Subspace Identification of Linear Time-Periodic Systems With Periodic Inputs

This paper proposes a new methodology for subspace identification of linear time-periodic (LTP) systems with periodic inputs. method overcomes the issues related to computation frequency response LTP by utilizing time-lifted system time-invariant structure instead. The is estimated an ensemble input-output data allows frequency-domain technique be extended systems. time-aliased impulse can then and order-revealing decomposition block-Hankel matrix formulated. consistency proposed proved under mild noise assumptions. Numerical simulation shows that performs better than multiple widely-used time-domain methods when available.