| 引用本文: | 史凯丽,田有国,孟远夺,杨积忠,吴优,王书峰,郑文刚.肥料全产业链大数据分析应用中心设计与实现[J].中国农业信息,2025,37(4):107-120 |
| |
|
| 摘要: |
| 【目的】 肥料是保障国家粮食安全和农业绿色发展的关键生产资料，当前肥料产业存在数据分散、标准不一、挖掘应用不足等问题，亟须构建覆盖全产业链的大数据平台以提升行业管理与服务水平。【方法】 文章通过云计算、人工智能等技术，系统建设覆盖肥料生产、登记、流通、贸易、施用、质量监管和施肥指导全链条的大数据分析应用中心，集成专题数据库、分析预测与决策服务子系统、大数据公共服务子系统三大核心部分，实现数据整合、智能分析与多元服务。【结果】 （1）平台构建的基于长短期记忆网络（LSTM）的肥料价格指数预测模型，实现了主要肥料品种价格趋势的高精度预测（R2为0.965）。（2）平台构建的基于养分驱动的玉米、小麦和水稻生产潜力智能预测模型，其R2均高于0.940，展现出优异的拟合效果与跨作物、跨区域的泛化能力，为肥料产业生产与施用提供科学化决策依据。【结论】 建设肥料全产业链大数据应用中心对推进农业绿色发展、提升粮食安全保障能力具有重要现实意义。 |
| 关键词: 肥料大数据 系统设计 供需预测 价格分析 分析模型 |
| DOI：10.12105/j.issn.1672-0423.20250408 |
| 分类号: |
| 基金项目:云南省重大科技专项计划项目“土壤养分大数据构建及智能化全程服务关键技术研究与应用”（202202AE090013）；北京市农林科学院科研创新平台建设项目“农业物联网技术国家地方联合工程实验室建设”（PT2025-25）；国家现代农业产业技术体系“智慧农业共性技术创新团队”(CARS-54)；北京市农林科学院重大科技成果培育项目“智慧灌溉关键技术装备及产业化” |
|
| Design and application of a big data analysis and application center for the entire fertilizer industry chain |
|
Shi Kaili1,2, Tian Youguo3, Meng Yuanduo3, Yang Jizhong4, Wu You3, Wang Shufeng1,5, Zheng Wengang1,2
|
|
1Nongxin Technology（Beijing）Co.，Ltd.， Beijing 100097，China;2Intelligent Equipment Research Center，Beijing Academy of Agriculture and Forestry Sciences，Beijing 100097，China;3National Agricultural Technology Extension and Service Center，Beijing 100125，China;4Yunnan Supply and Marketing Industry Investment Co.，Ltd.，Kunming 650051，Yunnan，China;5Information Technology Research Center，Beijing Academy of Agriculture and Forestry Sciences，Beijing 100097，China
|
| Abstract: |
| [Purpose] Fertilizers are key means of production for ensuring national food security and promoting green agricultural development. Currently，the fertilizer industry faces challenges such as fragmented data，inconsistent standards，and insufficient data mining and application. There is an urgent need to establish a big data platform covering the entire industrial chain to enhance industry management and service levels.[Method] Through the application of technologies such as cloud computing and artificial intelligence，a big data analytics and application center covering the entire chain of fertilizer production，registration，circulation，trade，application，quality supervision，and fertilization guidance was established. It integrated three core components：a specialized database，an analytical prediction and decision support subsystem，and a big data public service subsystem，enabling data integration，intelligent analysis，and diversified services.[Result] （1）The long short-term memory（LSTM）-based fertilizer price index prediction model developed by the platform achieved high-precision forecasting of price trends for major fertilizer varieties，with a R2 of 0.965. （2）The nutrient-driven intelligent prediction model for corn，wheat，and rice production potential constructed by the platform demonstrated excellent fitting performance and strong generalization capabilities across crops and regions，with R2 values all above 0.940，providing scientific decision-making support for fertilizer industry production and application.[Conclusion] The study demonstrates that establishing a big data application center for the entire fertilizer industry chain holds significant practical importance for advancing green agricultural development and enhancing food security capabilities. |
| Key words: fertilizer big data system design supply and demand prediction price analysis analytical model |
