基于核心零部件聚类的飞机现场可更换单元划分

为研究军用飞机测试性设计过程中现场可更换单元（LRU）划分工作，以功能独立性和结构独立性为基本原则，以降低飞机维修保障成本和时间为主要目标，设计了一种军用飞机LRU划分方案。方案在首先实现了飞机从整机到零部件的功能结构分层划分基础上，从设备零部件层级出发，以综合重要度作为各零部件的定量化指标，运用帕累托（Pareto）原则筛选出核心零部件。然后基于核心零部件开发了一种LRU划分聚类算法，该算法以军用飞机设计研制阶段、使用保障阶段、退役处置阶段全寿命周期内成本和时间为优化目标构造综合评判因子，实现了非核心零部件与LRU模块之间的聚类组合，得到最优的LRU划分结果。最后以某型军用飞机上蒸发循环制冷装置为例，运用本文设计的方案实现了该装置LRU的划分，并将划分结果与该装置的实际LRU清单进行对比，通过维修保障成本和时间综合评判因子对两种划分结果的优劣性进行分析，得出本文设计方案得到的LRU划分结果对飞机维修保障成本和时间的优化效果较为均衡的结论。

Partition of line replaceable units in aircraft based on clustering of key components

In order to study the module partition of Line Replaceable Unit (LRU) in the process of military aircraft testability design, a module partition scheme of LRU for military aircraft is designed based on the principle of reducing the fault location time and fault isolation time in the maintenance and support stage of military aircraft. Firstly, a macro-level initial partition model based on Product Lifecycle Management database is constructed to realize the initial partitioning from the whole aircraft to the system and then to the subsystem or equipment. Then, a module partition clustering algorithm of LRU based on the key components of equipment is developed. In the algorithm, on the basis of obtaining the key components of the equipment by using Pareto principle, a comprehensive evaluation factor is constructed aiming at the cost and time during the military aircraft's whole life cycle of design and customization stage, maintenance and support stage, end-of-life stage. By the comprehensive evaluation factor, the optimal combination of non-key components and the interacted targets with them is realized, and finally a LRU partition scheme is formed. Taking the system layer design of a military aircraft as an example and the subsystem or equipment layer design process of the aircraft's environmental control system as an example, the preliminary division module of the macro-level is validated and the verification results are consistent. The LRU module of the evaporative cycle refrigeration device in the aircraft environmental control system is divided by using the LRU module partition clustering algorithm, and the partition results are compared with the actual LRU inventory of the device, comparisons show that the results are consistent with each other.