【作者】 林金；

【导师】 钟志华； 邓乾旺；

【作者基本信息】 湖南大学 ， 车辆工程， 2011， 硕士

【摘要】 随着制造业的飞速发展,质量、成本、效率越来越受人们关注。特别对于具有代表性的汽车行业,汽车厂商之间的竞争日益激烈,只有不断提高生产质量和效率、降低成本才能不会被淘汰。汽车装配线的不平衡会严重影响汽车的装配质量和效率。本文以中气轿车小批量试制规划为背景,提出汽车装配线平衡优化的方法。建立了一个新装配线平衡模型,将作业难度参数融入到装配线平衡中,使得装配线的平衡优化结果更加人性化,给工人产生公平感,调动劳动积极性,这对产品装配线的平衡优化有一定的指导意义。利用DELMIA三维数字化仿真系统,分析中气轿车的装配工艺,从而来确定装配作业元素之间的关系。通过采用虚拟装配技术,实现了产品、资源、工艺数据的统一管理,同时可以实现零部件设计、工装夹具设计与工艺制定的并行化,真正实现了DFA,缩短产品开发周期。根据装配线平衡问题的特点,设计了一种基于遗传算法的高效求解策略。在种群的初始化设计时编码满足作业先后关系,使得所有个体都是可行的作业序列,实现算法处理问题的高效性。译码的设计根据问题类型的特点兼顾了约束条件与目标函数。杂交与变异都是以不违背作业元素的先后关系为前提,杂交的后代继承了父代优良的顺序属性,变异实现了后代的多样性。对设计的遗传算法的个体选择策略与变异概率提出了改进策略:采用了模拟退火算法的思想对个体适应度函数进行了改进,改善了个体的选择方式;随着进化的进行,个体的变异概率设计成变化的,增强了算法的全局搜索能力。实例验证了改进后遗传算法的实用性,其不仅可以获得比较理想的装配线平衡效果,而且具有很高的计算效率。采用文章介绍的装配线平衡方法,求解了中气轿车总装试制线平衡问题,在具体的分析过程中考虑了实际资源约束,得到良好的优化结果,可以指导汽车厂商对汽车装配线的平衡优化。更多还原

【Abstract】 With the rapid development of the manufacturing industry, people attach more and more attention to the quality, cost and efficiency. Especially in the typical motor industry, automobile manufacturers compete more and more fiercely nowadays. They will be eliminated if not to improve production quality and efficiency and reduce costs. The imbalance of Automobile assembly line will seriously affects the quality and efficiency of vehicle assembly. The paper proposed the methods of optimizing car assembly line balancing in the background of the plan that trial-produce small quantities of Zhongqi Sedan.Establishing a new assembly line balance model, infusing operation difficulty parameters into assembly line balance, make a more humanistic results of optimization of the assembly line balance, and give workers a sense of fairness and motivate them, which lead a conduct significance to the optimization of the realistic assembly line balance.The method of three-dimensional digital simulation is adopted to analyze the assembly process through using the DELMIA system, and then the relationship of task elements can be fixed. By using the visual assembly technology, we realized unified management of the product, resource and process data, and parallelization of the vehicle modification, tolls and fixture design and technology formulating, to achieve truly DFA and shorten the product development cycle.A genetic algorithm was designed to solve assembly line balancing problem efficiently. The coding result can meet tasks relationships in the design of population initialization, realizes the high efficiency of the algorithm processing problem. The decoding principles are based on the constraint conditions and objective functions. Both crossing and mutating are not against the constraints of tasks relation. The offspring succeed the excellent task order after crossover between their parents. Mutating diversifies individuals. Improvement strategies on the selection principle and mutating were put forward. Individual fitness function was modified according to the idea of the simulated annealing algorithm, which improves the way of choosing individuals. To enhance the capability of global search, the probability of mutating was designed to change with the generation. The improved genetic algorithm was verified efficient and acceptable with an example.The trial-produce line balancing of Zhongqi sedan is analyzed by using the introduced method of assembly line balancing in the paper. The actual resource constraints are taken into consideration in the analysis process and the results are optimized. Therefore, the method can guide automobile manufacturers to optimize their car assembly line balancing.更多还原