Recursive-State-Estimation

  为了得到具有一般相关量测噪声线性系统的递推滤波算法, 将该问题转化为具有相关量测 单值随机向量的滤波问题,根据单值随机向量的线性无偏最小方差估计算法,导出了量测噪声为一 般相关鞅差序列的线性系统的最优递推状态估计滤波算法.通过数值仿真,将该算法与假定量测噪 声不相关时的Kalman 滤波算法进行了比较,证明了该算法的有效性(In order to obtain the recursive f i ltering alg orithm for a linear system w ith general correlat
ed measurement noises, the problem is t ransformed to that of filtering for a single random vector w ith correlated measurements. According to the linear unbiased minimum variance estimation algorithm for a single random vector, a recursive filtering algori thm is presented. A digital simulat ion technique is used such that the new algorithm can be compared w i th the Kalman f i ltering algorithm, assuming that the measurement noises are uncorrelated. Validi ty of the proposed algorithm is thus proved)