基于双层遗传算法的飞机燃油热管理系统架构寻优

Optimization of Aircraft Fuel Thermal Management System Architecture Based on a Double-Layer Genetic Algorithm

摘要:

新型飞机燃油热管理系统换热网络排布情况复杂多样，传统的设计方法依赖于设计人员的经验，而简单的排布形式换热能力不足、燃油热沉利用不充分。针对上述问题，开展飞机燃油热管理系统拓扑架构优化研 究，提出基于等势点思想的燃油热管理架构表征方法，采用深度搜索算法生成燃油热管理系统所有可能的拓扑架构；利用双层遗传算法对拓扑图进行优化，以解决架构数量过多、遍历优化耗时过长的问题。结果表明：相比于采用纯遍历的搜索算法，双层遗传算法的优化速度提升833 倍，6 个子系统总计38 703 种架构花费1.29×103 s 完成计算；优化的架构在保持较高总吸热量下，平均散热方差最小、散热耗油量最小，证明了所提燃油热管理系统拓扑架构优化方法的有效性。

Abstract:

The aircraft fuel thermal management system architectures’ arrangement is complex. Traditional design methods use thermal capacity insufficiently and underutilize fuel heat sinks. This paper conducts research on the optimization of the topological architecture of aircraft fuel thermal management systems. A characterization method for fuel thermal management architectures based on the concept of equipotential points is proposed, and a depth-first search algorithm is employed to generate all possible topological architectures of the fuel thermal management system. To address the issue of excessive architectures and time-consuming exhaustive optimization, a dual-layer genetic algorithm is applied to optimize the topological graphs. Results demonstrate that compared to pure exhaustive search, the dual-layer genetic algorithm achieves an 833-fold speed increase, completing the calculation for 38,703 architectures across 6 subsystems in 1.29×103 seconds. The optimized architectures exhibit minimal average heat dissipation variance and the lowest fuel consumption for heat dissipation while maintaining high total heat absorption. This prove the method"s effectiveness.