In this code, we implemented a truck-drone collaborative delivery service for 100 customers in Harbin over a four-day period. Notably, the optimization algorithm used is the independently developed GOaMEA—a multi-objective optimization algorithm that integrates graph learning and multi-agent techniques. The model incorporates real-world road network data—including vehicle travel distance, duration, and speed—all obtained from the AutoNavi Map API. The routing paths for vehicles and drones are generated by the GOaMEA optimizer based on customer-specific parameters, while the route visualization results are provided by AutoNavi Map. The real-world case of Harbin.py is an executable program that utilizes the AutoNavi Map API to generate drone delivery route plans for vehicles based on actual road conditions. (The accompanying file 总数据.xlsx contains the complete dataset.) The file Harbin_Case_PVRPDTW.json is used to record authentic customer information for the Harbin case, including location, service frequency requirements, time windows, and other details.