Evolving Fuzzy Model Predictive Control based on Optimization for Nonlinear Plate Heat Exchanger

In this paper, a novel evolving Fuzzy Model Predictive Control based on Optimization (eFMPC) is proposed. The controller adresses model adaptation in case of abrupt concept shifts in the system parameters. The model of the system is evolved during control based on online learning approaches with a selfmonitoring approach to determine the quality of the local models. The control action is a result of optimization based on the Particle Swarm Optimization method. The control principle was tested on the real Plate Heat Exchanger pilot plant.