基于端边云协同与双模控制的多棚果蔬管控系统设计与实现

Design and Implementation of a Multi-Span Vegetable and Fruit Greenhouse Management and Control System Based on Device-Edge-Cloud Collaboration and Dual-Mode Control

摘要:

针对多栋果蔬大棚管控中本地自动控制与远程手动操作难以协同、视频监控与物联网数据相互孤立等问题，提出了基于端边云协同架构、以PLC为边缘核心、融合萤石云开放平台与阿里云物联网平台的多源数据集成与双模协同控制方法，设计并实现了一套多棚果蔬综合管控系统。系统以PLC为边缘处理核心，融合环境传感器、虫情测报灯、放风机、卷帘机等实现本地闭环控制，通过FH5111B串口服务器部署采集脚本完成设备数据上云，自研业务后台集成萤石云开放平台API，将实时视频与传感器时序数据在应用层融合，并提供移动APP、Web管理端与DataV大屏三端交互。从系统集成角度设计了本地自动与远程手动双模协同控制、多源异构数据松耦合融合、基于低代码大屏的作物差异化可视化等方案。系统在山西沃勃农业发展有限公司负责运营的多栋果蔬大棚部署应用，以番茄与青椒种植为典型场景，经超过3年的连续运行验证，人工巡检频次降低约70%，异常响应时间缩短至分钟级，设备在线率保持在99.5%以上。该系统为多棚果蔬种植的远程协同、精准调控与精细化管理提供了一套可复制的工程方案。

Abstract:

To address challenges such as the difficulty in coordinating local automatic control with remote manual operation and the isolation between video surveillance and IoT data in the management of multiple vegetable and fruit greenhouses, a multi-source data integration and dual-mode coordinated control method based on a device-edge-cloud collaborative architecture, with a PLC as the edge core and integrating the EZVIZ Cloud Open Platform and the Alibaba Cloud IoT Platform, was proposed, and a comprehensive multi-greenhouse management and control system was designed and implemented. With a PLC as the edge processing core, the system integrates environmental sensors, insect-trapping lamps, ventilation fans, and roller shutter motors to achieve local closed-loop control. Device data are uploaded to the cloud through data acquisition scripts deployed on an FH5111B serial server. A self-developed backend integrates the APIs of the EZVIZ Cloud Open Platform to merge real-time video streams with sensor time-series data at the application layer, and provides interaction through three terminals: a mobile APP, a web management console, and DataV large-screen displays. From the perspective of system integration, schemes were designed for dual-mode coordinated control combining local automatic and remote manual operation,loosely coupled fusion of multi-source heterogeneous data, and crop-specific visualization using low-code dashboards. The system has been deployed in multiple greenhouses operated by Shanxi Wobo Agricultural Development Co., Ltd., with tomato and bell pepper cultivation as typical application scenarios. After more than three years of continuous operation, the system has reduced the frequency of manual inspections by approximately 70%, shortened the response time to abnormal conditions to the minute level, and maintained the equipment online rate above 99.5%.The system offers a replicable engineering solution for remote collaboration, precise regulation, and refined management in multi-span vegetable and fruit cultivation.