A smart greenhouse monitoring and control system for agarwood seedlings

Abstract

This invention provides an intelligent greenhouse monitoring and control system for agarwood seedlings. It employs a three-tiered cloud-edge-device collaborative architecture. The perception layer deploys an array of environmental parameter acquisition sensors and an array of environmental parameter control devices to achieve multi-dimensional data acquisition and device control of air temperature and humidity, soil parameters, light intensity, and carbon dioxide concentration within the greenhouse. The transmission and edge computing layer uses an STM32 series main control unit as its core, equipped with a lightweight AI model for agarwood-specific light-temperature-water-fertilizer coupling control with less than 1MB of parameters, completing local data preprocessing, real-time decision-making, and closed-loop control. The cloud platform enables big data storage, AI model iterative training, and blockchain-based full-process traceability, along with an application interface for human-computer interaction and remote management. The beneficial effects of this invention are: it enables precise multi-parameter collaborative control of the agarwood seedling growth environment, full-link automated control, autonomous operation even when offline, and full-process product traceability.

Description

Technical Field

[0001] This invention relates to the field of intelligent greenhouse technology, specifically to an intelligent greenhouse monitoring and control system for agarwood seedlings. Background Technology

[0002] Agarwood is a precious traditional Chinese medicine and natural fragrance, possessing extremely high medicinal, economic, and cultural value. In recent years, the domestic agarwood market has continued to expand. According to the "2024 Annual Report of the China Association of Traditional Chinese Medicine," the average annual growth rate of domestic agarwood demand reached 15%, but the shortage of high-quality agarwood raw materials is as high as 40%, resulting in a significant supply-demand imbalance. Agarwood seedlings have stringent requirements for their growth environment. Even slight fluctuations in parameters such as temperature, humidity, light, soil water and fertilizer, and carbon dioxide concentration can directly affect the survival rate of seedlings and the quality of later agarwood. Therefore, large-scale seedling cultivation highly depends on a stable and precise greenhouse cultivation environment.

[0003] Currently, agricultural IoT monitoring systems are gradually being promoted in China's facility agriculture sector, enabling the collection and remote viewing of basic greenhouse environmental parameters. Several mature greenhouse automation control products have also emerged internationally, such as the greenhouse control system from the Dutch company Priva and the farm management system from Argus. However, when these existing technologies are applied to agarwood seedling cultivation, numerous unavoidable shortcomings are exposed. First, existing agricultural intelligent systems generally employ general threshold control strategies, failing to establish specific control models tailored to the different growth characteristics of agarwood seedlings at various stages, such as budding, growth, and seedling development. This prevents the coupling and control of multiple factors, including light, temperature, water, fertilizer, and air, resulting in large fluctuations in greenhouse environmental parameters. Under traditional planting methods, the survival rate of agarwood seedlings has consistently been below 60%, leading to inconsistent seedling quality and severely hindering the standardized development of the agarwood industry. Second, most existing systems only complete data collection and display functions, with the sensing, data processing, and equipment control modules isolated from each other. Data cannot be deeply integrated with the execution equipment, and daily operations such as shading, ventilation, irrigation, and fertilization still rely on manual judgment and operation by farmers, resulting in low work efficiency and a high rate of human error.

[0004] Meanwhile, most current mainstream smart agriculture systems adopt a centralized cloud processing architecture, with all data processing and control decisions relying on cloud servers. However, agarwood planting bases are mostly located in remote areas, where network signals are unstable and outages are frequent. Once the network is interrupted or the cloud server malfunctions, local equipment will stop operating directly, compromising system reliability. Furthermore, existing greenhouse monitoring systems only focus on environmental control and lack a product traceability system. Data from the entire process of agarwood cultivation, from seedling to finished product distribution, cannot be documented, making it impossible for consumers to distinguish product quality, and hindering the realization of brand premiums and intellectual property protection for high-quality agarwood. In terms of practical application, imported high-end greenhouse systems cost over 500,000 yuan per greenhouse, a prohibitively high cost threshold unsuitable for small and medium-sized agarwood growers in China. Domestic general-purpose smart systems have complex operating logic, are generally difficult for older agarwood growers to learn and use, and offer a poor human-computer interaction experience, further hindering the popularization of smart technologies.

