基于神经网络的加速度传感器动态模型参数辨识

Identifying the Dynamic Model Parameters Based on FLANN for Accelerometer

提出了利用函数连接型人工神经网络(FLANN)实现加速度传感器动态模型参数辨识的方法,该方法以加速度传感器动态标定实验数据为基础,通过FLANN训练来确定加速度传感器传递函数参数,文中介绍了辨识原理以及算法,给出了利用FLANN辨识的加速度传感器动态数学模型.结果表明,这种辨识方法具有精度高、鲁棒性好等优点.

A new approach is presented to identifying the dynamic model parameters of an accelerometer based on FLANN. The parameters are identified by using FLANN based on the dynamic experiment data. The principle and algorithms of FLANN are introduced and the dynamic mathematics model is given for accelerometer. The results show that the approach has the character of high precision and strong robustness.