Abstract:  An automatic aircraft tank pipeline layout method was proposed based on the improved spotted hyena algorithm to address the problem of internal aircraft tank pipeline path planning. To improve the global search ability and convergence rate of the spotted hyena algorithm， the diffusion search mechanism was introduced to spread the optimal solution in the iteration processes， and then it was introduced to randomly search in the optimal solution to improve the algorithms convergence rate. The grid method was used to build the laying space mathematical model and the pipeline layout optimization mathematical model was established with the shortest pipeline path， the pipeline bending angle and the pipeline bending radius were as constraints. Furthermore， considering the pipeline arms layout problem， with arm length and arm posture as the arm layout double target optimization function， using a genetic algorithm（NSGA-Ⅱ）to code the initial individual， design cross， variation rules， to solve the Pareto solution of the pipeline arm layout sets， and to obtain the pipeline arm layout scheme. Finally， numerical examples and pipeline laying simulation were used to validate the effectiveness of the proposed method. 

Key words: pipeline layout, spotted hyena algorithm, bracket, genetic algorithm, multi-objective optimization

摘要： 针对飞机油舱内部管路路径规划问题，提出一种基于改进斑鬣狗算法的飞机油舱管路自动布局方法。为了提高斑鬣狗算法的全局搜索能力以及收敛速度，在种群进化过程中引入扩散搜索机制对迭代过程中最优解进行扩散，以达到提高全局搜索能力的目的；在算法迭代后期引入深度包围机制，使其在最优解中随机搜索以提高算法的收敛速度。采用栅格法构建敷设空间数学模型，以管路路径最短为优化目标，以管路路径不与障碍物发生干涉、管路折弯角度及管路折弯半径为约束，建立管路布局优化数学模型。进一步，考虑管路的支臂布局问题，以支臂总长和支臂姿态作为支臂布局双目标优化函数，运用遗传算法NSGA-Ⅱ对管路支臂布局方案进行编码生成初始个体，设计交叉、变异规则，求解管路支臂布局的Pareto解集，进而获得管路支臂布局方案。最后，通过数值算例及管路敷设仿真验证所提方法的有效性。

关键词: 管路布局, 斑鬣狗算法, 支臂, 遗传算法, 多目标优化