Multidisciplinary System Design Optimization for Lunar Transportation Missions via Surrogate-Assisted Evolutionary Algorithms

In response to the advancement of lunar exploration programs, demands on transportation to cislunar space is increasing. In such missions, transporting as much payload as possible within a short period of time is a critical issue. However, finding an optimal system design is not easy, because a spacecraft system consists of multiple subsystems that are tightly coupled with each other, where independent techniques and experiences are required. Multidisciplinary design optimization (MDO), which simultaneously considers multiple disciplines to search for an optimal design, is one of the promising solutions to the present problem. Many of existing works on MDO deal with aircraft, launch vehicles, and Earth-orbiting satellites. In deep space missions, however, there are an infinite number of possible trajectories with different geometries on which the spacecraft system design strongly depends. This paper proposes a multidisciplinary system design optimization technique for lunar transportation missions with electric propulsion. A lunar transfer trajectory optimization problem is formulated incorporating design parameters of the spacecraft system. A surrogate model that predicts the response of trajectory optimization to the design parameters is developed. Using the surrogate model, an optimal system design with maximal payload mass and minimum time of flight is obtained. Results reveal the structure of the solution space of the spacecraft system design problem, making it possible to find feasible designs and global optimum.