首页|期刊导航|蔬菜|设施蔬菜生长模型研究进展与未来趋势

设施蔬菜生长模型研究进展与未来趋势OA

Research Progress and Future Trends in Growth Models of Protected Vegetables

中文摘要英文摘要

基于作物生长模型在设施蔬菜提质增效、资源利用、环境调控等方面作用日益突出,本文系统梳理了设施蔬菜生长模型演进历程,对比阐释机理模型、经验统计模型、数据驱动模型及包含数字孪生在内的多源耦合模型4类框架的建模原理、优劣与适用场景,从产量品质预测、栽培管理决策、作物表型解析三大维度总结现有模型的产业化应用场景;剖析当前领域存在的模型跨场景通用性差、参数标定与验证成本高、模型与温室环控系统耦合程度低等突出问题;提出4项发展对策,包括搭建标准化建模与多源异构数据共享体系、依托数字孪生实现模型与温室环控深度融合、融合多维高通量表型技术与人工智能开展多目标协同优化及构建跨学科产学研用一体化创新体系;最后展望行业发展趋势,指出设施蔬菜生长模型将逐步进化为自进化智能体、拓展至设施—气候—能源全域耦合、升级为数字孪生全自主闭环管控核心,相关结论可为智慧农业中设施蔬菜模型的标准化与智能化应用提供理论与技术参考.

Given the increasingly prominent role of crop growth models in improving quality-and-efficiency,enhancing resource utilization,and regulating greenhouse environments for protected vegetables,this paper systematically reviews the evolution of protected-vegetable growth models.It comparatively illustrates the modeling principles,advantages,limitations,and applicable scenarios of four model frameworks:mechanistic models,empirical-statistical models,data-driven models,and multi-source coupled models including digital twins.Industrial application scenarios of existing models are summarized from three dimensions:yield and quality prediction,cultivation-management decision-making,and crop phenotyping analysis.Prominent challenges in this field are analyzed,including poor cross-scenario generalization of models,high costs for parameter calibration and validation,and low coupling degree between models and greenhouse environmental control systems.Four development countermeasures are proposed:establishing a standardized modeling system and a sharing platform for multi-source heterogeneous data,realizing deep integration of models and greenhouse environmental control via digital twins,conducting multi-objective collaborative optimization by combining multi-dimensional high-throughput phenotyping technology and artificial intelligence,and building an interdisciplinary innovation system integrating industry,university,research and end-user application.Finally,future industry trends are prospected.It is pointed out that protected-vegetable growth models will gradually evolve into self-evolving intelligent agents,expand to full-domain coupling of facility-climate-energy systems,and serve as the core for fully autonomous closed-loop management and control based on digital twins.This results can provide theoretical and technical references for the standardized and intelligent application of growth models for protected vegetables in smart agriculture.

徐德敏;王华阳;王亮;张嘉鹏;马韫韬;朱晋宇

中国农业大学土地科学与技术学院,北京 100193||中国农业科学院蔬菜花卉研究所/蔬菜生物育种全国重点实验室,北京 100081中国农业大学土地科学与技术学院,北京 100193中国融通农业发展集团有限公司,北京 100036中国农业大学土地科学与技术学院,北京 100193中国农业大学土地科学与技术学院,北京 100193中国农业科学院蔬菜花卉研究所/蔬菜生物育种全国重点实验室,北京 100081

农业科技

设施蔬菜作物生长模型机理模型智能驱动数字孪生表型监测人工智能

protected vegetablecrop growth modelmechanistic modelintelligence-drivendigital twinphenotypic monitoringartificial intelligence

《蔬菜》 2026 (8)

81-91,11

国家重点研发计划项目(2023YFD2000600)国家自然科学基金面上项目(32372799、32271987)北京市乡村振兴农业科技项目(NY2501010200).

10.26924/j.cnki.sczz.2026.8006

评论