Robust Output-Feedback Control for Microelectromechanical System Gyroscopes with Guaranteed Transient Performance

In this article, a robust output-feedback control strategy capable of guaranteeing transient performance is proposed for microelectromechanical system (MEMS) gyroscopes. First, to ensure the system stability in the presence of the total uncertainties stemming from external disturbances and parameter perturbations, an estimator-based minimum-learning-parameter neural network (EMLPNN) observer actuated by estimated errors rather than tracking errors is investigated, which significantly mitigates the system chattering and computational complexities inherent in existing neural network (NN) observers. Subsequently, acknowledging the stringent time constraints prevalent in practical MEMS measurement tasks, a modified guaranteed performance control (MGPC) approach is introduced to enhance system transient responses. By devising a pair of novel performance envelopes characterized by appointed time and compact transient range, the presented technique is capable of ensuring overshoot-free convergence of position tracking errors within prescribed time requirements. The meaningful merits of the proposed scheme lie in circumventing the computational complexities associated with existing NN observers and mitigating chattering issues induced by tracking errors. Moreover, the system states can be effectively constrained within predefined regions by applying the suggested MGPC strategy, facilitating a prior adjustment. Furthermore, the stability of the closed-loop system is substantiated through Lyapunov analysis, while the effectiveness and superiority are demonstrated by simulation and experimental results.