“农-宅”直销模式下基于时空网络的鲜果采摘与配送协同优化研究

The collaborative optimization of fresh fruit picking and distribution under “Farm-to-Door” mode, is an essential issue to promote fresh agricultural products distributed to the upside, especially in the context of agricultural supply-side reforming and rural revitalization. This project focuses on the collaborative optimization problem of fresh fruit picking and distribution under “Farm-to-Door” mode. The breakthrough of this project includes maturity fitting, and collaborative optimization condition between maturity and picking and distribution decisions. The research methodologies include correlation and regression analysis, time-space network and Benders decomposition algorithm. The research problem contains four aspects. Firstly, build the maturity fitting model of fresh fruit, and investigate how to link it into picking and distribution decisions. Secondly, formulate the collaborative optimization model of fresh fruit picking and distribution. Thirdly, design an efficient algorithm to solve the proposed collaborative optimization model. Finally, take Wuxi peaches as an example to verify the effectiveness of the proposed model and solution algorithm, and to carry out case analysis. The research achievements can enrich operations management theory of fresh agricultural products, improve the quality and efficiency in the agricultural supply-side, and provide theoretical foundation and decision-making support for agricultural modernization. Besides, this research is also useful to compress intermediate links for fresh agricultural products distributed to the upside, so as to reduce the decay cost and increase farmers’ income. Therefore, this project not only has great value theoretically, but also has high importance practically.

“农-宅”直销模式下鲜果采摘与配送的协同优化，是我国农业供给侧改革和乡村振兴背景下生鲜农产品上行体系建设亟需解决的核心问题。本项目以鲜果为研究对象，聚焦“农-宅”直销模式下鲜果采摘与配送的协同优化决策，以挖掘成熟度拟合、成熟度与采摘配送决策的耦合关系为突破口，综合运用相关分析与回归分析、时空网络和Benders分解的理论方法，重点开展四个方面的研究，即鲜果成熟度拟合模型以及与采摘配送决策的耦合关系研究；鲜果采摘与配送协同优化模型研究；协同优化模型快速求解算法研究；以无锡水蜜桃为例的成果验证和案例分析研究。项目研究成果预期可丰富生鲜农产品运营管理理论，促进农业供给侧的提质增效，为实现农业现代化发展提供理论依据与决策支持，同时也有利于压缩生鲜农产品上行环节、减少生鲜损耗成本和增加农户收入，具有重要的理论价值和现实意义。

本项目以番茄、水蜜桃为例，开展“农-宅”直销模式下基于时空网络的鲜果采摘与配送协同优化研究，其核心目的是挖掘影响鲜果成熟度的主要因素以及各因素的相关性，分析鲜果成熟特性的变化规律，建立成熟度与采摘配送决策的耦合规则，揭示采摘与配送的时空关联关系，实现鲜果产地采摘与“农-宅”配送的协同优化。项目研究成果不仅有利于压缩生鲜农产品上行环节、减少生鲜损耗成本和增加农户收入，也丰富了生鲜农产品运营管理理论，为农业现代化发展提供理论依据与决策支持。.研究发现：（1）番茄成熟过程与表皮颜色、内部颜色、硬度、可溶性固形物等理化指标密切相关，综合考虑内外部生理属性能够减少相邻成熟阶段的混淆性，有效提高番茄成熟度判别效果；（2）基于聚核模糊分类的水蜜桃成熟度判别方法能够融合评价出汁率、糖度、可溶性固形物等多个成熟度指标信息，考虑成熟度的模糊性和区间性，减少指标数据映射成熟度过程的不确定性；（3）控制鲜果采后质量损失因素是降低总成本的有效方法，如冷链运输或食品保鲜技术；（4）基于成熟度的鲜果采摘与配送联合优化模型能够较好地满足鲜果配送成熟度要求，有效提高鲜果零售商服务能力与服务水平；（5）改进的遗传算法能够克服容易陷入局部最优的缺点，避免过早收敛问题，提高算法的搜索能力，高效地生成鲜果采摘与配送的协同方案；（6）基于交替方向乘子法的求解算法具有高效计算性能，能够在较短时间内解决中大规模算例。.本研究在国内外重要学术期刊上已发表学术论文8篇（其中，SCI/SSCI检索的国际期刊论文6篇，EI检索期刊论文1篇）；投稿在审期刊论文2篇；培养硕士研究生10名（已毕业6人，在读4人）。.