[0005] In summary, existing technologies suffer from a series of problems, including low precision in environmental control, weak intelligent linkage capabilities, insufficient ability to operate without network connectivity, lack of traceability systems, high costs, and poor usability. There is an urgent need for a smart greenhouse monitoring and control system that is specifically designed for agarwood seedling cultivation, offering high precision, high reliability, low cost, and traceability capabilities. Summary of the Invention

[0006] To address the problems in existing technologies, this invention provides an intelligent greenhouse monitoring and control system for agarwood seedlings. It employs a three-tiered cloud-edge-device collaborative architecture, consisting of a sensing layer, a transmission and edge computing layer, and a cloud platform and application layer. The sensing layer deploys an array of environmental parameter acquisition sensors and an array of environmental parameter control devices to achieve multi-dimensional data acquisition and device control of air temperature and humidity, soil parameters, light intensity, and carbon dioxide concentration within the greenhouse. The transmission and edge computing layer uses an STM32 series main control unit as its core, equipped with a lightweight AI model for agarwood-specific light-temperature-water-fertilizer coupling control with less than 1MB of parameters. This model performs local data preprocessing, real-time decision-making, and closed-loop control, while also possessing the capability to interrupt... With autonomous network operation capabilities, it can work independently for more than 72 hours when the network is interrupted; the cloud platform realizes big data storage, AI model iterative training, and blockchain full-process traceability, and together with the application operation interface, it realizes human-computer interaction and remote control, enabling precise control of multiple parameters of the agarwood seedling growth environment, full-link automated control, autonomous operation when the network is interrupted, and full-process traceability of products, improving seedling survival rate and product quality, reducing manual management costs and equipment investment costs, promoting the intelligent, standardized and large-scale development of the agarwood seedling industry, and solving the problems of low precision of environmental control in agarwood cultivation greenhouses, insufficient system intelligence, high network dependence, lack of quality traceability and high promotion costs in existing technologies.

[0007] This invention provides an intelligent greenhouse monitoring and control system for agarwood seedlings, comprising a sensing layer, a transmission and edge computing layer, and a cloud platform and application layer that cooperate with each other. The sensing layer is wirelessly connected to the transmission and edge computing layer, and the transmission and edge computing layer is wirelessly connected to the cloud platform and application layer. The sensing layer includes an environmental parameter acquisition sensor array and an environmental parameter control device array. The environmental parameter acquisition sensor array is used to collect multiple environmental parameters within the agarwood seedling greenhouse. The sensing layer is also used to receive control commands to drive the environmental parameter control device array within the greenhouse to complete environmental control actions. The transmission and edge computing layer includes a built-in human... The edge computing gateway for the artificial intelligence model includes a transmission and edge computing layer that receives environmental parameter data collected by the perception layer and generates control commands through the edge computing gateway, which are then sent to the perception layer. It can also automatically cache local data and operate autonomously when disconnected from the cloud platform and application layer. The cloud platform and application layer includes a cloud platform terminal and multiple application operation interfaces. The cloud platform and application layer stores all historical data, trains and iterates the artificial intelligence model of the edge computing gateway offline, and constructs a blockchain traceability system. The application operation interfaces provide users with data viewing, remote control, alarm notifications, and traceability query services.

[0008] In a further improvement, the sensing layer further includes a wireless transmission module. The environmental parameter acquisition sensor array includes an air temperature and humidity sensor, a soil moisture sensor, a soil conductivity sensor, a light intensity sensor, and a carbon dioxide concentration sensor, used to collect environmental parameter data within the intelligent greenhouse in real time. The wireless transmission module uses an MX02 Bluetooth module, based on BLE 5.0 technology, to transmit the environmental parameter data within the intelligent greenhouse to the transmission and edge computing layer, and to receive control commands from the transmission and edge computing layer. The environmental parameter control device array includes mutually cooperating shading nets, ventilation windows, irrigation devices, supplemental LED lights, integrated water and fertilizer machines, and heating and cooling equipment.

[0009] The present invention is further improved in that the transmission and edge computing layer also includes a local data caching module, a WiFi communication module and an alarm push module that cooperate with each other. The edge computing gateway is a main control unit using an STM32F103C8T6 microprocessor, which serves as the core processing unit of the transmission and edge computing layer. The artificial intelligence model is a lightweight AI inference engine. The lightweight AI inference engine has a built-in light-temperature-water-fertilizer coupling control model for agarwood. The parameters of the light-temperature-water-fertilizer coupling control model for agarwood are less than 1MB, and the latency of a single inference is less than 50ms.

