中文摘要：

针对有机蔬菜网上直销的“农一宅”配送难题，以提高物流配送方案的科学性、有效性和及时性为目标，以在线实时地生成配送方案为突破口，引入状态空间搜索理论和运筹学建模技术，提出带控制策略的深度优先搜索算法来快速、高效地生成可行的车辆路径方案集合，并运用运筹学中经典的“混合下料问题”的优化思想求解车辆路径方案的最优组合，为求解车辆路径问题这一NP难题提供了新思路．开发了“农一宅”配送车辆路径方案智能生成系统，实现由计算机自动完成车辆路径方案的生成、建模和求解全过程．最后，通过将车辆路径方案智能生成方法与精确求解方法进行对比，验证了所提方法的有效性．

英文摘要：

The paper focuses on the ＇farm-to-door＇ delivery problem of organic vegetables in online direct sales, with the objective of enhancing the scientific, efficient and on-time processing level. Taking online routing scheme generation as a breakthrough point, and by applying state space search theory and modeling technologies of operations research, it addresses a depth-first search algorithm with control rules to generate the set of feasible routing schemes quickly and efficiently. Then it obtains the optimal combination of routing schemes by utilizing the optimization idea of mixed cutting problem in operations research, which provides a new solution for vehicle routing problem. Furthermore, it develops an intelligent routing scheme genera- tion system for ＇farm-to-door＇ delivery problems, which realizes an automatic solution process from scheme generation, to modeling, and to solution. Finally, it compares the results from the intelligent routing scheme generation method with those from the exact solution algorithm, and the results demonstrate the efficiency of the intelligent approach.