0

Full Content is available to subscribers

Subscribe/Learn More  >

Support Optimization for Piping System With Machine Learning

[+] Author Affiliations
Jongho Ham, Jungeun An, Bongjae Kim, Jaewoong Choi, Booki Kim

Samsung Heavy Industries, Seongnam-si, Korea Republic

Paper No. OMAE2017-62356, pp. V07AT06A043; 10 pages
doi:10.1115/OMAE2017-62356
From:
  • ASME 2017 36th International Conference on Ocean, Offshore and Arctic Engineering
  • Volume 7A: Ocean Engineering
  • Trondheim, Norway, June 25–30, 2017
  • Conference Sponsors: Ocean, Offshore and Arctic Engineering Division
  • ISBN: 978-0-7918-5773-1
  • Copyright © 2017 by ASME

abstract

Piping stress analysis is performed by the manipulations of support type, location and pipe arrangement based on many specific design criteria. A classical way to find good engineering solutions satisfying design criteria among lots of combinations is obviously time-consuming work. In field practice, it also highly depends on engineer’s experiences and abilities.

This paper proposes a hybrid method by combining several global search optimization algorithms and predication model generation in order to automatically control the combinations of support types as the engineering solutions. Here, we use some efficient and popular algorithms such as genetic algorithm, swarm intelligence and Gaussian pattern search to develop initial design of experiments. From the set of the initials, we build and update a prediction model by applying machine learning algorithm such as artificial neural network. As a result of using the hybrid method, the engineering solution is sufficiently optimized for the classical solution.

Design variables for this problem are the types of restraints (or the pipe support type). The nonlinearity conditions such as gaps and frictions are also treated as key design variables. Each restraint is initially identified as a binary set of design variables, and transformed to integer numbers to run on the n-dimensional design space. The number of dimension corresponds to the number of pipe supports. Currently, pipe stress analysis problems are divided into a certain size that is enough to run on one computer for project management purpose. If we have bigger system with more design variables to consider, the hybrid machine learning method plays a key role in saving computation time with the help of additional parallel computation technique.

Copyright © 2017 by ASME

Figures

Tables

Interactive Graphics

Video

Country-Specific Mortality and Growth Failure in Infancy and Yound Children and Association With Material Stature

Use interactive graphics and maps to view and sort country-specific infant and early dhildhood mortality and growth failure data and their association with maternal

NOTE:
Citing articles are presented as examples only. In non-demo SCM6 implementation, integration with CrossRef’s "Cited By" API will populate this tab (http://www.crossref.org/citedby.html).

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In