Many-objective optimisation is a design problem, having more than 3 objective functions, which is found to be difficult to solve. Implementation of such optimisation on aircraft conceptual design will greatly benefit a design team, as a great number of trade-off design solutions are provided for further decision making. In this paper, a many-objective optimisation problem for an unmanned aerial vehicle (UAV) is posed with 6 objective functions: take-off gross weight, drag coefficient, take off distance, power required, lift coefficient and endurance subject to aircraft performance and stability constraints. Aerodynamic analysis is carried out using a vortex lattice method, while aircraft component weights are estimated empirically. A new self-adaptive meta-heuristic based on decomposition is specifically developed for this design problem. The new algorithm along with nine established and recently developed multi-objective and many-objective meta-heuristics are employed to solve the problem, while comparative performance is made based upon a hypervolume indicator. The results reveal that the proposed optimiser is the best performer for this design task.

多目标优化是一个设计问题，具有3个以上的目标函数，发现很难解决。在飞机概念设计上实施这种优化将极大地有益于设计团队，因为提供了许多折衷的设计解决方案以供进一步决策。本文针对无人飞行器（UAV）的多目标优化问题提出了6个目标函数：起飞毛重，阻力系数，起飞距离，所需动力，升力系数和承受飞机性能的耐久性以及稳定性约束。空气动力学分析是使用涡流格子法进行的，而飞机部件的重量是凭经验估算的。针对此设计问题专门开发了一种新的基于分解的自适应元启发式算法。该新算法以及已建立和最近开发的9种多目标和多目标元启发式算法被用来解决该问题，同时基于超量指标做出了比较性能。结果表明，所提出的优化器是该设计任务的最佳执行者。