摘要
研究捷联惯组误差系数预测建模问题,由于惯组误差系数漂移因素的复杂性、不确定性以及测试数据的小子样特性,时间序列分析方法难以准确描述漂移规律,预测精度不高。为提高惯组误差系数预测精度,将时间序列分析方法和卡尔曼滤波结合,研究了一种卡尔曼滤波的时间序列预测模型。首先采用时间序列分析方法对惯组测试数据建立时间序列预测方程,然后根据预测方程建立卡尔曼滤波方程,最后依靠卡尔曼递推方程不断地修正预测方程的系数,采用修正的预测模型进行预测。仿真结果表明,采用改进方法可以有效地提高预测精度,能够很好地满足对惯组测试数据分析的要求。
In order to improve the prediction accuracy of inertial measurement unit error coefficient, a time series prediction model based on kalman filtering was proposed, integrating time series analysis with kalman filtering. First- ly, the paper used time series analysis method to build the time series forecasting equation for inertial measurement u- nit test data, then established kalman filtering equation, and finally, revised constantly coefficients of time series model by use of kalman recursion equation. The modified model was used to forecasting for inertial measurement unit test data. The simulation shows that the established forecasting model has high precision which satisfies the index de- mand for inertial measurement unit test data.
出处
《计算机仿真》
CSCD
北大核心
2013年第5期46-49,共4页
Computer Simulation
关键词
捷联惯组
预测
时间序列分析
卡尔曼滤波
误差系数
Strapdown inertial measurement unit
Forecasting
Time series analysis
Kalman filtering
error coefficient
