| Multidisciplinary Design Optimization (MDO) has emerged as a new technology dealing with the design of complex systems. This thesis focuses on development of an effective and efficient collaboration mechanism for coordinating coupled disciplines, as couplings are dominant issues in MDO. In this thesis, an innovative collaboration model is developed to handle these couplings. Based on this newly proposed collaboration model, two novel methods are developed for solving MDO problems. The first method is named the Boundary Search and Simplex Decomposition Method (BSSDM). It proposes a new methodology to geometrically depict the coupling information. Difficulties encountered with the couplings are alleviated through the obtained geometric information by the BSSDM. The second method is named the Collaboration Pursuing Method (CPM). It is a sampling-based MDO method, which aims to deal with relatively large-scale MDO problems. Both new methods are successfully tested. A conceptual aircraft design is implemented with the CPM and the results show that the CPM is competitively efficient as compared with other MDO methods. Limitations of the newly proposed MDO methods are discussed, along with suggestions for future work. |
