method

Coriolis vibration gyroscopes (CVG) are among the high-tech navigation systems used in submarines and spacecraft that require particularly precise rotation measurements with high performance parameters. Therefore it is of critical importance. CVG is a sensor that measures the rotation by reading the change in the resonator element vibrating at the resonance frequency due to the Coriolis effect, which occurs as a result of the external rotation, through the electrode elements. In this study, the CVG containing a ring-shaped resonator was modeled as a 2-DOF coupled system with fast dynamics. The resulting coupled model was modeled in terms of slow dynamics using the method of averaging. The electrodes in the system were modeled linearly and a model of the entire system was obtained. After the model was created, peak-picking method which is a well-known modal system identification method was applied to the ring-shaped resonator. This method was applied to models of resonators containing varying natural frequency, quality factor, natural frequency azimuth and time constant azimut mismatch angles establish in the MATLAB-Simulink environment, and the operating limits were examined, the applicability of the method was confirmed and the effects were analyzed. After carrying out the method through simulations, a ring-shaped resonator was tested in a laboratory environment. Then, the peak-picking method was applied to the ring-shaped resonator and system parameters were estimated by correction of the frequency response function obtained from the time domain steady-state solution. Force-to-Rebalance (FTR) control method with four parallel control loops was applied to the system model including the parameter estimation results and the model was examined through the numerical simulations. Feed-through capacitance effects, assembly and electronic components' errors which could not model over the ring-shaped resonator were added to the system as noise. Transient responses and performance values of the FTR control applied system were examined. Then, the Kalman augmented FTR method was design in order to performance enhancement by suppressing the errors that are reflected in the system as noise. In this method, a Kalman filter design was performed with a variable state transition matrix due to the change in the natural frequency and the rotation. The robustness of the designed control method was demonstrated by using different values in the Kalman filter design and the system model. By comparing the FTR and K-FTR methods through the MATLAB-Simulink numerical simulations, an improvement in the performance parameters was achieved. When the K-FTR method is compared with the FTR method, performance enhancement is achieved in the rotation sensitivity and angular random walk (ARW) value, decreased from 6 °/s to 0.06 °/s and from 0.765 °/√hr to 0.046 °/√hr, respectively.