The multi-depot vehicle routing problem with profit fairness

This data is generated in order to investigate the multi-depot vehicle routing problem with profit fairness (MDVRP-PF), a bi-objective optimization problem that adds a fairness objective function to the classical cost minimization function. By studying the MDVRP-PF , we explore the effects of integrating fairness in the optimization process.In order to perform the desired experiments, artifcial MDVRP-PF instances are generated. These instances represent different configurations that could be suitable for carrier coalitions, especially with respect to customer locations and stand-alone revenue share of each carrier. In this sense, we differentiate between two types of customer locations (clustered vs. uniform) and two types of initial revenue share distribution (balanced vs. unbalanced). In clustered instances customers are placed closer to the depots of the carriers, while being randomly located in the uniform type. Both types represent possible realistic situations, where partners are located in different distant industrial/commercial regions or within the same urban area. Regarding revenue share, in balanced instances, all carriers contribute a similar amount of revenue. Contrarily, in unbalanced instances, notable differences exist in the initial revenues contributed by each carrier. For each pair of location-revenue share configurations (from now on coded as C B, C U, U B and U U), we generate a set of instances. Each set contains three instances of different sizes: 2 depots and 100 customers (2D 100C), three depots and 150 customers (3D 150C), and four depots and 200 customers (4D 200C). To keep simplicity of the experiments, all customers have the same demand (10) and the same revenue (100).本数据集旨在研究考虑收益公平性的多仓库车辆路径问题（multi-depot vehicle routing problem with profit fairness, MDVRP-PF），这是一类在经典成本最小化目标函数基础上新增公平性目标函数的双目标优化问题。通过对MDVRP-PF的研究，我们旨在探究将公平性纳入优化流程后所产生的影响。为开展预期实验，我们生成了人工构建的MDVRP-PF测试实例。这些实例涵盖了适用于承运人联盟的不同配置场景，尤其体现在客户点位分布与各承运人的独立收益占比两个维度。据此，我们区分了两类客户点位分布模式：集群式与均匀式，以及两类初始收益占比分布模式：均衡式与非均衡式。集群式实例中，客户被布置在距离承运人仓库较近的区域；而均匀式实例中的客户点位则为随机分布。两类分布模式均对应了现实中可能存在的场景：合作伙伴分别位于不同的远程工业/商业区域，或是处于同一城市圈内。关于收益占比，在均衡式实例中，所有承运人的收益贡献规模相近；与之相反，非均衡式实例中，各承运人的初始收益贡献存在显著差异。针对每一组「客户点位-收益占比」配置组合（后文将其编码为CB、CU、UB及UU），我们生成了若干测试实例。每个实例集合包含三种不同规模的实例：2个仓库与100个客户（2D100C）、3个仓库与150个客户（3D150C），以及4个仓库与200个客户（4D200C）。为简化实验流程，所有客户的需求均设为10，收益均设为100。