Nonlinear Modeling and Control of Full Recuperative MGT-CHP System Based on Hammerstein Model

This paper studies the simulation modeling and control of the full recuperative microturbine-based combined heat and power (MGT-CHP) system. Aiming at the fact that physical model is too complicated to reproduce and the simulation is very time-consuming, an easily-reproduced block-oriented model is proposed in order to design the control strategy. The multivariate Hammerstein model of the MGT-CHP system is obtained by the model identification method based on recursive least squares (RLS). Based on the Hammerstein predictive model, a nonlinear generalized predictive controller (H-GPC) is designed for tracking the heat and electrical loads of the MGT-CHP system. The simulation results show that the Hammerstein model fits well with the physical model, and the proposed nonlinear generalized predictive controller has better performance than the traditional PID controller with optimal parameters.