In the competitive e-commerce environment, self-operated firms independently manage procurement, logistics and sales, yet often experience inefficiencies due to weak coordination, forecast errors and supply chain disruptions. This study investigates the multi-period, multi-product inventory routing problem with procurement decisions under quantity discounts (MMIRP-QD), aiming to jointly optimize procurement, transportation and warehousing decisions to improve replenishment efficiency, operational profitability and customer satisfaction.
An integrated optimization framework is developed for the MMIRP-QD, which incorporates quantity discounts, replenishment lead times, stockout penalties and a flexible replenishment strategy. First, an innovative demand forecasting model is proposed to enhance prediction accuracy. Second, a hybrid two-level heuristic with a crossover-repair operator improves computational efficiency. Third, a rolling-horizon approach is developed to handle forecast errors and dynamic disruptions. The proposed framework is illustrated through a case study.
Computational experiments demonstrate that the proposed approach outperforms benchmark methods in solution quality and efficiency. Furthermore, the rolling-horizon approach effectively mitigates forecast-error propagation. Sensitivity analysis further explores the effects of replenishment strategy and stockout penalties, yielding actionable managerial insights.
This study extends the standard MIRP by incorporating procurement decisions under quantity discounts, thereby formulating the MMIRP-QD, and proposes an integrated framework that combines demand forecasting, hybrid heuristic optimization and a rolling-horizon approach. By linking prediction with optimization, the framework provides a robust and scalable decision support tool for self-operated e-commerce decision making.
