Fuzzy model-based Smith predictive control and its application in superheated steam temperature control

This paper proposes a Smith predictive controller based on fuzzy dynamic model. Smith predictor is very sensitive to the model error For overcoming parameter variations of the controlled object at every different load, different from adaptive control, this paper uses one constant Smith predictive model all long to construct mismatched Smith predictor fuzzy dynamic model is applied to compensate model error This procedure includes two steps: first, construct reduced models through GA according to step response of plant at several different loads; second, aim at these reduced models, establish fuzzy dynamic model to compensate the model mismatch. Simulation results show that the method proposed in this paper is feasible and effective for a class of large time delay, time-varying complex processes.