Modelling and Exploring Hierarchical Systems Engineering Models with a Preliminary Ship Design Focus

Abstract

In contributing to the body of knowledge for decision-based design (DBD), the work reported in this paper highlights a work in progress at the Australian Defence Force Academy (ADFA). The notion of a system being represented by a hierarchical set of sub-system design problems is not new. However, the ability to deal with the system's sub-problems generically and direct them to different solvers as appropriate within a design environment has not been well explored. It follows that with an interest in genetic algorithms, our current research objective may be summarised as an effort to answer the following question. Given that decision support and design optimisation techniques are generally applicable in design, can better real world decisions be modelled and explored using a hybrid recursive solver, a method employing (for demonstration purposes) sequential linear and genetic algorithm methods? In general, much has been done in the area of optimisation in design. While not necessarily easy in their own right, applications in this area are served well by the variables being continuous. What has proved hard is dealing with configuration / layout issues because the variables are discrete and functions are discontinuous. Thus, the innovation, significance and value derived from the work begun and reported is the development of an efficient and effective method to address shape and layout issues concurrently. This implies the ability to computationally deal with the decomposition and synthesis of a system design. A genetic algorithm (GA) is used to deal with discrete aspects of a design model (e.g., allocation of space to functions) and a sequential linear programming (SLP) method for the continuous aspects (e.g., sizing).