Numerical Test for Stability Evaluation of Discrete-Time Systems

In this paper, a new numerical test for stability evaluation of discrete-time systems is presented. It is based on modern root-finding techniques at the complex plane employing the Delaunay triangulation and Cauchy's Argument Principle. The method evaluates if a system is stable and returns possible values and multiplicities of unstable zeros of the characteristic equation. For state-space discrete-time models, the developed test evaluates complex function related to the characteristic equation on the complex plane, so it does not require computation of state-matrix eigenvalues. The proposed method is general as it allows to analyze systems whose characteristic equations are not only polynomials. The verification of the algorithm is presented in benchmarks for both integer- and fractional-order systems.