.png){kind=icon}
.png){kind=icon}
Return
Digital twin-based online structural optimization? Yes, it's possible!
.png){kind=icon}
PRE
AI
DOI：10.1016/j.tws.2024.112796.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
In structural design, simulation technology has been extensively applied. However, simulation-based offline optimization methods encounter two primary limitations: (1) they typically optimize parameters based on a structure's mechanical properties under extreme, static conditions, and (2) they involve complex, timeconsuming performance calculations. To address these challenges, this paper introduces a structural online optimization framework driven by shape-performance integrated digital twins (SPI-DTs), which breaks the connection barrier between the real-time dynamics of digital twins and the offline characteristics of traditional optimization under dynamic operating conditions. Initially, the structure's remaining life is calculated online using real-time performance data from the digital twin, with the constraint stress and equivalent load derived via the static load equivalent method. A hierarchical radial basis function is then proposed, which integrates multi- source data for rapid calculation of the structure's target performance under the equivalent load. Finally, structural optimization parameters, determined through topology optimization, are solved using the multi- objective genetic algorithm. The feasibility of the proposed method is demonstrated through typical case studies: the wing cantilever beam and a three-dimensional flat plate. The results show that the weight of the wing cantilever beam case based on size optimization is reduced by 12.07 %, and the weight of the flat plate case based on combined topology optimization and size optimization is reduced by 66.66 %. In conclusion, the proposed framework offers a viable approach to achieving lightweight design and extending the operational life of structures under real-world conditions.
Keywords:
Digital twin
Online optimization
Fatigue analysis
Predictive maintenance
Journal
T
IF:
6.6
Papers:
1.1W
Citations:
4.0W