摘要
目前,抗差Kalman滤波一般采用独立等价权形式,观测值相关时等价权的确定是一个难题.从抗差估计的原始定义出发,首先直接对观测残差进行限制,然后利用观测残差与状态预报值残差的关系对状态预报值残差进行限制,这样可同时消除观测模型误差和状态模型误差的影响.此外,本方法无需考虑观测值是独立的还是相关的.最后给出了算例,结果表明,该方法是切实可行的.
The equivalent weights of robust Kalman filtering generally use an independent form, and the determining of values of equivalent weights is a problem when observations are relevant. Starting from the original definition of robust estimation,the direct observation residuals are restricted firstly,then the state prediction residuals are restricted by using the relationship of state prediction residual and observation residuals. The observation model errors and state model errors can be eliminated at the same time by this method. In addition, this method does not need to consider the observations are independent or related. Finally, the results of an example show that the method is feasible.
出处
《河南科学》
2012年第9期1237-1242,共6页
Henan Science
基金
国家自然科学基金项目(40974009
40474007)
河南省基础与前沿技术研究计划项目(082300410240)
郑州市科技计划攻关项目(0910SGYG21198)
关键词
抗差滤波
动态导航
等价权
残差
robust filtering
dynamic navigation
equivalent weight
residual
