A Novel Temperature Drift Error Estimation Model for Capacitive MEMS Gyros Using Thermal Stress Deformation Analysis

Abstract

Because the conventional Temperature Drift Error (TDE) estimation model for Capacitive MEMS Gyros (CMGs) has inadequate Temperature Correlated Quantities (TCQs) and inaccurate parameter identification to improve their bias stability, its novel model based on thermal stress deformation analysis is presented. Firstly, the TDE of the CMG is traced precisely by analyzing its structural deformation under thermal stress, and more key decisive TCQs are explored, including ambient temperature variation ∆T and its square ∆T2, as well its square root ∆T1/2; then, a novel TDE estimation model is established. Secondly, a Radial Basis Function Neural Network (RBFNN) is applied to identify its parameter accurately, which eliminates local optimums of the conventional model based on a Back-Propagation Neural Network (BPNN) to improve bias stability. By analyzing heat conduction between CMGs and the thermal chamber with heat flux analysis, proper temperature control intervals and reasonable temperature control periods are obtained to form a TDE precise test method to avoid time-consuming and expensive experiments. The novel model is implemented with an adequate TCQ and RBFNN, and the Mean Square Deviation (MSD) is introduced to evaluate its performance. Finally, the conventional model and novel model are compared with bias stability. Compared with the conventional model, the novel one improves CMG’s bias stability by 15% evenly. It estimates TDE more precisely to decouple Si-based materials’ temperature dependence effectively, and CMG’s environmental adaptability is enhanced to widen its application under complex conditions.