Model predictive control: a data-driven approach using simple fuzzy tools

Advances in model predictive control using fuzzy tools are presented. Research results are aggregated to present a complete approach based on data-driven fuzzy tools. A fuzzy model of the system is identified from sampled data using supervised fuzzy clustering for rule extraction. This model is used in model predictive control. The non-convex optimization problem introduced by a nonlinear plant model is solved by applying discrete search techniques. The trade-off between computational time and performance that follows from the discretization is addressed by using fuzzy predictive filters. The global fuzzy predictive control approach is applied to a small real-world process.