A New Recurrent Radial Basis Function Network-based Model Predictive Control for a Power Plant Boiler Temperature Control

In this paper, a new radial basis function network-based model predictive control (RBFN-MPC) is presented to the steam temperature of power plant boiler. For first time in paper Laguerre polynomials are used obtain local boiler models based on different load modes. Recursive least square (RLS) method as observer coefficient. Then locally recurrent neural network with self-organizing mechanism these transfer and it estimate future behavior. The RBFN tracks system dynamic online updates model. RBFN, output hidden layer nodes at past moment modelling, So behaves exactly like real Various uncertainties have been added immediately recognized by RBFN. simulation, proposed has compared traditional MPC (based mathematical model). Simulation results showed that RBFN-based perform better than model-based MPC. This due network's tracking dynamics, while way always constant. As amount uncertainty increases, difference between our existing methods can clearly be observed.