The method of MEMS gyroscope random error compensation based on ARMA

In order to suppress the random error of micro electro-mechanical system (MEMS) gyroscope, the method based on optimized auto regressive moving average (ARMA) model is proposed to solve problems such as poor extraction of trend items during signal preprocessing and strong subjectivity in ARMA ordering. Firstly, the trend item of MEMS gyroscope data is extracted by empirical mode decomposition (EMD) to obtain a stable and zero-mean time series. In order to reduce the subjective influence of the traditional ARMA ordering method, genetic algorithm theory is used to optimize the ordering process. Choose the order corresponding to the highest fitness function as the modeling order and use the recursive augmented least squares method to estimate the model parameters. So far, a complete mathematical model can be established. Finally, adaptive Kalman filter (AKF) is used to compensate the random error of the MEMS gyroscope. Compensate the same MEMS gyroscope signal before and after optimization. Experiments show that the accuracy of EMD de-trending and optimized ARMA is better than the traditional method. After AKF compensation, under the premise that the mean does not change, the standard deviation is reduced by 27% in this paper. The coefficients of the main random error terms are significantly reduced, which further improves the accuracy of the MEMS gyroscope.