Recursive Identification of Linear Systems

摘要: Let the three matrices $\sum (N) = (G(N),F(N)H(N))$ define a linear constant system of least degree which realizes set numbers $f_1 , \cdots ,f_N $ regarded as partial impulse response system. An algorithm has been developed for recursively calculating minimal realizations each $N 1,2, such that \[ \sum {(N - k)} \subseteq {(N) .} \] This differs from previous ones, B. L. Ho’s, in there is recursion on N well. Because this, no priori guess order required. Moreover, an addition terms to initial sequence causes computation only few new elements. When combined with another factoring covariance given permits recursive identification random systems. No earlier methods seem appear literature. Finally, categorical description abstract given.