Study of an algorithm of RBF-GPC for CFBB Unit

Circulating Fluidized Bed Boiler (CFBB) is a multivariable, strong-coupling, long-time delay, nonlinear and time-variable systems. The accurate model is difficult to establish. So it is hard to get optimum effect with conventional control strategy. In this study, RBF network is used to identify the CFBB control system by establishing a predictive model in generalized predictive control strategy, and achieve predictive control with online rolling optimization and real time feedback revision. The simulation result shows that it can adapt the object change characteristic. This method has a good control effect and the strong robustness.