SUBSPACE-BASED METHODS FOR THE IDENTIFICATION OF LINEAR TIME-INVARIANT SYSTEMS

Subspace-based methods for system identification have attracted much a ttention during the past few years. This interest is due to the abilit y of providing accurate state-space models for multivariable linear sy stems directly from input-output data. The methods have their origin i n classical state-space realization theory as developed in the 1960s. The main computational tools are the QR and the singular-value decompo sitions. Here, an overview of existing subspace-based techniques for s ystem identification is given. The methods are grouped into the classe s of realization-based and direct techniques. Similarities between dif ferent algorithms are pointed out, and their applicability is commente d upon. We also discuss some recent ideas for improving and extending the methods. A simulation example is included for comparing different algorithms. The subspace-based approach is found to perform competitiv e with respect to prediction-error methods, provided the system is pro perly excited.