This is documentation for an obsolete product.
Current products and services

 Documentation /  Advanced Numerical Methods /  Model Identification /  Time-Domain System Identification /

OverviewIdentification Using Markov Parameters

8.1 Time-Domain System Identification

The problem of system identification in the time domain amounts to identifying matrices A, B, C, and D of the state-space system given a set of measured time-domain data. Only the discrete-time state-space system is considered here. Thus, given a set of output responses, which correspond to a set of input signals, the problem is to identify the matrices , , , and that satisfy the discrete-time state-space system:

Available are an infinite number of models that can predict the identical responses for any particular input. The one with the smallest state-space dimensions, called the minimal state-space realization, is of interest.

Two methods are implemented here. The first method is based on the explicit construction of Markov parameters. The second is a direct method in the sense that the system matrices are computed directly from the input-output data. Both methods are numerically effective.

OverviewIdentification Using Markov Parameters