Carrier-based unmanned aerial aircraft (UAV) structure is subjected to severe tensile load during takeoff, especially the drawbar, which affects its fatigue performance and structural safety. However, the complex structural features pose great challenges for the engineering design. Considering this situation, a structural design, fatigue analysis, and parameters optimisation joint working platform are urgently needed to solve this problem. In this study, numerical analysis of strain fatigue is carried out based on the laboratory fatigue failure of the carrier-based aircraft drawbar. Taking the sensitivity of drawbar parameters to stress and life into account and optimum design of drawbar with fatigue life as a target using the parametric method, this study also includes cutting-edge parameters of milling cutters, structural details of the drawbar and so on. Then an experimental design is applied using the Latin hypercube sampling method, and a surrogate model based on RBF neural network is established. Lastly, a multi-island genetic algorithm is introduced for optimisation. The results show that the error between the obtained optimal solution and simulation is 0.26%, while the optimised stress level is reduced by 15.7%, and the life of the drawbar is increased by 122%.

中文翻译：

---

基于应变疲劳分析的舰载无人机牵引杆参数优化

舰载无人机（UAV）结构在起飞过程中承受着较大的拉伸载荷，尤其是牵引杆，影响其疲劳性能和结构安全。然而，复杂的结构特征给工程设计带来了巨大的挑战。考虑到这种情况，迫切需要一种结构设计、疲劳分析和参数优化的联合工作平台来解决这个问题。本研究基于舰载机牵引杆的实验室疲劳失效进行应变疲劳数值分析。考虑到牵引杆参数对应力和寿命的敏感性，采用参数化方法以疲劳寿命为目标优化设计牵引杆，还包括铣刀的切削刃参数，拉杆的结构细节等等。然后采用拉丁超立方抽样方法进行实验设计，建立了基于RBF神经网络的代理模型。最后，引入了多岛遗传算法进行优化。结果表明，得到的最优解与仿真的误差为0.26%，而优化后的应力水平降低了15.7%，牵引杆的寿命提高了122%。