“…This algorithm may use lookup tables [1], radial basis functions (RBF) [2], artificial neural networks (ANN) [3,4] and its derivations such as fuzzy logic (FL) [5] and neural-fuzzy network (NF) [6], and regression [7,8]. Model generators can also be categorized into the black, grey and white box approaches, depending on the level of existing knowledge of the system's structure and parameters Unfortunately AMG may produce high order models of excessive complexity [9 -11], in which case model order reduction (MOR) techniques may be required [12]. A recent paper [13] discusses MOR techniques with respect to various model types, and in a variety of contexts: LTI MOR [14], LTV MOR [15,16], weakly nonlinear methods including polynomial-based [17,18], trajectory piecewise linear (TPWL) [19], and piecewise polynomial (PWP) [20].…”