Investigating the surrogate worth trade-off method to facilitate technology selection for new systems

Aria Iwasawa, Naohiko Kohtake, Nobuaki Minato, William Crossley

Research output: Contribution to journalArticlepeer-review

6 Citations (Scopus)


When designing a new system, engineers must often select from a set of discrete technologies available for use. Once the engineers select these technologies, they subsequently determine the value of continuous variables (e.g., lengths, thicknesses, other dimensions) that describe the new system. This mix of discrete and continuous choices can make it difficult to identify the best design. Additional difficulties arise when—as for most new systems—a trade-off exists between multiple, competing objectives. From a design optimization perspective, the resulting problem is a Multi-Objective, Mixed-Discrete Non-Linear Programming (MO-MDNLP) problem. The solution to an MO-MDNLP problem is not a single design; it is a set of non-dominated designs. In this set, the performance of one objective cannot improve without degrading performance in the other objective(s). However, the design process requires that a single design emerge as the best candidate; this best design needs to reflect the decision-maker's preferences. The Surrogate Worth Trade-off (SWT) method is one approach that provides an interface between the decision-maker's preferences and the mathematical models. This paper applied the SWT method to a simple example MO-MDNLP problem to determine how this might support a decision-maker in selecting new technologies during the early phases of design.

Original languageEnglish
Pages (from-to)339-349
Number of pages11
JournalTopics in Safety, Risk, Reliability and Quality
Publication statusPublished - 2014


  • Multi-objective design
  • Technology selection
  • The surrogate worth trade-off method

ASJC Scopus subject areas

  • Safety, Risk, Reliability and Quality


Dive into the research topics of 'Investigating the surrogate worth trade-off method to facilitate technology selection for new systems'. Together they form a unique fingerprint.

Cite this