一种MEMS加速度计误差分析与校准方法

MEMS传感器测量中一般存在确定性偏差和随机噪声两类误差,为提高其测量精度和稳定性,以MEMS加速度计为例,在建立其测量误差模型的基础上,提出一个综合的校准方案,同时对这两类误差进行补偿。其中使用Allan方差对随机噪声项进行量化分析,并综合采用了最小二乘法计算确定性校准系数,扩展卡尔曼滤波降低随机噪声等方法。通过实验验证,表明该方案可以有效校准加速度计测量的确定性偏差并降低随机噪声干扰,最终经过误差补偿后的测量数据在精度和稳定性方面都有明显提升。

An error analysis and calibration method of MEMS accelerometer

MEMS sensors are usually susceptible to deterministic biases and stochastic noises, to improve their measurement accuracy and stability, take accelerometer sensor as an example, after a error model is built for it, an integrated calibration method to compensate both kinds of errors at the same time is proposed, in which Allan variance is used to analysis the random noise quantificationally. Besides, several kinds of methods including least square method to calculate deterministic calibration parameters, extended Kalman filter to decrease random noise are used. The experiment result shows that this method can efficiently calibrate the deterministic biases and decrease the influence of random noises, the error compensated measurement data has notable improvement in accuracy and stability.