Application of Adaptive Local Linear Model Tree for Nonlinear Identification of Heat Recovery Steam Generator System Based on Experimental Data

In this paper, the Local Linear Model Tree (LOLIMOT) founded on Takagi-Sugeno-Kang fuzzy notion is employed for model estimation of a nonlinear HRSG (Heat Recovery Steam Generator) process. This method involves a heuristic search to choose the input partitions space by axis-orthogonal splits. The aim of this work is to enhance accuracy of the dynamic model without increasing its complexity. The boiler model presented here displays all the crucial features of the actual boiler dynamics, including nonlinearities, nonminimum-phase behavior, instabilities, noise spectrum in the same frequency range as significant plant dynamics, time delays, and load disturbances. The results show that the LOLIMOT gives the smallest error on the unseen data. On the other hand, the limited flexibility of the local estimation reduces the variance error due to the bias/variance dilemma.