基于预测误差法的加速度传感器动态模型参数辨识

加速度传感器动态模型对研究与分析加速度传感器的动态特性与动态误差补偿具有重要作用。针对加速度传感器动态模型的参数辨识,提出了一种基于预测误差法的加速度传感器动态模型参数辨识方法,该方法将加速度传感器的状态空间模型转化为线性带外生输入的自回归滑动平均(ARMAX)模型,获得其最优一步预测输出的表达式,并通过求解加速度传感器最优一步预测输出极小化误差准则函数,实现加速度传感器动态模型参数的最优辨识。实验结果表明,该方法有效地实现了加速度传感器动态模型的参数辨识,所得加速度传感器动态模型具有较高的精度,能描述加速度传感器的动态特性。

Dynamic model parameter identification of the acceleration sensor based on the prediction error method

The dynamic model of the acceleration sensor plays an important role in dynamic characteristic analysis and dynamic error compensation. For the parameter identification of dynamic model of the acceleration sensor, a parameter identification method using the prediction error method for the acceleration sensor dynamic model is proposed. The method builds a linear autoregressive moving average with exogenous input(ARMAX) model according to the state space model of acceleration sensor, in order to obtain the optimal predicted output expression. Finally, it can achieve optimal parameters of the acceleration sensor dynamic model by solving the system optimal step prediction output to minimize the error criteria function. The experimental results show that the method can effectively realize the parameter identification of the acceleration sensor dynamic model, and the dynamic model has higher accuracy, which can describe the dynamic characteristics of the acceleration sensor.