频率型硅微传感器是一类基于谐振式测量原理直接输出频率量的传感器，具有稳定可靠、精度高，易与计算机匹配的优势，在导航、制导与传感领域具有极为广泛的应用价值与战略意义。然现有工作存在以下难题制约：1、现有表征方法大都针对静电激励下线性振动本体模型，忽略内外部综合因素影响的特征信息，难以实时反映电热激励下非线性振动特性；2、甚少量化分析非线性振动影响的不确定性和定量分析谐振梁非线性振动影响的强弱，难以准确锁定影响因素引起的测量误差；3、甚少考虑基于数据驱动的自适应补偿算法，难以高效在线生成具备精细化抵消和补偿的模型训练数据。因此本项目针对上述问题开展电热激励下的非线性振动表征工作；完成非线性振动影响的不确定性量化与可抑制性分析；研究基于数据驱动的非线性振动误差自适应补偿方法；在此基础上，通过实验验证和评估非线性振动特性分析方法的有效性，为宽量程和高精度硅微传感器研制提供理论与实践基础。

Frequency-Silicon-Micro sensor is a frequency sensor of high precision based on the principle of resonance. It is of great significance in application occasions such as navigation, guidance and sensing, by virtue of its incomparable advantages on stability, reliability, high accuracy, and computer integration easily. However, current research in this field faces deficiency which is recognized as following: 1. The representation that reflects nonlinear vibration characteristics under electrothermal excitation in real time is absent in most works. Instead, only linear vibration model under electrostatic excitation were concerned, ignoring the characteristic information affected by internal and external factors; 2. There is quite few quantitative analysis of the uncertainty of the impacts by nonlinear vibration, as well as the effects of nonlinear vibration of the resonant beam, leading to the difficulty in how to lock the measurement errors accurately; 3. There is very little consideration of the data-driven adaptive compensation algorithm; also, it is rarely been proposed to generate online model training data including fine offset and compensation efficiently.Hereby, this project aims to tackle the fore-mentioned problems by carrying out nonlinear vibration representation under electrothermal excitation. The uncertainty quantification and restrain analysis of nonlinear vibration effects is expected to be completed at first, followed by studying on data-driven adaptive compensation method of nonlinear vibration error. Based on that, experiments are designed and carried out to verify and evaluate the effectiveness of the nonlinear vibration characteristic analysis method. This project has strong faith in providing theoretical and practical basis for the development of wide range and high precision for Silicon-Micro-Sensors.