MIMO ARX-SDP model estimation proposal and application to a hydraulic system

This work deals with the ARX-SDP model estimation problem; it implies on discovering the state-parameter dependences. This model is always non-linear and cannot be solved by conventional methods. Young's proposal, based on a special data-reordering and on an iterative-recursive method, is modified considering a fixed and low forgetting factor and unitary regressors. A MIMO version of our proposal is also presented. A hydraulic application to four interconnected tanks shows the applicability of our proposal.