The development of on-demand global transportation service procurement (oGTSP) through digital trading platforms has accelerated due to frequent fluctuations in transport capacity. In the oGTSP model, the exporter must consider logistics service quality and transport prices when sourcing global logistics services. To satisfy the continuous transport needs, procurement is conducted sequentially throughout multiple auction cycles. For a single auction, we constructed a service-level weight-scoring function and analysed the trading parties’ behavioural strategies to obtain an auction equilibrium strategy in a specific context. Then, we developed a multi-cycle sequential decision method based on a single-cycle equilibrium decision by forwarders that can dynamically adjust the auction lot size to help the exporter obtain optimal utility. Finally, based on the real case of a large electronic product exporter, the proposed approach was verified. The results demonstrated that exporters should pay more attention to the quality of service when choosing freight forwarders to improve the utility of transportation service procurement. The exporter can attract more forwarders to participate in auctions to obtain more capacity supply by increasing the weighting of service levels. Besides, the proposed auction system could effectively accommodate strategic forwarders with learning abilities. The exporter’s utility will significantly improve if the freight forwarders have learning ability. There is a marginal diminishing effect in that the benefits from additional participation of learning-oriented bidders are initially large but eventually stabilized. The strategic auction participation of learning-oriented freight forwarders smooths the capacity supply trend, reduces extreme fluctuations and makes multi-cycle predictions more accurate.

中文翻译：

---

顺序多属性全球运输服务采购中的最佳批次规模和服务水平加权

由于运输运力的频繁波动，通过数字交易平台进行按需全球运输服务采购 （oGTSP） 的发展加速。在 oGTSP 模式中，出口商在采购全球物流服务时必须考虑物流服务质量和运输价格。为了满足连续运输需求，采购在多个拍卖周期中按顺序进行。对于单个拍卖，我们构建了一个服务级别的权重评分函数，并分析了交易方的行为策略，以获得特定背景下的拍卖均衡策略。然后，我们开发了一种基于货运代理单周期均衡决策的多周期顺序决策方法，可以动态调整拍卖手数，以帮助出口商获得最佳效用。最后，基于某大型电子产品出口商的真实案例，对所提方法进行了验证。结果表明，出口商在选择货运代理时应更加关注服务质量，以提高运输服务采购的效用。出口商可以通过增加服务水平的权重来吸引更多的货运代理参与拍卖，以获得更多的运力供应。此外，拟议的拍卖系统可以有效地容纳具有学习能力的战略货运代理。如果货运代理具有学习能力，出口商的效用将显着提高。存在边际递减效应，因为以学习为导向的投标人的额外参与带来的好处最初很大，但最终趋于稳定。 学习型货代的战略性拍卖参与，平滑了运力供应趋势，减少了极端波动，使多周期预测更加准确。