Closed loop identification of MIMO Hammerstein models using LS-SVM

Abstract

In this paper we present an algorithm to identify MIMO Hammerstein systems under open and closed-loop conditions. To do so, we formulate the optimized predictor based subspace identification algorithm in the dual space. In this dual space we utilize ideas from support vector machines to estimate the state sequence. With the state sequence known, we use the same machinery to estimate the system matrices and the static nonlinearity. The effectiveness of the approach is illustrated with a closed-loop simulation example.