Fuzzy modeling of nonlinear uncertain dynamic systems

Person in charge

M.Sc. Salman Zaidi


October 2011 - December 2015


Brief description

In modeling of real world physical systems from input-output data, some system dynamics, inherent stochasticity or external disturbances may remain unmodeled causing uncertainty on how close the model is to the “true system”. Dependent on the characteristics of the uncertainties, this information could be max/min boundaries of the output or statistical information on its spreading.

Conventional fuzzy systems do not possess this capability. In contrast, a type-2 FLS (T2 FLS) can handle higher levels of uncertainties. However the interpretation, reliability and accuracy of their ability to provide additional information in the form of type reduced set are rarely discussed in the literature. Moreover, how the uncertainty in data for identification can be translated into Type-2 fuzzy sets is still an open research question. In this research project, data driven modeling approach for nonlinear dynamic systems with uncertainties will be developed and tested for experimental test benches e.g., eletro-mechanical and servo-pneumatic actuators with friction.