The implementation of a direct search approach for the resolution of complex and changing rule-based problems

Medland, Anthony John; Matthews, Jason

Abstract

During the evolution of constraint modeling approaches, they have increased in their ability to resolve more and more complex problems. They all rely upon their ability to define the design problem by a set of constraint rules, which are true when the problem is solved, by the manipulation of selected free variables. However, as they have advanced differing techniques, they have been applied to address problems of increasing complexity. This study has been directed toward addressing those that are not only complex but also ill structured and evolving. In order to address such problems, an approach has been developed that employs sensitivity analysis and problem strategies to form an evolving direct search technique. While this is generic approach, which has been applied to a range of engineering problems, it is illustrated here through its use in a study into the posture modeling of humans. In this, it was recognized that such a new approach was required due to the complex description, limits, and postures possible in the human body. © 2009 Springer-Verlag London Limited.

• Journal Article Type: Article
• Publication Date: Apr 1, 2011
• Journal: Engineering with Computers
• Print ISSN: 0177-0667
• Electronic ISSN: 1435-5663
• Publisher: Springer Verlag
• Peer Reviewed: Peer Reviewed
• Volume: 27
• Issue: 2
• Pages: 105-115
• DOI: https://doi.org/10.1007/s00366-009-0148-z
• Keywords: direct search, rule-based problems, constraints
• Public URL: https://uwe-repository.worktribe.com/output/963600
• Publisher URL: http://dx.doi.org/10.1007/s00366-009-0148-z