基于客户满意度的配送中心车辆调度优化研究

发布时间：2018-11-09 12:44

【摘要】：随着经济蓬勃发展,物流在各行各业的地位愈发重要,可谓是国民经济的命脉。然而,国内物流水平与经济发展速度并不匹配,长此以往势必拖累经济发展。时下,国家各项政策都提到需要提高国内物流水平。提高物流水平需要多方面努力,但根本目的是要降低物流费用。而物流费用中,运输费用占比达到一半,科学地进行车辆调度,可以降低运输费用、提高车辆使用率、减少空气污染等。本文在研究配送中心车辆调度优化问题时,增加了客户满意度这一因素。增加的原因在于如今的市场是买方市场,客户有众多选择,提高客户满意度可以留住现有客户并且吸引潜在客户。但是客户对企业的贡献程度或者说企业对客户的依赖程度不同,企业在如今的激烈市场环境中想要立足必须准确识别客户,重点服务那些重要客户,这里引入“客户重要程度”这一概念表示客户对企业的重要程度,而这一重要程度也将作为客户满意度的权重系数,如此更为合理。至于在实际中如何计算“客户重要程度”,将根据专家打分并运用模糊综合评价法求得各客户的重要程度。如何将满意度与调度优化问题融合在一起,文章利用满意度模糊隶属函数表示满意度,二者之间依靠车辆抵达客户处的时间联系起来。之后根据相应条件,以总配送距离最短、满意度最高为目标建立数学模型。因为遗传算法具有很强的鲁棒性和快速寻优能力,已被前人证实了其有效性,所以本文选用遗传算法求解此类问题。最后,在求解具体案例时,引入企业A旗下配送中心的案例。考虑到案例实际的客户点数量不多,如果按照多目标优化寻优,可能会导致解空间缩小,算法可能得到局部寻优结果。所以文章将分别以运输距离最短、满意度最高为单目标进行优化。首先以运输距离最短为单目标进行优化后,得到几个优良染色体,利用满意度隶属函数与满意度权重系数计算染色体的加权客户满意度,再利用“单位满意度行驶距离”指标选取最优;之后以客户满意度最高为单目标进行优化,得到几条优良染色体,再结合“单位满意度行驶距离”指标选取最优。最后再次运用“单位满意度行驶距离”指标选取前后二者间的最优结果作为最终方案。
[Abstract]:With the booming development of economy, logistics is becoming more and more important in various industries, which is the lifeblood of national economy. However, the domestic logistics level and the economic development speed does not match, in the long run is bound to drag down the economic development. Nowadays, national policies all mention the need to improve the level of domestic logistics. It takes many efforts to improve the level of logistics, but the fundamental purpose is to reduce the cost of logistics. However, transportation costs account for half of the logistics costs. Scientific vehicle scheduling can reduce transportation costs, improve vehicle utilization rate, and reduce air pollution, and so on. In this paper, customer satisfaction is increased when studying vehicle scheduling optimization problem in distribution center. The increase is due to the fact that today's market is a buyer's market, with a wide range of choices, and increased customer satisfaction can retain existing customers and attract potential customers. However, the degree of customer's contribution to the enterprise or the degree of dependence of the enterprise on the customer is different. In order to establish a foothold in today's fierce market environment, enterprises must accurately identify customers and focus on serving those important customers. The concept of "customer importance" is introduced here to indicate the importance of the customer to the enterprise, and this important degree will be the weight coefficient of customer satisfaction, so it is more reasonable. As to how to calculate "customer importance" in practice, the importance degree of each customer will be obtained by using fuzzy comprehensive evaluation method and scoring by experts. In this paper, the fuzzy membership function of satisfaction degree is used to express the satisfaction degree, which depends on the time when the vehicle arrives at the customer. Then, according to the corresponding conditions, the mathematical model is established with the shortest distance and the highest satisfaction. Because genetic algorithm has strong robustness and fast optimization ability, it has been proved to be effective by predecessors, so this paper chooses genetic algorithm to solve this kind of problem. Finally, in solving specific cases, the introduction of enterprise A distribution center case. Considering that the actual number of customer points in the case is not many, if the optimization is based on multi-objective optimization, the solution space may be reduced, and the algorithm may get local optimization results. Therefore, the article will be the shortest transportation distance, the highest degree of satisfaction for single-objective optimization. First of all, after optimizing with the shortest transportation distance as a single objective, several excellent chromosomes are obtained, and the weighted customer satisfaction degree of the chromosome is calculated by using the degree of satisfaction membership function and the satisfaction degree weight coefficient. Secondly, the optimum is chosen by using the index of "driving distance per unit satisfaction degree". After that, the highest customer satisfaction is used as a single objective to optimize, and several excellent chromosomes are obtained, and then the optimum is selected by combining with the index of "driving distance per unit satisfaction". Finally, the final scheme is to select the optimal result between the two factors.
【学位授予单位】：成都理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F274;F426.4

【参考文献】