Open AccessCompetitive co‐evolutionary algorithm for constrained robust design
Author(s) -
Li Min,
Guimarães Frederico,
Lowther David A.
Publication year - 2015
Publication title -
iet science, measurement and technology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.418
H-Index - 49
eISSN - 1751-8830
pISSN - 1751-8822
DOI - 10.1049/iet-smt.2014.0204
Subject(s) - computer science , evolutionary algorithm , algorithm , mathematical optimization , artificial intelligence , mathematics
A competitive co‐evolutionary algorithm is proposed to solve constrained robust design problems. The method uses three populations, among which one population evolves the values of the design variables and the other two evolve the values of two uncertainty variables, accounting for the target and feasibility robustness, respectively. An application to the robust design (minimisation of torque ripples) of an interior permanent magnet machine is presented. Through a worst case analysis, the result from the co‐evolutionary algorithm is proven to have a more robust performance than that of the non‐robust optimisation if manufacturing tolerance is taken into account. The computation time of the co‐evolutionary algorithm may be significantly reduced through the use of parallel computing environments.