Optimization techniques with knowledge based control in ship concept design

This thesis was submitted for the degree of Doctor of Philosophy and awarded by Brunel University.An integrated computational approach to Ship Concept Design using optimization techniques and a knowledge base to control the optimization process has been developed. The system automates both synthesis and analysis; analysis by the repeated sequential use of Design Theory Modules and synthesis through the optimization process, which compromises conflicting requirements, subject to constraints. The intention of this work has been to find a better approach to automated design synthesis and at the same time employ detailed analytical tools such as a three-dimensional hull-form definition and engineering analysis modules. Optimization techniques and a knowledge base are combined to achieve the desired capabilities, taking advantage of the benefits optimization can bring using goal oriented methods and exploratory searches, alongside a knowledge base that controls the synthesis process rather than the design. A function mapping strategy has been developed to provide a multiple-parametric view of regions of the optimization objective function and constraints. A discussion is included on the role of further applications of expert systems to design systems in both synthesis and analysis and their possible interference with creativity and innovation. Two design examples are provided, one showing the application of the system using optimization and the other adding the use of the knowledge base. The results are compared and discussed.National Council of Research (CNPq - Brazil