Evaluating memetic building spatial design optimisation using hypervolume indicator gradient ascent

K. van der Blom, S. Boonstra, H. Wang, H. Hofmeyer, Michael T.M. Emmerich

Onderzoeksoutput: Hoofdstuk in Boek/Rapport/CongresprocedureConferentiebijdrageAcademicpeer review

1 Citaat (Scopus)
3 Downloads (Pure)

Uittreksel

In traditional, single objective, optimisation local optima may be found by gradient search. With the recently introduced hypervolume indicator (HVI) gradient search, this is now also possible for multi-objective optimisation, by steering the whole Pareto front approximation (PFA) in the direction of maximal improvement. However, so far it has only been evaluated on simple test problems. In this work the HVI gradient is used for the real world problem of building spatial design, where the shape and layout of a building are optimised. This real world problem comes with a number of constraints that may hamper the effectiveness of the HVI gradient. Specifically, box constraints, and an equality constraint which is satisfied by rescaling. Moreover, like with regular gradient search, the HVI gradient may overstep an optimum. Therefore, step size control is also investigated. Since the building spatial designs are encoded in mixed-integer form, the use of gradient search alone is not sufficient. To navigate both discrete and continuous space, an evolutionary multi-objective algorithm (EMOA) and the HVI gradient are used in hybrid, forming a so-called memetic algorithm. Finally, the effectiveness of the memetic algorithm using the HVI gradient is evaluated empirically, by comparing it to an EMOA without a local search method. It is found that the HVI gradient method is effective in improving the PFA for this real world problem. However, due to the many discrete subspaces, the EMOA is able to find better solutions than the memetic approach, albeit only marginally.
Originele taal-2Engels
TitelNumerical and Evolutionary Optimization – NEO 2017
RedacteurenOliver Schutze, Leonardo Trujillo, Yazmin Maldonado, Paul Valle
Plaats van productieCham
UitgeverijSpringer
Pagina's62-86
Aantal pagina's25
ISBN van geprinte versie978-3-319-96103-3
DOI's
StatusGepubliceerd - 2019

Publicatie series

NaamStudies in Computational Intelligence
Volume785
ISSN van geprinte versie1860-949X

Vingerafdruk

Gradient methods
Multiobjective optimization
Design optimization

Citeer dit

van der Blom, K., Boonstra, S., Wang, H., Hofmeyer, H., & Emmerich, M. T. M. (2019). Evaluating memetic building spatial design optimisation using hypervolume indicator gradient ascent. In O. Schutze, L. Trujillo, Y. Maldonado, & P. Valle (editors), Numerical and Evolutionary Optimization – NEO 2017 (blz. 62-86). (Studies in Computational Intelligence; Vol. 785). Cham: Springer. https://doi.org/10.1007/978-3-319-96104-0_3
van der Blom, K. ; Boonstra, S. ; Wang, H. ; Hofmeyer, H. ; Emmerich, Michael T.M. / Evaluating memetic building spatial design optimisation using hypervolume indicator gradient ascent. Numerical and Evolutionary Optimization – NEO 2017. redacteur / Oliver Schutze ; Leonardo Trujillo ; Yazmin Maldonado ; Paul Valle. Cham : Springer, 2019. blz. 62-86 (Studies in Computational Intelligence).
@inproceedings{773c095f3ada45f098f8ce2d56168b9d,
title = "Evaluating memetic building spatial design optimisation using hypervolume indicator gradient ascent",
abstract = "In traditional, single objective, optimisation local optima may be found by gradient search. With the recently introduced hypervolume indicator (HVI) gradient search, this is now also possible for multi-objective optimisation, by steering the whole Pareto front approximation (PFA) in the direction of maximal improvement. However, so far it has only been evaluated on simple test problems. In this work the HVI gradient is used for the real world problem of building spatial design, where the shape and layout of a building are optimised. This real world problem comes with a number of constraints that may hamper the effectiveness of the HVI gradient. Specifically, box constraints, and an equality constraint which is satisfied by rescaling. Moreover, like with regular gradient search, the HVI gradient may overstep an optimum. Therefore, step size control is also investigated. Since the building spatial designs are encoded in mixed-integer form, the use of gradient search alone is not sufficient. To navigate both discrete and continuous space, an evolutionary multi-objective algorithm (EMOA) and the HVI gradient are used in hybrid, forming a so-called memetic algorithm. Finally, the effectiveness of the memetic algorithm using the HVI gradient is evaluated empirically, by comparing it to an EMOA without a local search method. It is found that the HVI gradient method is effective in improving the PFA for this real world problem. However, due to the many discrete subspaces, the EMOA is able to find better solutions than the memetic approach, albeit only marginally.",
keywords = "Building spatial design, Hypervolume indicator gradient, Memetic algorithm, Multi-objective optimisation",
author = "{van der Blom}, K. and S. Boonstra and H. Wang and H. Hofmeyer and Emmerich, {Michael T.M.}",
year = "2019",
doi = "10.1007/978-3-319-96104-0_3",
language = "English",
isbn = "978-3-319-96103-3",
series = "Studies in Computational Intelligence",
publisher = "Springer",
pages = "62--86",
editor = "Oliver Schutze and Leonardo Trujillo and Yazmin Maldonado and Paul Valle",
booktitle = "Numerical and Evolutionary Optimization – NEO 2017",
address = "Germany",

