In this high-risk and high-payoff project, a plan to establish the feasibility of a new method for engineering design is presented. The new method will be based on the synthesis of several computational mechanisms inspired by biological processes of coevolution. Specifically, the research will focus on efficient methods of decomposition of complex engineering systems and on ways of making engineering designs more robust. Biological processes of cooperative and competitive coevolution will inspire the methods of achieving these goals. The effectiveness and efficiency of the proposed mechanisms will be verified using several examples of both simple and relatively complex design problems involving structural and mechanical systems. One of the expected contributions of the research would be the development of a modern computational tool supporting and enhancing design processes. So far, nature-inspired engineering design was almost exclusively focused, with relatively few exceptions, on design optimization issues in which the principles of evolution encoded as evolutionary algorithms were used to optimize designs. The goal of this project is to show how other computational mechanisms inspired by biological processes of coevolution can be efficiently used to decompose complex engineering designs (cooperative coevolution) and to make designs more robust (competitive coevolution). Thus, this research will build a bridge between modern biology, computer science, and engineering design. The generality of models, procedures, and algorithms proposed in this project will make them well suited for a wide range of engineering design applications.
This research is also expected to have a broader impact on science and education. It will make significant contributions to the theory of coevolutionary systems by providing the first comprehensive application of its concepts, models, and procedures to engineering design.
