Intelligent Engineering Systems through Artificial Neural Networks, Volume 20
27 Parameter Estimation of the Duffing Oscillator Using Poincaré Map and an Elitist Genetic Algorithm
Download citation file:
This paper deals with unsupervised genetic algorithm based clustering in multidimensional parameter spaces for the non-linear Duffing dynamic system. Numerical integration of the governing differential equations provides the basis for the development of Poincaré maps which is used for automatic feature extraction and data clustering in the parameter space to distinguish regions of periodic behavior from chaotic regions. An elitist genetic algorithm is used to search the parameter space to optimize a combination of oscillator parameters.