Abstract: In e-commerce logistics, the stacking method of cargo boxes in delivery vehicles impacts the overall efficiency of order picking and delivery. This paper considers the constraints of the unloading sequence during the delivery process and analyzes how different stacking methods affect unloading efficiency, picking order, and batching strategies. Aiming to minimize the time spent on picking and unloading, an Integrated Order Picking and Delivery Scheduling model (IOPDS-BUS) that takes into account cargo box unloading sequence constraints has been developed. Furthermore, an Improved Simulated Annealing algorithm (ISA) has been proposed to find more efficient stacking methods, batching strategies, and picking sequences. The results demonstrate that the ISA algorithm can effectively address the IOPDS-BUS problem and outperforms traditional algorithms. Moreover, the IOPDS-BUS model can enhance the overall efficiency of logistics centers. However, in practical logistics operations, optimizing picking and unloading times simultaneously is challenging, but it is feasible to improve unloading time at the expense of picking time, thus lowering logistics costs and enhancing overall delivery efficiency.