TY - GEN
T1 - Tower crane layout planning through Generative Adversarial Network
AU - Li, R.
AU - Chi, H. L.
AU - Peng, Z.
AU - Chen, J.
N1 - Publisher Copyright:
© 2023 the Author(s).
PY - 2023/3
Y1 - 2023/3
N2 - Tower cranes are globally utilized in construction projects to transport components vertically and horizontally, which governs the construction schedule and requires proper locations. However, in practice, the layout of the tower crane is mainly decided by the experience of construction contractors or managers, lacking quality assurance. Generative adversarial network (GAN) is an emerging deep learning technology to generate synthetic images with predictive nature, applied in many research areas, especially in automatic design. Given that, this paper proposed a TC-GAN to identify an appropriate tower crane layout. Information on construction projects was gathered and calculated to obtain a high-quality training dataset considering efficiency and safety. Then, a framework derived from cGAN was applied for the TC-GAN generator and discriminator, training on the massive dataset. The learning rate selection was conducted based on evaluating the quality and rationality of the generated image, which validated the TC-GAN performance in tower crane layout planning.
AB - Tower cranes are globally utilized in construction projects to transport components vertically and horizontally, which governs the construction schedule and requires proper locations. However, in practice, the layout of the tower crane is mainly decided by the experience of construction contractors or managers, lacking quality assurance. Generative adversarial network (GAN) is an emerging deep learning technology to generate synthetic images with predictive nature, applied in many research areas, especially in automatic design. Given that, this paper proposed a TC-GAN to identify an appropriate tower crane layout. Information on construction projects was gathered and calculated to obtain a high-quality training dataset considering efficiency and safety. Then, a framework derived from cGAN was applied for the TC-GAN generator and discriminator, training on the massive dataset. The learning rate selection was conducted based on evaluating the quality and rationality of the generated image, which validated the TC-GAN performance in tower crane layout planning.
UR - http://www.scopus.com/inward/record.url?scp=85160422265&partnerID=8YFLogxK
U2 - 10.1201/9781003354222-49
DO - 10.1201/9781003354222-49
M3 - Conference article published in proceeding or book
AN - SCOPUS:85160422265
SN - 9781032406732
T3 - eWork and eBusiness in Architecture, Engineering and Construction - Proceedings of the 14th European Conference on Product and Process Modelling, ECPPM 2022
SP - 382
EP - 388
BT - eWork and eBusiness in Architecture, Engineering and Construction - Proceedings of the 14th European Conference on Product and Process Modelling, ECPPM 2022
A2 - Hjelseth, Eilif
A2 - Sujan, Sujesh F.
A2 - Scherer, Raimar J.
PB - CRC Press/Balkema
T2 - 14th European Conference on Product and Process Modelling, ECPPM 2022
Y2 - 14 September 2022 through 16 September 2022
ER -