}

van der Blom, K, Boonstra, S, Wang, H, Hofmeyer, H & Emmerich, MTM 2019, Evaluating memetic building spatial design optimisation using hypervolume indicator gradient ascent. in O Schutze, L Trujillo, Y Maldonado & P Valle (redactie), Numerical and Evolutionary Optimization – NEO 2017. Studies in Computational Intelligence, vol. 785, Springer, Cham, blz. 62-86. https://doi.org/10.1007/978-3-319-96104-0_3

Evaluating memetic building spatial design optimisation using hypervolume indicator gradient ascent. / van der Blom, K.; Boonstra, S.; Wang, H.; Hofmeyer, H.; Emmerich, Michael T.M.

Numerical and Evolutionary Optimization – NEO 2017. redactie / Oliver Schutze; Leonardo Trujillo; Yazmin Maldonado; Paul Valle. Cham : Springer, 2019. blz. 62-86 (Studies in Computational Intelligence; Vol. 785).

Onderzoeksoutput: Hoofdstuk in Boek/Rapport/CongresprocedureConferentiebijdrageAcademicpeer review

TY - GEN

T1 - Evaluating memetic building spatial design optimisation using hypervolume indicator gradient ascent

AU - van der Blom, K.

AU - Boonstra, S.

AU - Wang, H.

AU - Hofmeyer, H.

AU - Emmerich, Michael T.M.

PY - 2019

Y1 - 2019

N2 - In traditional, single objective, optimisation local optima may be found by gradient search. With the recently introduced hypervolume indicator (HVI) gradient search, this is now also possible for multi-objective optimisation, by steering the whole Pareto front approximation (PFA) in the direction of maximal improvement. However, so far it has only been evaluated on simple test problems. In this work the HVI gradient is used for the real world problem of building spatial design, where the shape and layout of a building are optimised. This real world problem comes with a number of constraints that may hamper the effectiveness of the HVI gradient. Specifically, box constraints, and an equality constraint which is satisfied by rescaling. Moreover, like with regular gradient search, the HVI gradient may overstep an optimum. Therefore, step size control is also investigated. Since the building spatial designs are encoded in mixed-integer form, the use of gradient search alone is not sufficient. To navigate both discrete and continuous space, an evolutionary multi-objective algorithm (EMOA) and the HVI gradient are used in hybrid, forming a so-called memetic algorithm. Finally, the effectiveness of the memetic algorithm using the HVI gradient is evaluated empirically, by comparing it to an EMOA without a local search method. It is found that the HVI gradient method is effective in improving the PFA for this real world problem. However, due to the many discrete subspaces, the EMOA is able to find better solutions than the memetic approach, albeit only marginally.

AB - In traditional, single objective, optimisation local optima may be found by gradient search. With the recently introduced hypervolume indicator (HVI) gradient search, this is now also possible for multi-objective optimisation, by steering the whole Pareto front approximation (PFA) in the direction of maximal improvement. However, so far it has only been evaluated on simple test problems. In this work the HVI gradient is used for the real world problem of building spatial design, where the shape and layout of a building are optimised. This real world problem comes with a number of constraints that may hamper the effectiveness of the HVI gradient. Specifically, box constraints, and an equality constraint which is satisfied by rescaling. Moreover, like with regular gradient search, the HVI gradient may overstep an optimum. Therefore, step size control is also investigated. Since the building spatial designs are encoded in mixed-integer form, the use of gradient search alone is not sufficient. To navigate both discrete and continuous space, an evolutionary multi-objective algorithm (EMOA) and the HVI gradient are used in hybrid, forming a so-called memetic algorithm. Finally, the effectiveness of the memetic algorithm using the HVI gradient is evaluated empirically, by comparing it to an EMOA without a local search method. It is found that the HVI gradient method is effective in improving the PFA for this real world problem. However, due to the many discrete subspaces, the EMOA is able to find better solutions than the memetic approach, albeit only marginally.

KW - Building spatial design

KW - Hypervolume indicator gradient

KW - Memetic algorithm

KW - Multi-objective optimisation

UR - http://www.scopus.com/inward/record.url?scp=85050355768&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-96104-0_3

DO - 10.1007/978-3-319-96104-0_3

M3 - Conference contribution

SN - 978-3-319-96103-3

T3 - Studies in Computational Intelligence

SP - 62

EP - 86

BT - Numerical and Evolutionary Optimization – NEO 2017

A2 - Schutze, Oliver

A2 - Trujillo, Leonardo

A2 - Maldonado, Yazmin

A2 - Valle, Paul

PB - Springer

CY - Cham

ER -

van der Blom K, Boonstra S, Wang H, Hofmeyer H, Emmerich MTM. Evaluating memetic building spatial design optimisation using hypervolume indicator gradient ascent. In Schutze O, Trujillo L, Maldonado Y, Valle P, redacteurs, Numerical and Evolutionary Optimization – NEO 2017. Cham: Springer. 2019. blz. 62-86. (Studies in Computational Intelligence). https://doi.org/10.1007/978-3-319-96104-0_3