Genetic algorithm with peaks adaptive objective function used to fit the EPR powder spectrum

Authors:
Sebastian Grzegorz Żurek
Affiliations:
Institute of Physics, University of Zielona Góra, ul. Szafrana 4a, 65-516 Zielona Góra, Poland
Venue:
Applied Soft Computing
Year:
2011

Citing 5
Cited 1

The GENITOR algorithm and selection pressure: why rank-based allocation of reproductive trials is best

Proceedings of the third international conference on Genetic algorithms
Genetic Algorithms in Search, Optimization and Machine Learning

Genetic Algorithms in Search, Optimization and Machine Learning
Uniform Crossover in Genetic Algorithms

Proceedings of the 3rd International Conference on Genetic Algorithms
Predictive models for the breeder genetic algorithm i. continuous parameter optimization

Evolutionary Computation
SCE toolboxes for the development of high-level parallel applications

ICCS'06 Proceedings of the 6th international conference on Computational Science - Volume Part II

Application of genetic algorithm for unknown parameter estimations in cylindrical fin

Applied Soft Computing

Quantified Score

Hi-index	0.00

Visualization

Abstract

An interpretation of the EPR spectra of powder samples is generally accomplished after the determination of a set of spin-Hamiltonian parameters and the line-shape of the paramagnetic object. Various local optimization techniques have been used to match the mathematical model to the experimental data-set and as a rule, the result depends on the proper choice of the starting parameters entered by the spectroscopist. The method of genetic algorithms, modified with peaks adaptive objective function is shown to successfully perform the global optimization search process. Furthermore, the right definition of the genotype as well as the modification of the objective function, which should account for the EPR data peaks adaptation, is shown to dramatically improve the search in the multidimensional parameter space. To convert the genotype data to the phenotype representation the EPR simulation tool, Easyspin Matlab Toolbox, is used. To speed up the computation the LAM/MPI is implemented to set up a small computer cluster, with Javier Fernandez Baldomero's MPI Toolbox serving as a Matlab interface. To apply the genetic algorithm routines the Hartmut Pohlheim's GEATbx Matlab Toolbox is used. If a proper genotype - phenotype conversion is implemented, the method described below is expected to be useful in analysing a wide variety of spectroscopy data.