A serialization-based partial decoupling approach for multidisciplinary design optimization of complex systems

The optimization for multidisciplinary engineering systems is highly complicated, which involves the decomposing of a system into several individual disciplinary subsystems for obtaining optimal solutions. Managing the coupling between subsystems remains a great challenge for global optimization as the existing methods involve inefficient iterative solving processes and thus have higher time cost. Some strategies such as discipline reorder, coupling suspension and coupling ignoring can to some extent reduce the execution cost. However, there are still some deficiencies for these approaches such as uniform handling of the couplings, complete decoupling and heavy burden of system optimizer. To overcome the above drawbacks, a serialization-based partial decoupling approach is proposed in this study, which consists of three main steps. First, different disciplines are clustered into some subsystems by analyzing the interdisciplinary sensitivities. Then, for each subsystem, a serialization process is proposed ...