军工企业生产与物流系统优化关键问题研究

发布时间：2017-12-27 04:29

本文关键词：军工企业生产与物流系统优化关键问题研究　出处：《沈阳理工大学》2016年硕士论文　论文类型：学位论文

【摘要】：以智能制造、智慧物流为特征的新型工业化改革已是大势所趋,我国在该领域的研究也正在全面展开。在制造型企业中,军工制造企业由于其特殊的作用,在我国的经济发展中扮演着重要角色。限于我国特殊国情,该型企业在智能化、信息化方面的建设普遍处于落后状态。企业从接到客户订单直至订单运输至客户的全过程,多采用以人工经验为基础的粗放型运作管理方式,导致其生产环节,特别是最后的产品装配过程效率低下;其物流环节,特别是从企业到客户的运输成本过高,客户满意度较低。提高军工企业的智能化、信息化水平有助于提高我国军工制造行业整体竞争力,并实现顺利转型。本文以我国某引信制造企业为研究平台,针对企业中存在的问题,结合当前的研究热点,从引信的生产和物流过程中提炼出两个需要优化的问题,建立优化模型,设计求解算法。最终给出可以指导企业生产运作和物流管理的决策依据,以提高企业的智能化,信息化管理水平。主要研究两部分问题:引信装配序列优化问题研究;具有随机特征的引信订单生产和运输协调调度问题。对两部分具体内容做如下概述:(1)针对引信的机械结构,分析零件之间的几何约束关系,以最大化满足几何约束次数,以及最小化装配方向改变次数为目标建立优化模型。基于离散粒子群算法,提出产生初始解的启发式算法,以及以深度临域搜索为特征的改进策略,通过实例验证此算法在求解速度和稳定性方面优于其他算法。结合军工企业对产品质量的要求,从装配质量角度,建立装配序列评价指标体系,最终从多个解中得到有助于提高装配质量的装配序列。(2)针对引信订单生产和运输两个环节,考虑到订单生产过程中可能出现不可预测情况,将订单加工时间考虑为随机变量,订单的加工环境考虑为单机环境。订单运输过程采用批运输的形式,客户数量考虑单客户和多客户两种情况,分别建立了以订单期望完工时间和与总运输费用之和最小化为目标的优化模型。对于两种情况下,分别证明了问题是多项式时间可解和NP-hard的,并分别设计了带有随机变量的动态规划算法和粒子群优化算法予以解决。通过实际算例,验证了算法的有效性和稳定性。
[Abstract]:The new industrialization reform, which is characterized by intelligent manufacturing and intelligent logistics, has been the trend of the times, and the research in this field is also being carried out in an all-round way. In the manufacturing enterprises, the military manufacturing enterprises play an important role in the economic development of our country because of their special functions. Limited to the special national conditions of our country, the construction of this type of enterprise is generally in the backward state in the construction of intelligence and information. The whole process of receiving customer orders until they are transported to the customer from the enterprise, the extensive operation management mode based on human experience, the production process, especially the low efficiency of the final product assembly process; the logistics links, especially from the enterprise to the customer the high transport costs, low customer satisfaction. Improving the intelligence and information level of military industrial enterprises will help to improve the overall competitiveness of our military manufacturing industry and achieve a smooth transition. In this paper, a fuze manufacturing enterprise in China is taken as the research platform, aiming at the existing problems in the enterprise, combined with the current research hotspots, we extract two optimization problems from the fuze production and logistics process, establish the optimization model and design the solving algorithm. Finally, the decision basis which can guide the operation of enterprise production and logistics management is given in order to improve the intelligence of the enterprise and the level of information management. This paper mainly studies the two part of the problem: the optimization of fuse assembly sequence; the problem of order production and transportation coordination scheduling with random characteristics. The following two parts are summarized as follows: (1) aiming at the mechanical structure of fuzes, the geometric constraint relationship between parts is analyzed, and the optimization model is established to maximize the number of geometric constraints and minimize the number of changes in assembly direction. Based on discrete particle swarm optimization (PSO), a heuristic algorithm to generate initial solution and an improved strategy based on deep in search are proposed. The algorithm is proved to be superior to other algorithms in solving speed and stability. Considering the requirements of military enterprises for product quality, the assembly sequence evaluation index system is established from the perspective of assembly quality, and finally, assembly sequences that are helpful to improve assembly quality are obtained from multiple solutions. (2) in view of the two links of fuze order production and transportation, considering the unpredictable situation of the order production process, the order processing time is considered as a random variable, and the processing environment of the order is considered as a single machine environment. The order transportation process takes the form of batch transportation. Considering the two situations of the number of customers, single customer and multiple customers, the optimization model is established to minimize the sum of the expected time of completion of the order and the total transportation cost. For the two cases, it is proved that the problem is polynomial time solvable and NP-hard, respectively, and dynamic programming algorithm with random variables and particle swarm optimization algorithm are designed respectively to solve them. The effectiveness and stability of the algorithm are verified by a practical example.
【学位授予单位】：沈阳理工大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：F426.48;F273;TP18

【参考文献】