Intermittently Dynamic Fuzzy Learning-Based Tracking Control of Amplitude Signals for Vibratory Gyroscopes with Composite Identification

This paper investigates an intermittently dynamic fuzzy learning-based tracking control (IDFL-TC) method of amplitude signals for vibratory gyroscopes. Compared with learning-based tracking control (L-TC) and fuzzy learning-based tracking control (FL-TC) methods, the IDFL-TC method with dynamic pruning technology can improve the identification accuracy of gyroscope dynamics and reduce calculation times by intermittently updating the output of nodes and dynamically recruiting or removing nodes. Subsequently, a composite identification mechanism is introduced into the control method to further improve the identification accuracy. A novel tracking controller is developed to well track the reference amplitude signal and guarantee the boundedness of the tracking error. Finally, comparative experimental results demonstrate that the proposed IDFL-TC method can obtain higher control precision of amplitude signals for vibratory gyroscopes with fewer calculation time in comparison to the L-TC and FL-TC methods.