MIMO closed-loop identification of an MSW incinerator

In this paper, the application of a specific system identification procedure to a municipal solid waste (MSW) incinerator is discussed. This procedure is a combination of, on the one hand, a particular closed-loop identification method called the two-stage method and, on the other hand, the approach of high-order multiple input multiple output (MIMO) ARX model estimation followed by model reduction. MIMO ARX model estimation is performed by means of a, so-called, multiple data set identification method, i.e. a method by means of which it is possible to estimate a model on the basis of several data sets instead of just one data set. Model reduction is applied to each transfer function of the resulting MIMO ARX model separately. It is shown that with the proposed identification procedure a model of the MSW incinerator is obtained which, according to system identification validation measures, is good. Using the estimated model, the influence of the disturbances on the identification and control of an MSW incinerator is discussed. Furthermore, the validation of a first-principles model of the MSW incineration process by means of the resulting low-order SISO models is discussed. The results show that the proposed way of validating a first-principles model is a powerful tool for determining its quality. r 2002 Elsevier Science Ltd. All rights reserved.