TY - JOUR
T1 - Lego-inspired reconfigurable modular blocks for automated construction of engineering structures
AU - Barhemat, Rojyar
AU - Mahjoubi, Soroush
AU - Li, Victor C.
AU - Bao, Yi
N1 - Publisher Copyright:
© 2022 Elsevier B.V.
PY - 2022/7
Y1 - 2022/7
N2 - Reconfigurable modular structures are able to be assembled using prefabricated modules and reconfigured to promote automated construction and to improve sustainability and resilience of infrastructure, while the computer-aided design and modeling of the modules are unclear. This study develops a many-objective optimization approach to design the modules made using strain-hardening cementitious composite. The proposed approach integrates a sequential surrogate model, Latin hypercube sampling method, Unified Non-dominated Sorting Genetic Algorithm III, and Technique for Order of Preference by Similarity to Ideal Solution to predict and optimize the properties of assemblages of the modules. Four objective functions were defined using the load-carrying capacity, deformability, stiffness, and volume. Results showed that the proposed method had reasonable prediction accuracy. The optimal design increased the load-carrying capacity, deformability, and stiffness by 22.8%, 11.5%, and 129.2%, respectively, and reduced the volume by 51.6%. This study is expected to effectively improve the design of reconfigurable modular structures.
AB - Reconfigurable modular structures are able to be assembled using prefabricated modules and reconfigured to promote automated construction and to improve sustainability and resilience of infrastructure, while the computer-aided design and modeling of the modules are unclear. This study develops a many-objective optimization approach to design the modules made using strain-hardening cementitious composite. The proposed approach integrates a sequential surrogate model, Latin hypercube sampling method, Unified Non-dominated Sorting Genetic Algorithm III, and Technique for Order of Preference by Similarity to Ideal Solution to predict and optimize the properties of assemblages of the modules. Four objective functions were defined using the load-carrying capacity, deformability, stiffness, and volume. Results showed that the proposed method had reasonable prediction accuracy. The optimal design increased the load-carrying capacity, deformability, and stiffness by 22.8%, 11.5%, and 129.2%, respectively, and reduced the volume by 51.6%. This study is expected to effectively improve the design of reconfigurable modular structures.
KW - Kriging model
KW - Many-objective optimization
KW - Modular blocks
KW - Reconfigurable modular structures sequential surrogate modeling
KW - Unified non-dominated sorting genetic algorithm III
UR - http://www.scopus.com/inward/record.url?scp=85129972924&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85129972924&partnerID=8YFLogxK
U2 - 10.1016/j.autcon.2022.104323
DO - 10.1016/j.autcon.2022.104323
M3 - Article
AN - SCOPUS:85129972924
SN - 0926-5805
VL - 139
JO - Automation in Construction
JF - Automation in Construction
M1 - 104323
ER -