XML Persian Abstract Print

Abstract:   (1806 Views)

A nonlinear model predictive control (NMPC) algorithm based on neural network is designed for boiler- turbine system. The boiler–turbine system presents a challenging control problem owing to its severe nonlinearity over a wide operation range, tight operating constraints on control move and strong coupling among variables. The nonlinear system is identified by MLP neural network and neural predictive control is designed by MIMO neural network model. Using neural network as predictive model in predictive control results in steady state error. A disturbance model is used in neural predictive model for elimination of steady state error and disturbance rejection. Simulation results on a boiler turbine system illustrate that a satisfactory closed-loop performance and offset-free property can be achieved by using the proposed method.

Full-Text [PDF 1168 kb]   (1580 Downloads)    
Type of Article: Research | Subject: Power
Received: 2017/02/1 | Accepted: 2017/02/1 | Published: 2017/02/1