关键词: 经济性修理级别分析, 民机维修, 二进制粒子群算法, 人为因素成本, 混合惩罚函数

Abstract: Modular design is widely used in modern civil aircraft products. It has multi-indenture characteristics, and its corresponding maintenance sites are also divided into multi-echelon maintenance levels. When the existing repair level analysis model is moved or discarded, it may have the problem of repeated accumulation of the cost of module and its internal packaging subunit. In this paper, the following behavior matrix is introduced to mark the decisions made by the hierarchy constraints of the parent unit, which can effectively avoid the repeated calculation of the costs of the parent-unit and the child-unit. On this basis, the influences of human factors such as No Fault Found (NFF) event and secondary maintenance on LORA’s decision aere considered when building the repair level analysis model suitable for civil aircraft. Compared with the existing model, the proposed model can more accurately reflect the estimated cost of decision-making more truly, not only improving the credibility of the decision but also increasing the applicability of the project. Finally, the mixed penalty function method is used to transform the original problem into an equivalent unconstrained optimization problem. An improved Binary Particle Swarm Optimization (BPSO) algorithm is used to simulate and optimize the three-indenture and three-echelon case with multi-fault mode. The results verified the rationality of the economic analysis model and the effectiveness of the algorithm.

Key words: economic repair level analysis, civil aircraft maintenance, bpso algorithm, human factor cost, mixed penalty function