摘要
通过对现有加速度传感器静态模型参数辨识方法进行分析,指出传统的算术平均值法在实现加速度传感器静态模型参数辨识中存在的缺陷,提出了数据融合方法,并对其进行了讨论。数据融合方法是将来自同一目标的多源数据加以智能化合成,从而产生比单一数据源更精确更完全的估计和判决。通过分析表明,采用数据融合方法进行参数辨识得到的参数的离散度小于传统的算术平均值法所求参数的离散度,使模型参数的精度有明显提高。因此数据融合方法优于传统的算术平均值法,特别适合于加速度传感器静态模型参数的辨识。该方法的提出为控制系统的实时补偿提供了良好的条件。
By analyzing current methods of the accelerometer stationary model parameter identification, this paper indicates the limitation of the traditional arithmetic average value method, then it puts forward and discusses a new data fusion method. This method intelligently synthesizes the multi - source data to get more exact and more complete estimate and verdict. It is found that the dispersed degree of the parameter acquired from the data fusion method is less than that acquired from the traditional method and the precision of the model parameter is improved greatly. Data fusion method is better than the traditional arithmetic average value method, especially fitting for the accelerometer stationary model parameter identification. This method provides a good condition for the real -time compensation of the control system.
出处
《航天控制》
CSCD
北大核心
2006年第1期79-81,86,共4页
Aerospace Control
关键词
加速度传感器
静态模型
参数辨识
数据融合
Accelerometer Stationary model Parameter identification Data fusion
