Goal programming applications in multidisciplinary design optimization

In this paper, we begin to explore the ramifications of a goal programming approach for multidisciplinary design optimization based on the Collaborative Optimization framework. Working within an existing computing architecture for simulation-based design, we propose a bi-level goal programming formulation of Collaborative Optimization to automate the design synthesis of complex systems, with particular emphasis on undersea vehicles. A brief overview of the simulation-based design computing architecture that has been developed is given. To demonstrate the proposed bi-level goal programming approach, the multidisciplinary design and optimization of an undersea vehicle containing four subsystem level analyses and a system-level analysis is presented. A traditional single-level optimization solution is also presented to provide a benchmark for comparison. While the bi-level goal programming approach is computationally more expensive, it does improve overall system performance by automating subsystem synthesis and arbitration during optimization.