A汽车4S店零部件需求预测研究

发布时间：2018-04-20 13:08

本文选题：汽车零部件 + 需求预测　；参考：《北京交通大学》2017年硕士论文

【摘要】：最近几年新车的盈利前景不被看好,新品牌车辆的销售更是令人担忧,而汽车售后市场方面的业务逐渐发展,成为了汽车4S店的重点发展核心。本文以单个新品牌汽车4S店为出发点,发现该4S店零部件需求预测不当引发的一系列问题。本文旨在选择适合该汽车4S店售后零部件的分类方法和需求预测模型,解决零部件需求预测问题,为该4S店零部件库存管理提供数据和模型支持。通过实地调研,发现该4S店售后零部件库存管理不当,经验主义较为严重,分类方法不合理、需求预测不准确,导致其存储很多不必要的零部件,同时还会采购一些不必要的零部件。部分零部件缺货现象时有发生,零部件库存积压现象较为严重,长期这样会造成库存停滞的恶性局面,降低客户满意度,降低该4S店的盈利能力,影响该4S店的发展。考虑到汽车零部件随机性强且需求波动较大,而需求的波动是影响库存决策的关键,因此,本文希望从优化售后零部件的分类和需求预测角度来改善A汽车4S店零部件的库存管理现状。本文在相关理论研究的基础上,对需求预测分两个阶段进行研究。第一阶段利用数据包络分析法与聚类分析相结合的方法对该4S店的售后零部件进行分类,找出维修零部件中具有预测意义的关键备件,提高零部件分类的有效性和零部件管理的针对性。通过与原有分类方法做比较,证明了该分类方法的有效与合理。第二阶段利用自适应变异粒子群参数寻优的最小二乘支持向量机算法预测该4S店零部件的需求量,提高该4S店汽车零部件需求预测准确率,为售后零部件的库存计划与管理优化提供科学的数据支持,为降低库存及物流成本、减少库存的滞后和积压、提高售后服务质量与客户满意度提供科学的理论支持。通过与BP神经网络模型和多元回归分析得出的需求预测值进行比较,分析并比较其误差,验证了最小二乘支持向量机用于预测4S店汽车零部件需求量的相对准确性。研究结果表明,该方法可以用于4S店进行零部件需求预测,进而改善零部件库存管理现状。本文包括图16幅,表13个,参考文献54篇。
[Abstract]:In recent years, the profit prospects of new cars have been low, the sales of new brand vehicles are even more worrying, and the development of after-sale business has become the core of the development of 4S stores. In this paper, a series of problems caused by improper forecasting of spare parts demand in 4S shop of a new brand automobile are found as the starting point. The purpose of this paper is to select the classification method and demand forecasting model suitable for the after-sale parts of the 4S shop, to solve the demand forecasting problem of the parts, and to provide data and model support for the inventory management of the 4Sshop parts. Through field investigation, it was found that the 4S store had improper inventory management of after-sale parts, more serious empiricism, unreasonable classification methods and inaccurate demand prediction, which led to the storage of many unnecessary parts. At the same time will also purchase some unnecessary parts. Part of the spare parts out of stock phenomenon occurs from time to time, parts inventory backlog phenomenon is more serious, such a long-term will lead to the vicious situation of inventory stagnation, reduce customer satisfaction, reduce the profitability of the 4S store, affect the development of the 4S store. Considering the randomness of automobile parts and the large fluctuation of demand, the fluctuation of demand is the key to the decision-making of inventory. This paper hopes to improve the inventory management status of the parts in 4S shop from the point of view of optimizing the classification of aftermarket parts and forecasting the demand. On the basis of relevant theoretical research, this paper studies demand forecasting in two stages. In the first stage, the method of data envelopment analysis and cluster analysis is used to classify the after-sale parts of the 4S shop, and to find out the key parts in the maintenance parts which have predictive significance. Improve the effectiveness of parts classification and the pertinence of parts management. By comparing with the original classification method, it is proved that the method is effective and reasonable. In the second stage, the least squares support vector machine (LS-SVM) algorithm is used to predict the demand of the 4S shop parts, and to improve the accuracy of the 4Sshop auto parts demand prediction. It provides scientific data support for after-sale parts inventory planning and management optimization, scientific theoretical support for reducing inventory and logistics costs, reducing inventory lag and backlog, improving after-sales service quality and customer satisfaction. Compared with BP neural network model and multivariate regression analysis, the error is analyzed and compared, and the relative accuracy of least square support vector machine (LS-SVM) in predicting the demand of automobile parts in 4Shop is verified. The results show that this method can be used to predict the demand of parts in 4S shop and improve the inventory management situation. This paper includes 16 figures, 13 tables and 54 references.
【学位授予单位】：北京交通大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F426.471

【参考文献】