I. Introduction
In THE literature dedicated to linear time-invariant (LTI) model identification [12], [25], many reliable techniques are now available to determine black-box state-space models effectively. For instance, the subspace-based techniques [11], [26], [27] are considered as good solutions for estimating a discrete-time [17] or continuous-time (CT) [1], [8], [18] fully parameterized state-space form as well as the system order accurately. The success of the MOESP, CVA, or N4SID-like techniques probably relies on the use of tools, such as the RQ factorization or the singular value decomposition [7] which are well known for their numerical robustness as well as the nonnecessity to involve iterative gradient-based search algorithms. In order to bypass the suboptimality of the subspace-based identification methods, it can be suggested using the subspace-based state-space matrices estimates as a reliable initial guess for a prediction-error-based technique [12] (see [16], [20], [27], [28] for further details about the theoretical and numerical properties of this standard two-step approach).