Predictive functional control with modified Elman neural network for reheated steam temperature

For complex system of reheated steam temperature with large time constant, long time delay and multivariable characteristics, a predictive functional control (PFC) strategy based modified Elman neural network (IOMENN), which consists of input and output context nodes, is designed. PFC based on base function overcomes time delay characteristics to ensure good control performance; IOMENN Elman networks not only act as decouplers and identifier, and also supply base function response for PFC algorithm. Comparing with typical modified Elman neural network, IOMENN structure proposed in the paper improves the dynamic characteristics and converge speed due to employing input and output context nodes. Simulation results prove effectiveness of proposed control strategy.