Efficient predictive control algorithm based on fuzzy Hammerstein models: A case study

An efficient fuzzy predictive control algorithm based on Hammerstein models is proposed in the paper. The algorithm uses the DMC (Dynamic Matrix Control) technique and a Hammerstein model, in which fuzzy static block precedes a linear dynamic block. The static block may be identified easily using, e.g. heuristic approach and/or fuzzy neural networks. The dynamic part of the model has the form of control plant step responses. The proposed algorithm is little complicated and numerically effective (the main part of calculations is performed off-line) but it offers better control performance than a classical algorithm (based on a linear model). It is demonstrated in the example control system of a nonlinear control plant with significant delay. © Springer-Verlag Berlin Heidelberg 2010.