Internal Model Control based on Parallel Self-learning Neural Network for Superheated Steam Temperature System

As to the superheated steam temperature control system has large time constant, long time-delay and time-varying in thermal power plant, a control strategy of internal model control based on parallel self-learning neural network is presented in this paper, which recurs to the identification ability for complex nonlinear of the neural network to identify the model and inverse model of the object. It divides the control system into two processes: control process and parallel self-learning process. Control process realizes the function of the internal model control, which includes the NNM, NNC and a feedback robust stable controller (RC). The parallel self-learning process is used to train the NNC and then its weights are copied to control process online. Simulation results show that this strategy has perfect control performances, strong robustness and self-adaptive ability