梳状音叉MEMS陀螺漂移时间序列建模方法研究

根据一种典型结构的梳状音叉MEMS陀螺的非随机性误差特性,借助ARIMA模型对其进行描述。在ARIMA模型参数估计的过程中,引入虚拟白噪声来补偿参数估计过程中的由于噪声未知而引入的误差,根据样本的自相关函数、偏相关函数,并结合AIC与BIC准则确定模型阶数,利用最大似然估计对模型参数进行估计,在一定近似条件下,得到梳状音叉MEMS陀螺非随机性误差模型的形式。实验结果表明:采用该方法所确定的模型能够精确地描述MEMS陀螺的漂移特性,预测陀螺的输出。

Research on random drift time series modeling for comb-drive tuning fork MEMS gyroscope

Based on the ARIMA model,the non-stochastic error feature of the comb-drive tuning fork MEMS gyroscope is studied.In the process of the estimating of the parameters of the model,the pseudo white noise is introduced to compensate the error by the unknown error.The order of the ARIMA model is determined according to the combining with autocorrelation function,the partial autocorrelation function and both AIC rules and BIC rules.The maximum likelihood function is employed to estimate the parameters of the model.On the approximate condition,the error model of the comb-drive tuning fork MEMS gyroscope is found,and the experimental results demonstrate that the model can accurately describe the drift feature and predict the output of the MEMS gyroscope respectively.