Closed-loop system identification using subspace-based methods

System identification using subspace-based methods has been a hot research topic in the past few years. Subspace methods fit state space models directly from the input and output signals measured from systems. In this paper we present a new way to identify a LTI system operating in a closed-loop experiment. One of the MOESP identification methods is adapted to identify the deterministic part of system operating in closed-loop experiment with nonwhite process noise and white output noise. A state space model will be computed to provide the relationship between the input and output of the plant. A simulation example is also provided to show how the algorithm works.