考虑B2C平台运作效率损失的多品采销协同优化研究  

A novel collaborative optimization for B2C multi-item replenishment and sales considering operational efficiency loss on the tripartite platform

作　　者：黄成忱 许茂增[1] 崔利刚[1] 刘孝林 HUANG Chengchen;XU Maozeng;CUI Ligang;LIU Xiaolin(School of Economics and Management,Chongqing Jiaotong University,Chongqing 400074,China)

机构地区：[1]重庆交通大学经济管理学院,重庆400074

出　　处：《管理工程学报》2022年第3期236-244,共9页Journal of Industrial Engineering and Engineering Management

基　　金：国家自然科学基金资助项目(72172022);教育部人文社会科学研究青年项目(21YJC630016);重庆市技术预见与制度创新项目(cstc2020jsyj-zdxwt B0003)。

摘　　要：随着B2C行业生态链的完善,竞争加剧,平台对网络零售商的运作效率提出了更高的要求,而退货和延时服务都会造成运作效率的损失。本文紧密结合B2C运作实践,提出一个考虑平台运作效率损失的多品采销协同优化模型,将“采-存-销”三个供应链运作的核心环节进行集成优化。研究通过JRP、UJRP、EJRP三种模型的参数灵敏度对比分析,发现B2C运作模式在抵抗需求波动上具有一定优势。其中,对于EJRP模型而言,商品平均退货率对总成本的扰动较大。因此,本文建议决策者加强对退货率、延时服务率等影响平台运作效率系列指标的控制与管理。同时,本文设计了一种混合蝙蝠差分算法(BADE),通过不同问题规模的算例实验,证明算法的有效性及鲁棒性。另外,该算法着重关注可行解的多样性,在多变的网络零售环境下,给予决策者灵活的决策空间,更具实用性。As the B2 C industry ecosystem improves and competition intensifies,platforms are placing greater demands on the operational efficiency of online retailers.Frequent returns and delayed services both will cause operational efficiency losses.Based on B2 C practices,this paper proposes a collaborative optimization model for multi-item replenishment and online operations of retailers(EJRP),which takes the loss of the operational efficiency into account,aiming to realize integrated optimization among replenishment,inventory and online distribution through B2 C platforms.First of all,based on the multi-item replenishment model(JRP)and the multi-item replenishment model on random demands(UJRP),we have put core items of the total operating costs of online retailers into consideration,and realized collaborative optimization on costs among reduced operational efficiency,purchasing,and inventory,thus,enabled a more systematic decision-making with strong guiding significance in practice.Secondly,considering that the EJRP model is essentially an NP-hard problem,this paper designs a two-stage optimization algorithm(BADE)to solve this model.As for the BADE algorithm,which integrates both advantages of bat algorithm(BA)and differential evolution algorithm(DE),achieves optimized resolution for classified types of decision variables.Thirdly,according to the real case data of 6 product items,the JRP,UJRP,and EJRP models are solved respectively,and the three algorithms of GA,DE,BA and BADE algorithm are selected so as to compare and analyze the calculation results of the optimal cost value.At the same time,through the comparison of the value of k;(product purchase frequency)in the solution plan,the diversity of the plan is examined and understood.Fourthly,by analyzing the sensitivity of model parameters,we discussed the effect on the optimal cost result from perturbations of input parameters.Finally,applied cases further validated performances of the algorithm in problems of different scales.Based on above analysis,our research foun

关 键 词：决策科学 运作效率 随机需求 联合补货 差分算法 

分 类 号：F252[经济管理—国民经济]