A CLASS OF SUBSPACE MODEL IDENTIFICATION ALGORITHMS TO IDENTIFY PERIODICALLY AND ARBITRARILY TIME-VARYING SYSTEMS

Subspace model identification algorithms that allow the identification of a linear, time-varying (LTV) state space model from an ensemble se t of input-output measurements are presented in this paper. Each pair of input and output sequences in this ensemble is recorded when the un derlying system to be identified undergoes the same time-varying behav ior. The algorithms operate directly on the available ensemble of inpu t-output data and are a generalization of the recently proposed Multiv ariable Output Error State sPace (MOESP) class of algorithms to this e nsemble type of identification problems. A special case is considered in this paper, where the repetition of this time-varying behavior is i ntrinsic, namely in periodically time-varying systems. An example of i dentifying a multirate sampled data system from a recorded input and o utput sequence demonstrates some of the capabilities of the presented subspace model identification algorithms.