考虑客户随机取消订单的电商快递配送优化模型

Optimization Model for E-Commerce Parcel Delivery Considering Random Customer Order Cancellations

摘要: 为解决电商快递配送中客户随机取消订单引发的需求不确定性问题，突破传统确定性模型的局限，本文构建了考虑需求随机波动的机会约束配送优化模型。将订单需求建模为服从泊松分布的随机变量，通过分位数方法将随机机会约束转化为确定性形式。采用改进的NSGA-II算法求解，设计基于订单分配的整数编码方案，并结合部分匹配交叉、基因交换变异与2-opt局部搜索策略。基于PDPTW基准算例的数值试验表明，所提模型与算法在总成本、车辆使用量与计算效率之间取得了较优平衡。

Abstract: To address the demand uncertainty issue arising from customers’ random order cancellations in e-commerce express delivery scenarios, and to break through the limitations of deterministic assumptions in traditional delivery optimization models, this paper constructs an opportunity-constrained express delivery optimization model that considers random demand fluctuations. The order demand is modeled as a random variable following a Poisson distribution, and the random opportunity constraints are transformed into solvable deterministic forms through quantile methods. An improved NSGA-II algorithm is employed, incorporating an integer encoding scheme based on order allocation, partially matched crossover, gene swap mutation, and 2-opt local search strategies. Numerical experiments based on the PDPTW benchmark show that the proposed model and algorithm achieve a favorable balance among total cost, vehicle usage, and computational efficiency.