微机械陀螺温度混合线性回归补偿方法

Mixed linear regression temperature compensation method for annular-vibrating MEMS gyroscope

环形振动微机电陀螺受温度影响较大,并且还具有很强的自回归特性。针对传统的分段拟合等温度模型均难以精确补偿陀螺受温度影响的问题,提出了一种基于混合线性回归的温度补偿模型。该方法根据混合线性回归模型的特点将陀螺自身的影响以及温度变化等因素考虑到温度补偿模型中,采用多元线性回归方法确定各项的系数,通过对残差的正态检验确定模型是否能够较好的符合陀螺数据的变化规律。验证试验结果表明:补偿后的均值可以减小1～2个数量级,并且该温度误差补偿方法重复性好,具有重大的工程应用参考价值。

Annular-vibrating MEMS gyroscope is significantly affected by temperature and it also has strong self-return characteristics.In view of the problem that traditional method（such as sectional fitting） could not accurately compensate the gyroscope influenced by temperature,a mixed linear regression model is proposed to set up a temperature compensation model.According to the characteristics of mixed linear regression model,this method takes the influences of gyroscope itself and the temperature changes as well as other factors into account.The multiple linear regression method is also used in determining the coefficients.At last,normality test for the residual is used to check if the model can meet the change rule of gyro data.Experimental results show that the mean of the gyroscope can be reduced by one to two orders of magnitude.The compensation method has a good reproducibility,so it has a great value of engineering reference.