[0010] The present invention is further improved in that the local data caching module is a non-volatile storage unit used to cache environmental parameter data and device operation logs when the network is interrupted. The transmission and edge computing layer runs autonomously for no less than 72 hours when the network is disconnected, and automatically synchronizes the cached data to the cloud platform and application layer after the network is restored.

[0011] The present invention is further improved in that the cloud platform includes a cloud data server, a deep learning model training platform, and a blockchain traceability service module. The cloud data server is equipped with a MySQL relational database and an InfluxDB time series database. The MySQL relational database is used to store business log data, and the InfluxDB time series database is used to store environmental parameter data streams from high-frequency sensors.

[0012] The present invention is further improved in that the deep learning model training platform is based on a GPU cluster, which can use the long-term accumulated agarwood growth data to complete the retraining and parameter optimization of the agarwood-specific light-temperature-water-fertilizer coupling regulation model, and distribute the updated agarwood-specific light-temperature-water-fertilizer coupling regulation model to the transmission and edge computing layer. The blockchain traceability service module is a FISCO BCOS consortium blockchain framework, which can perform hash calculations on the environmental data and agricultural operation records of the entire growth cycle of agarwood seedlings and store them on the blockchain, and generate a unique traceability code for each batch of agarwood.

[0013] The present invention is further improved in that the application operation interface includes a computer web interface, a mobile terminal APP and a WeChat mini program. The application operation interface is used for real-time data display, historical data curve query, manual control of equipment, abnormal alarm reception, greenhouse distribution visualization and traceability code scanning query.

[0014] The present invention is further improved in that the perception layer, the transmission and edge computing layer, the cloud platform and the application layer together constitute a closed-loop control process of perception-decision-execution-feedback. The perception layer collects environmental parameter data at a preset frequency and uploads it. The transmission and edge computing layer receives the environmental parameter data and generates control commands to send to the perception layer. The perception layer continuously transmits environmental parameter data back until the environmental parameter data stabilizes within the target range set by the agarwood-specific light-temperature-water-fertilizer coupling regulation model. The overall closed-loop response time is less than 30 seconds.

[0015] The present invention is further improved in that the supplementary LED light is a stepless dimming LED light with PWM pulse width modulation, the irrigation device is a water and fertilizer supply device that combines drip irrigation and micro-spraying mode, the water and fertilizer integrated machine is adapted to the irrigation device, and the fertilizer dilution ratio of the water and fertilizer integrated machine is adjustable in the range of 1:50 to 1:500.

[0016] The present invention is further improved in that the WiFi communication module is an ESP01S communication module, which complies with the IEEE 802.11b / g / n standard and the MQTT IoT communication protocol, and is used for data interaction between the transmission and edge computing layer and the cloud platform and application layer.

