飞机装配生产线的科学规划和优化配置对于提高装配效率、降低生产成本具有重要意义。飞机结构、装配工艺和生产管理复杂，装配时间占整个制造周期的50%以上，因此，高效的生产线建模仿真与优化是提高生产线规划效果的有效途径。本论文以某型飞机部总装生产线规划为背景，针对现有生产线仿真软件建模效率低、生产线优化功能有限的问题，研究生产线仿真模型的快速构建，探索生产线工人配置、物料配送路径规划和生产线平衡问题的优化方法及算法，以实现生产线的快速建模仿真与优化。本文首先设计了一种基于格式化数据的生产线仿真模型快速构建方法，基于商品化生产与物流系统仿真软件Plant Simulation的二次开发接口设计开发了快速建模程序。该方法通过整合和提炼飞机装配生产线基本要素，构建能描述生产线属性的仿真模型模版，并通过建模数据与仿真模型匹配、仿真模块封装，支持仿真数据的高效读取与通信，实现仿真模型的一键快速构建。针对生产线关键要素中的工人配置和物料配送自动导引小车（Automated Guided Vehicle，AGV）路径规划两个优化问题开展研究。本文将工人优化配置建模为二维装箱问题，设计了一种改进的BFD（Best-Fit Decreasing）算法；建立了考虑装配工艺顺序约束的AGV路径最短优化模型，设计了一种基于贪心算法的优化流程，并进行了算法测试，获得了班组的最小工人数量和派工策略，给出了优化的AGV运输路径。面向飞机总装生产线平衡优化问题，通过分析飞机总装工艺流程及特点，建立了以生产节拍最短、平衡率最高为优化目标，装配工艺顺序为约束条件的生产线平衡优化数学模型，设计了一种利用作业优先图矩阵定量描述装配工艺顺序、采用遗传算法原理的生产线平衡优化流程及算法，能获得满足生产线产能和均衡生产需求的总装生产线站位划分结果。结合某型飞机部装和总装两条生产线的规划仿真需求，对本文设计开发的生产线仿真模型快速构建程序与三种优化方法及算法进行了应用验证，显著提升仿真模型的构建效率，获得了优化的工人配置方案和AGV路径，且相比人工规划方案，优化后的总装站位划分方案具有更短的生产节拍和更高的生产线平衡率，表明了优化方法及算法的有效性。
Aircraft structures, assembly processes, and production management are complex, with assembly time accounting for over 50% of the entire manufacturing cycle. Scientific planning and optimized configuration of aircraft assembly production lines are of great significance for improving assembly efficiency and reducing production costs. Therefore, efficient production line modeling, simulation, and optimization are effective ways to improve the effectiveness of production line planning. This thesis, set against the background of a specific aircraft final assembly line planning, addresses the issues of low modeling efficiency and limited optimization functions of existing production line simulation software. It explores methods and algorithms for the rapid construction of simulation models, optimization of worker allocation, material delivery path planning, and production line balancing, aiming to achieve rapid modeling, simulation, and optimization of the production line.For the problems of low efficiency and heavy workload in traditional production line model construction, this article first designs a fast construction method for production line simulation models based on formatted data. A rapid modeling program was designed and developed based on the secondary development interface of the commercial production and logistics system simulation software Plant Simulation. By integrating and refining the basic elements of the aircraft assembly production line, the method constructs simulation model templates that describe production line attributes. Through matching modeling data with simulation models and encapsulating simulation modules, it supports efficient reading and communication of simulation data, enabling one-click rapid construction of simulation models.The research focuses on two optimization problems in key production line elements: worker allocation and Automated Guided Vehicle (AGV) path planning for material delivery. The worker optimization allocation is modeled as a two-dimensional bin packing problem, and an improved Best-Fit Decreasing (BFD) algorithm is designed. An AGV path optimization model considering assembly process sequence constraints is established, and an optimization process based on a greedy algorithm is designed and tested, yielding the minimum number of workers per team and dispatching strategies, and providing optimized AGV transportation paths.For the production line balancing optimization problem, by analyzing the characteristics and processes of aircraft final assembly, a mathematical model for production line balancing optimization is established with the shortest production cycle and highest balance rate as optimization goals and assembly process sequences as constraints. A production line balancing optimization process and algorithm using a genetic algorithm and priority matrix to quantitatively describe assembly process sequences are designed. This approach can achieve production line station divisions that meet capacity and balanced production requirements.The rapid construction program for production line simulation models and the three optimization methods and algorithms developed in this thesis were applied and validated on two production lines of a specific aircraft‘s component assembly and final assembly. The results significantly improved the efficiency of simulation model construction, optimized worker allocation schemes, and AGV paths. Compared to manual planning schemes, the optimized final assembly station division scheme achieved shorter production cycles and higher production line balance rates, demonstrating the effectiveness of the optimization methods and algorithms.
