This paper presents the multi-objective optimization of a small aircraft turbine engine rotor system subjected to maneuver loads. Application of a clustering algorithm, an unsupervised machine learning technique, to the Pareto front developed from multi-objective optimization of maneuvering aircraft rotor system is the novelty of the present work. An in-house finite element code is developed using MATLAB for the analysis of rotor system. Hybrid Genetic Algorithm is employed to simultaneously minimize the rotor response at maximum speed during maneuver and rotor response at critical speed with restrictions imposed on critical speed. Shaft diameters and pedestal stiffness at both the bearing locations are identified as design variables. Pareto optimal solutions are generated, clustering is carried out in both objective space and decision space and the solution close to the utopia point is selected as final compromise solution. The average of the values of design variables for the selected cluster is compared with final compromise solution and is found in good agreement. The response of the rotor system and the critical speeds are verified by carrying out tests on ground.

本文介绍了受机动载荷作用的小型飞机涡轮发动机转子系统的多目标优化。将聚类算法（一种无监督的机器学习技术）应用于从机动飞机旋翼系统的多目标优化发展而来的帕累托前沿是当前工作的新颖之处。使用 MATLAB 开发了用于转子系统分析的内部有限元代码。混合遗传算法用于同时最小化机动期间最大速度下的转子响应和临界速度下的转子响应，并对临界速度施加限制。两个轴承位置的轴直径和基座刚度被确定为设计变量。产生帕累托最优解，在目标空间和决策空间都进行聚类，选择接近乌托邦点的解作为最终的折衷解。将所选集群的设计变量值的平均值与最终折衷解决方案进行比较，发现它们具有良好的一致性。转子系统的响应和临界速度通过在地面上进行测试来验证。