Intelligent design system for GA-based optimum design (Application to design of stiffener layout)

For searching an optimum design based on a genetic algorithm (GA), it is necessary to select a suitable algorithm for the design problem and to determine the genetic operations. This paper presents a framework for an intelligent design system which aids the selection and determination for GA-based optimization by integrating a GA and an expert system (ES). The ES effectively acts as a coordinator to control the optimization. A prototype system is developed, and the required knowledge about GA is acquired from numerical trials of stiffener layout design. The stiffener layouts obtained in the trials are also stored in the knowledge base of the system. The system is applied to the maximum stiffness design of a stiffened plate, and it is confirmed that an appropriate algorithm and genetic operations are interactively selected and the optimum layouts are searched successfully. For the effective search of the optimum design, the system functions to generate the initial population by referring to the previous design solution. The effectiveness of the system is demonstrated.