A new method for the pole estimation of linear time-invariant systems using singular value decomposition

A new method is presented for the pole estimation of linear time invariant (LTI) systems. For a single input case, this method gives estimation of each pole of the system. For a multiple input case, it estimates the repeated system poles with multiplicity equal to the number of inputs. The new method employs the singular value decomposition technique (SVD), and is shown to be numerically robust against measurement noises.