Adaptive neurofuzzy predictive control of nuclear steam generators

The main emphasis of this paper is the application of adaptive neurofuzzy model-based predictive control, to regulate the water level in the U-tube steam generating (UTSG) unit used for electricity generation. A nonlinear predictive controller is designed on the basis of a Takagi-Sugeno fuzzy model with B-spline membership function. By on-line adaptation of the neurofuzzy model, improvement of the control performance can be achieved with the time-variant process behaviour. For this purpose, a normalized least-square algorithm is utilized which exploits the local linearity of Takagi-Sugeno fuzzy models. An optimization approach with a quadratic programing technique is used to calculate predictions of the future control actions. The effectiveness and real-world applicability of the proposed approach are demonstrated by computer simulation. The control experiments were successfully conducted for this nonlinear process with satisfactory results and performances.