基于LM_BP神经网络模型的MEMS加速度计温度补偿方法研究

发布时间：2018-04-19 21:45

  本文选题：扭摆式硅微加速度计 + 神经网络　； 参考：《仪表技术与传感器》2015年11期

【摘要】：随着MEMS加速度计应用领域的不断广泛,其温度性能越来越受到重视。在研究扭摆式硅微加速度计结构与温度特性的基础上,采用改进LM_BP神经网络来构建MEMS加速度计的补偿模型,通过实时温度变化优化出温度补偿模型参数,进而实现实时温度补偿。实验结果表明,通过该方法补偿后的标度因数温度系数和全温零偏稳定性分别由252 ppm/℃和16.62 mg/h减小为100 ppm/℃和2.30 mg/h,证明了该温度补偿方法的有效性和可行性。
[Abstract]:With the wide application of MEMS accelerometer, more and more attention has been paid to its temperature performance. Based on the study of the structure and temperature characteristics of the torsion pendulum silicon microaccelerometer, an improved LM_BP neural network is used to construct the compensation model of the MEMS accelerometer, and the parameters of the temperature compensation model are optimized by real-time temperature variation. Furthermore, real-time temperature compensation is realized. The experimental results show that the scaling factor temperature coefficient and the total temperature zero bias stability are reduced from 252 ppm/ 鈩，

本文编号：1774850