摘要
为了提高Ｋａｌｍａｎ滤波器的计算效率，本文通过改变算法结构，提出了有效的串行和并行方法以解决目标跟踪和故障诊断问题．串行新算法计算量由ｏ（ｎ３）（ｎ为状态维数）下降到ｏ（ｎ２）；应用于单目标和多目标跟踪时可使计算效率达到普通Ｋａｌｍａｎ滤波算法的２．５６倍．
To get a highly efficient, Kalman filter. and some series and parallel methodsare developed to solve the problems of target tracking or failure detection. It shows that the number of operations in the series method is reduced from o(n3) too(n') (n is the dimension of state). When applied to the problems of target tracking, the efficiency of the new algorithm is 2.56 times higher than the ordinary kalman filter.
出处
《自动化学报》
EI
CSCD
北大核心
1994年第5期605-609,共5页
Acta Automatica Sinica
关键词
KALMAN滤波
并行算法
目标跟踪
Kalman filter
parallel algorithm
target tracking
computationalefficiency.
