【作者】 胡钰松；

【导师】 陈新；

【作者基本信息】 广东工业大学 ， 机械电子工程， 2012， 硕士

【摘要】 在市场经济中,成本和效率是一个企业能够生存壮大的关键。对于制造型企业而言,备件是企业生产所需要的重要物资,备件库存有利于维持企业的正常运行,促进生产过程的连续性,减少企业因备件短缺而造成的损失。备件管理应该做到把备件供应不及时所造成的损失减少到最低限度,同时把备件库存储备资金压缩到合理供应的最低水平。本文研究了模具制造企业的生产特征及备件、备件管理特征,结合某模具制造企业现状说明了传统ABc分类法在其备件库存管理中的局限性,分析了这种局限性产生的原因,借此提出了一种适合模具制造企业现状的备件库存ABCD分类模型,该模型把影响备件库存的六个重要因素作为分类指标,以A类、B类、C类、D类作为四种分类类别,同时给出了相应的库存管理策略。结合BP神经网络的优点求解备件库存ABCD分类模型,提取样本数据集,借助MATLAB软件仿真设计,通过大量网络训练,初步建立基于BP神经网络的备件库存ABCD分类模型。针对BP神经网络训练过程中出现的收敛慢、学习效率低、易陷入局部最小值的不足,引用遗传算法对BP神经网络算法进行优化改进,弥补神经网络的不足,取长补短,建立了基于遗传算法的BP神经网络备件库存ABCD分类模型,最后进行精度测试,通过精度比较,得出基于遗传算法的BP神经网络备件分类模型的优异性。以某模具制造企业的实际数据为依托,引入现有库存量作为判断条件,实例运算、对比验证了基于遗传算法的BP神经网络备件库存ABCD分类模型的合理性。最后,对本论文进行了总结,对模具制造企业备件库存管理的发展做出展望。更多还原

【Abstract】 Cost and efficiency are the key for an enterprise to survive and thrive in market-oriented economy. As these manufacturing enterprises, it is very important to maintain the normal operation for spare parts inventory. Spare parts inventory can keep manufacturing enterprises operation, promote the continuity of the production, and reduce the losses caused by shortages of spare parts. Spare parts inventory management should minimize the losses, and simultaneously reduce stock-out cost to the reasonable lowest level.The dissertation, researched characteristics of production, spare parts and spare parts management in these mold manufacturing enterprises, combined with current situation of a mold manufacturing enterprises, explained the limitations of the traditional ABC classification in the spare parts inventory management, analyzed the reasons. For this purpose, it proposed a suitable ABCD classification of model spare parts inventory for the mold manufacturing enterprises, set the six important influence factors as classification index of model and took Class A, class B, class C, class D as the four spare parts category of model, provided corresponding inventory management strategy.Combining with the advantages of BP neural network, the dissertation solved the ABCD classification model of spare parts inventory, though the sample data which was extracted and MATLAB software, designed the simulation, and made lots of network training, preliminary built the spare parts inventory classification model based on BP neural network.Because of the slow convergence and learning efficiency of BP neural network, it is easy to fall into the local minimum and make BP neural network invalid, The dissertation, put forward to apply genetic algorithm to optimize and improve BP neural network, made up the insufficient of neural network, learned from each other, built the spare parts inventory classification model based on genetic algorithm and BP neural network, though testing and comparing the precision, proved the advantage of the spare parts inventory ABCD classification model that is based on GA-BP. Though computing and contrasting the actual data of the mold manufacturing enterprises, the dissertation, introduced the existing inventory as the judgment condition, and verified the rationality of the spare parts inventory classification model based on GA-BP.Finally, the dissertation made the summary and made the prospect for inventory management development of spare parts in the mold manufacturing enterprises.更多还原