Return
Prediction of assembly posture errors between microhemispherical resonators and electrode plates
L
X
X
DOI:10.1088/1361-6501/ae6359.png)
Abstract
En 中文
As the core sensitive unit of a resonant gyroscope, the micro-hemispherical resonator has the advantage of a high quality factor, enabling the development of high-precision micro-hemispherical resonant gyroscopes. A key challenge lies in the assembly accuracy between the micro-hemispherical resonator and the electrode plate. Because assembly errors often require repeated adjustments, the assembly process is time-consuming and inefficient. In this work, an assembly posture error identification method for a micro-hemispherical resonator with in-plane electrodes is proposed. First, based on the defined assembly posture error parameters of the in-plane electrodes and the micro-hemispherical resonator, a mathematical model relating assembly posture errors to capacitance is established. Second, this model is used to analyze the influence of posture error parameters on capacitance and to generate capacitance data under different prescribed assembly posture errors. Third, a neural-network regression model is established to map capacitance values to assembly posture error parameters. The capacitance-pose dataset used for model training and testing is generated from the theoretical capacitance model, and the corresponding pose parameters are used as predefined reference labels during training and evaluation. The results show that the proposed model has good regression capability on the constructed capacitance-pose dataset. In addition, assembly experiments were carried out to compare the capacitance uniformity under coarse alignment and fine alignment. The experimental results show that fine alignment can significantly reduce assembly errors and improve capacitance uniformity. This work provides a theoretical basis for capacitance-guided assembly optimization of micro-hemispherical resonators.
Keywords:
microhemispherical resonators
electrode
assembly posture errors
Journal
IF:
3.4
Papers:
2.6K
Citations:
2.3W
