Using Metaheuristic Computations to Find the Minimum-Norm-Residual Solution to Linear Systems of Equations

Journal article

Authors/Editors

Jamisola, Rodrigo

Research Areas

ENGINEERING AND TECHNOLOGY (Frascati classification)

Publication Details

Subtitle: Using Metaheuristic Computations to Find the Minimum-Norm-Residual Solution to Linear Systems of Equations

Author list: Jamisola, Rodrigo

Publisher: IGI Global

Publication year: 2009

Journal: International Journal of Applied Metaheuristic Computing (1947-8283)

Start page: 1

End page: 9

Number of pages: 9

ISSN: 1947-8283

eISSN: 1947-8291

URL: file:///C:/Users/kethmoloi/Downloads/2009_PCJ09_UsingMetaheuristic.pdf

Languages: English-United States (EN-US)

Abstract

This work will present metaheuristic computations,namely, probabilistic artificial neural network, simulated annealing,and modified genetic algorithm in finding the minimumnorm-residualsolution to linear systems of equations. By demonstratinga set of input parameters, the objective function, and theexpected results solutions are computed for determined, overdetermined,and underdetermined linear systems. In addition, thiswork will present a version of genetic algorithm modified interms of reproduction and mutation. In this modification, everyreproduction cycle is performed by matching each individual withthe rest of the individuals in the population. Further, the offspringchromosomes result from crossover of parent chromosomeswithout mutation. The selection process only selects the best fitindividuals in the population. Mutation is only performed whenthe desired level of fitness cannot be achieved, and all the possiblechromosome combinations were already exhausted. Experimentalresults for randorrly generated matrices with increasing matrixsizes will be presented and analyzed. It will be the basis inmodeling and identifying the dynamics parameters of a humanoidrobot through response optimization at excitatory motions.

Keywords

No matching items found.

Documents