Frequency-domain worst-case identification of multiple input multiple output errors-in-variables models

This paper extends a frequency-domain worst-case identification method for single input single output (SISO) errors-in-variables (EIV) models to its multiple input multiple output (MIMO) case. Similar to the SISO case, the identified model set for a MIMO EIV model is described by an estimated nominal system model and its worst-case error bound. The estimated nominal system model is characterized by a normalized right graph symbol and its worst-case error bound with possibly less conservativeness is quantified by the v-gap metric using a priori and a posteriori information on the EIV model. As a consequence, such model set is well suited to subsequent robust controller design via the H∞ loop-shaping method proposed by Vinnicombe. Finally, the proposed identification method is verified by a numerical simulation example. © 2016, Editorial Department of Control Theory & Applications South China University of Technology. All right reserved.