Authors:
Aki Sorsa
;
Anssi Koskenniemi
and
Kauko Leiviskä
Affiliation:
University of Oulu, Finland
Keyword(s):
Differential Evolution, Identification, Nonlinear Model, Fuel Cell.
Related
Ontology
Subjects/Areas/Topics:
Engineering Applications
;
Evolutionary Computation and Control
;
Informatics in Control, Automation and Robotics
;
Intelligent Control Systems and Optimization
;
Robotics and Automation
;
Signal Processing, Sensors, Systems Modeling and Control
Abstract:
Evolutionary algorithms are optimization methods and their basic idea lies in biological evolution. They suit well for large and complex optimization problems. In this study, differential evolution is applied for identifying the parameters of the nonlinear fuel cell model. Different versions of the algorithm are used to compare the genetic operators they use. One problem with the studied algorithms is also in defining the internal parameters that regulate the development of the population. In this paper, entropy is used for defining the population size and other parameters are tuned using recommendations from the literature and by trial-and-error. The results show that DE/rand-to-best/1/bin is the most suitable algorithm for the studied problem. Selection of the crossover operator has no considerable effect on the results. The results also show that the studied identification problem has a lot of local minima that are very close to each other that makes the optimization problem even
more challenging.
(More)