卡尔曼滤波的心动阵列实现研究

A Study of the Systolic Array Implementation on Kalman Filtering

基于Faddeeva算法可以实现各种矩阵运算 ,并且可以方便地映射到心动阵列结构上这一事实 ,提出了一种基于Faddeeva算法的卡尔曼滤波心动阵列实现方法 ,并在此基础上设计了两种处理器拓扑结构。分析结果表明 ,这两种处理器结构具有效率高、规整性好和易于扩展等特点。

Based on the fact that the Faddeeva algorithm can practise various matrix manipulations and can be mapped easily into a systolic architecture, an approach of the systolic implementation on Kalman Filtering based on the Faddeeva algorithm is presented. Moreover, two architectures of processors were developed. The analysis results show that these two architectures are efficient,regular and expandable.