Adaptive control system for a vibrational MEMS gyroscope with time-varying rotation rates

This dissertation presents a new adaptive control system to control both axes of a vibrational MEMS gyroscope. Compared to the adaptive controllers introduced in current literature, our control system is specifically designed to handle a time-varying rotation rate. Such a control system must learn to control the gyroscope while it is in operation, and while some of the unknown parameters in gyroscope are time-varying. This adaptive control system can improve the performance of MEMS gyroscope and compensate for some of the fabrication defects.; The main system tasks are driving the drive axis to resonance, and sensing the Coriolis force created along the sense axis to determine the rotation rate. We apply a time-invariant adaptive controller on the drive axis to tune its natural frequency to resonance and regulate the amplitude of vibration. PI-like adaptation laws are used to update the controller parameters. A Lyapunov approach combined with an invariant set argument is employed to attain both the control and the adaptive laws. For sense axis control, we choose a force-to-rebalance mode of operation. We explore a new time-varying adaptive controller on sense axis to drive its output to zero. The resulting control signal also becomes the measurement of rotation rate. A polynomial approximation approach is used to represent the time-varying rotation rate. An adaptive law is developed to estimate the unknown coefficients of the polynomial. We apply another adaptive law to identify the unknown quadrature error term and remove it in our control effort. A Lyapunov approach is used to obtain both the control and the adaptive laws.; Finally we combine the independent adaptive controllers for both axes to implement the control system. We simulate this control system on a model of a Berkeley Z-axis gyroscope. The simulation results verified the effectiveness of our control system. We also built and tested an analog circuit for adaptively tuning and controlling the drive axis. The experimental results agree well with simulation results.