A multidisciplinary design optimization method for cross-domain vehicles based on distributed-centralized augmented Lagrangian coordination

A nested multidisciplinary design optimization method based on augmented Lagrangian coordination and joint geometric/aerodynamic/structural disciplines modeling is proposed to address the problem of inefficient solving in high time-consuming disciplines such as aerodynamic, structural, and trajectory, brought by the complexity of the cross-domain vehicle configurations. Firstly, integrated geometric-aerodynamic-structural parametric modeling based on the 3D CST method is proposed to realize joint geometric/aerodynamic/structural modeling. Second, a distributed-centralized augmented Lagrangian coordination method that balances efficiency and parallel computation is proposed with an adaptive asynchronous multiplier update strategy to update the penalty weights. Finally, the X37-like vehicle in the cross-domain vehicle is taken as an object, and its nested multidisciplinary design optimization process is established to carry out the optimization design with the goal of optimal mass. The optimization results show that during the whole mission, the optimized design shape increases the payload mass by 70.2 % based on the initial shape, and the performance is significantly improved, which verifies the effectiveness of the distributed-centralized augmented Lagrangian coordination multidisciplinary design optimization method proposed in this paper.