An augmented Lagrangian approach to non-convex SAO using diagonal quadratic approximations

A.A. Groenwold, L.F.P. Etman, S. Kok, D.W. Wood, S. Tosserams

    Onderzoeksoutput: Bijdrage aan tijdschriftTijdschriftartikelAcademicpeer review

    7 Citaten (Scopus)

    Samenvatting

    Successful gradient-based sequential approximate optimization (SAO) algorithms in simulation-based optimization typically use convex separable approximations. Convex approximations may however not be very efficient if the true objective function and/or the constraints are concave. Using diagonal quadratic approximations, we show that non-convex approximations may indeed require significantly fewer iterations than their convex counterparts. The nonconvex subproblems are solved using an augmented Lagragian (AL) strategy, rather than the Falk-dual, which is the norm in SAO based on convex subproblems.
    Originele taal-2Engels
    Pagina's (van-tot)415-421
    TijdschriftStructural and Multidisciplinary Optimization
    Volume38
    Nummer van het tijdschrift4
    DOI's
    StatusGepubliceerd - 2009

    Vingerafdruk

    Duik in de onderzoeksthema's van 'An augmented Lagrangian approach to non-convex SAO using diagonal quadratic approximations'. Samen vormen ze een unieke vingerafdruk.

    Citeer dit