陀螺随机漂移建模研究

Study on Building the DTG Random Drift Model

介绍了时间序列分析理论及利用时间序列分析方法建立陀螺随机漂移模型的步骤,利用实验测量数据,建立了挠性陀螺的随机漂移 ARMA 模型,并应用一种单隐层 BP 神经网络对 ARMA 模型的参数进行辨识。利用 MATLAB 设计了一套完整的软件,软件包括陀螺漂移数据的检验与预处理、模型形式的选择、模型参数的估计、模型适用性的检验。利用这个软件可以很方便的对所采集到的陀螺漂移数据的随机特性进行分析,并建立陀螺随机漂移误差的 ARMA 模型。

Time-series theory and the approach of Building the gyro random drift model are introduced in the papar, Based on the measured data,An ARMA model of the random drift of Dynamically tuned gyro(DTG)is presented.A Back-propagation neural network is used in identifying model.A suit of software on MATLAB is designed,checking up and pretreatment of the Gyro Random Drift data,choicing the model's form,estimating the model's parameter. checking up the model's applicability are included in the software.The characteristic of The Gyro Random drift can be easily analysed by the software and Building the Gyro Random Drift ARMA model.