0000000000522859
AUTHOR
Stelian Ciurea
Tuning a Mamdani Fuzzy Controller with an Imperialist Competitive Algorithm
We have implemented a fuzzy controller with a view to regulating a single-input and single-output second-order linear system. The fuzzy controller was a Mamdami proportional-derivative controller. To determine the parameters of the fuzzy controller we have used an imperialist competitive algorithm. This type of algorithm has a long running time so we implemented also a parallel version of the algorithm that we run on HPC Zamolxes located at the Engineering Faculty of “Lucian Blaga” University from Sibiu. Because we did not have on this computer a version of MATLAB allowing to write parallel algorithms, we implemented the entire application in the C language using the MPI library.
Imperialist competitive algorithm for determining the parameters of a Sugeno fuzzy controller
Abstract We used an imperialist competitive algorithm to determine the parameters of a fuzzy controller of type Sugeno that would ensure a good unit step response of a second-order single-input and single-output automatic system.