基于BP神经网络模型的MEMS加速度计误差补偿方法  

Error compensation method for MEMS accelerometer based on BP neural network model

作　　者：郭卫[1] 杨鹏飞[1] 张武刚[1] 王巍[1] GUO Wei;YANG Pengfei;ZHANG Wugang;WANG Wei(Xi'an University of Science and Technology,Xi,an 710054,China)

机构地区：[1]西安科技大学,陕西西安710054

出　　处：《中国测试》2018年第3期109-113,共5页China Measurement & Test

摘　　要：现阶段普遍采用多元线性回归对加速度计误差建模,并利用最小二乘法对模型参数辨识,但其对加速度计精度提高有限,因此该文提出一种基于BP神经网络模型的MEMS加速度计误差补偿方法。该方法利用BP神经网络建立加速度计误差模型,通过多位置翻滚进行实验数据测量,并对模型进行训练,最后利用训练好的模型对加速度计误差进行补偿。比较多元线性回归和BP神经网络建模对加速计误差补偿结果,其标准偏差分别为0.001 9 g和0.000 16 g。结果表明误差下降一个数量级,说明BP神经网络能有效地补偿加速度计误差。At present,multiple linear regression is commonly used for the modeling accelerometer of the error of and the least square method is used to identify the parameters of the model.However,this method is limited to improve the accuracy of accelerometer,thus an compensation error method for MEMS accelerometer based on BP neural network model is proposed.method The establishes a model for the error of accelerometer based on BP neural network,data measures through multi-position rolling and trains the model and finally compensates the error accelerometer of with the trained model.Comparison of the error compensation results of accelerometer the by using multiple linear regression and BP neural network modeling,the deviation standard is respectively 0.0019 g and 0.00016 g.Experimental results show that the error by decreases an order of magnitude and it shows BP neural network can effectively compensate the error accelerometer of accelerometer.

关 键 词：加速度计 BP神经网络 多元线性回归 误差补偿模型 

分 类 号：TH-39[机械工程] TP183[自动化与计算机技术—控制理论与控制工程]