ABSTRACT

While there are many possible approaches to the problem briefly described in the introduction, it is believed that a Bayesian decision framework provides the most appropriate platform from which a unified set of techniques can be developed. The motivation for using this framework is that prior knowledge can be used in a systematic manner in lieu of full-scale measurements and that decision-making can proceed in a manner that accounts for high uncertainties and costs of failure. The complexity of a structural system, however, and the resulting high-dimensional design space makes a direct application of Bayesian decision theory to structural design a significant computational challenge. Some details on approaching the computational and information research issues in developing a Bayesian decision framework are given in the following.