Engine control-oriented fuzzy modeling

This paper exposes a fuzzy model identification method, applied to an engine process. A model structure is defined and an original method is used to specify its different components. This method is tested on a simulated engine process in order to prove the algorithm efficiency and before applying it to a real engine. The results are fuzzy Sugeno's rulebase, with local linear models as conclusion of each rule. Local linear controllers can then be selected with the same fuzzy rule base as for the model.