Solving multiobjective optimization problems using quasi-separable MDO formulations and analytical target cascading

Cited 13 time in webofscience Cited 0 time in scopus
  • Hit : 461
  • Download : 0
One approach to multiobjective optimization is to define a scalar substitute objective function that aggregates all objectives and solve the resulting aggregate optimization problem (AOP). In this paper, we discern that the objective function in quasi-separable multidisciplinary design optimization (MDO) problems can be viewed as an aggregate objective function (AOF). We consequently show that a method that can solve quasi-separable problems can also be used to obtain Pareto points of associated AOPs. This is useful when AOPs are too hard to solve or when the design engineer does not have access to the models necessary to evaluate all the terms of the AOF. In this case, decomposition-based design optimization methods can be useful to solve the AOP as a quasi-separable MDO problem. Specifically, we use the analytical target cascading methodology to formulate decomposed subproblems of quasi-separable MDO problems and coordinate their solution in order to obtain Pareto points of the associated AOPs. We first illustrate the approach using a well-known simple geometric programming example and then present a vehicle suspension design problem with three objectives related to ground vehicle ride and handling
Publisher
SPRINGER
Issue Date
2014-11
Language
English
Article Type
Article
Keywords

DESIGN OPTIMIZATION; DECOMPOSITION; SYSTEM

Citation

STRUCTURAL AND MULTIDISCIPLINARY OPTIMIZATION, v.50, no.5, pp.849 - 859

ISSN
1615-147X
DOI
10.1007/s00158-014-1144-5
URI
http://hdl.handle.net/10203/212992
Appears in Collection
RIMS Journal Papers
Files in This Item
There are no files associated with this item.
This item is cited by other documents in WoS
⊙ Detail Information in WoSⓡ Click to see webofscience_button
⊙ Cited 13 items in WoS Click to see citing articles in records_button

qr_code

  • mendeley

    citeulike


rss_1.0 rss_2.0 atom_1.0