[0017] Compared with existing technologies, the beneficial effects of this invention are as follows: This invention provides an intelligent greenhouse monitoring and control system for agarwood seedlings. It adopts a three-level collaborative architecture of cloud-edge-device, consisting of a sensing layer, a transmission and edge computing layer, and a cloud platform and application layer. The sensing layer deploys an array of environmental parameter acquisition sensors and an array of environmental parameter control devices to achieve multi-dimensional data acquisition and device control of air temperature and humidity, soil parameters, light intensity, and carbon dioxide concentration within the greenhouse. The transmission and edge computing layer uses an STM32 series main control unit as its core, equipped with a lightweight AI model for agarwood-specific light-temperature-water-fertilizer coupling control with parameters less than 1MB, to complete local data preprocessing, real-time decision-making, and closed-loop control. It also possesses the ability to operate autonomously without network access. It can operate independently for over 72 hours during network interruptions; the cloud platform enables big data storage, AI model iterative training, and blockchain-based full-process traceability. Combined with the application interface, it facilitates human-computer interaction and remote control, achieving precise multi-parameter coordinated regulation of the agarwood seedling growth environment, automated end-to-end control, autonomous operation during network outages, and full-process product traceability. This improves seedling survival rates and product quality, reduces manual management and equipment investment costs, and promotes the intelligent, standardized, and large-scale development of the agarwood seedling industry. Furthermore, it boasts high environmental control precision, significantly enhancing seedling growth. It establishes a dedicated light-temperature-water-fertilizer coupled AI model for agarwood, abandoning the traditional single-threshold control mode to achieve dynamic coordinated regulation of multiple environmental parameters, controlling temperature fluctuations within ±1℃. Internally, water and fertilizer utilization rates are increased to over 85%, the survival rate of agarwood seedlings exceeds 80%, and yield per unit area increases by 35% to 45%, completely solving the problem of extensive environmental management in traditional greenhouses. The system features full-chain automated control with fast response times. It constructs a closed-loop control chain of "perception-decision-execution-feedback," relying on edge computing to achieve rapid local decision-making. The overall closed-loop response time is controlled within 30 seconds. The perception, decision-making, and execution modules are deeply interconnected, requiring no manual intervention and significantly reducing the workload and error rate of manual management. It also boasts strong self-regulation capabilities during network outages and high operational reliability. Lightweight AI models and local storage units are deployed at edge nodes, eliminating strong dependence on the cloud and network. The system can autonomously run for 72 hours after a network outage. After a period of time, offline data is automatically re-uploaded once the network is restored, perfectly adapting to application scenarios with unstable networks in remote planting bases; blockchain traceability empowers and enhances product added value by storing the entire agarwood seedling cultivation data on the blockchain, generating a unique traceability QR code, ensuring data immutability, realizing a visual link between the planting process and product quality, and achieving a premium rate of 50%~80% for certified products, helping agarwood brand building and quality protection; low cost and ease of use facilitate large-scale promotion; the entire system is built with domestically produced hardware, with a single greenhouse construction cost of only 120,000~180,000 yuan, reducing costs by more than 70% compared to imported high-end systems; it is equipped with lightweight interactive interfaces such as WeChat mini programs, reducing the learning cost for farmers by 70%, and adapting to the usage habits of middle-aged and elderly planting groups;Green and energy-saving with outstanding ecological benefits, the system supports dual power supply modes of photovoltaics and municipal grid power. Combined with precision drip irrigation technology, overall energy consumption is reduced by 30%, water savings reach 40%, and a single greenhouse can reduce wastewater discharge by 120 tons annually. The equipment shell uses biodegradable boards and recycled plastics, significantly reducing the carbon footprint and meeting the requirements of green and sustainable agricultural development. It solves the problems of low environmental control precision, insufficient system intelligence, high network dependence, lack of quality traceability, and high promotion costs in existing agarwood cultivation greenhouses. Attached Figure Description

[0018] To more clearly illustrate the solutions in this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0019] Figure 1 This is a schematic diagram of the intelligent greenhouse monitoring and control system for agarwood seedlings of the present invention. Figure 2 This is a schematic diagram of the intelligent greenhouse monitoring and control system for agarwood seedlings according to the present invention. Detailed Implementation

[0020] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs; the terminology used herein in the specification of the application is for the purpose of describing particular embodiments only and is not intended to limit the application; the terms “comprising” and “having” and any variations thereof in the specification, claims and foregoing description of the drawings are intended to cover non-exclusive inclusion.

[0021] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0022] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings.

[0023] like Figures 1-2As shown, this invention provides an intelligent greenhouse monitoring and control system for agarwood seedlings, comprising a sensing layer, a transmission and edge computing layer, and a cloud platform and application layer that cooperate with each other. The sensing layer and the transmission and edge computing layer are wirelessly connected, as are the transmission and edge computing layer and the cloud platform and application layer. The sensing layer is equipped with an environmental parameter acquisition sensor array and an environmental parameter control device array. The environmental parameter acquisition sensor array is used to collect multiple environmental parameters within the agarwood seedling greenhouse. The sensing layer is also used to receive control commands to drive the environmental parameter control device array within the greenhouse to complete environmental control actions. The transmission and edge computing layer is equipped with a built-in... The edge computing gateway for the artificial intelligence model receives environmental parameter data collected by the perception layer and generates control commands to be sent to the perception layer. It can also automatically cache local data and operate autonomously when disconnected from the cloud platform and application layer. The cloud platform and application layer have a cloud platform terminal and multiple application operation interfaces. The cloud platform and application layer are used to store all historical data, train and iterate the artificial intelligence model of the edge computing gateway offline, and build a blockchain traceability system. The application operation interfaces are used to provide users with data viewing, remote control, alarm reminders and traceability query services. In this embodiment, a three-tiered cloud-edge-device collaborative architecture is adopted. The perception layer deploys an array of environmental parameter acquisition sensors and an array of environmental parameter control devices to achieve multi-dimensional data acquisition and equipment control of air temperature and humidity, soil parameters, light intensity, and carbon dioxide concentration within the greenhouse. The transmission and edge computing layer uses an STM32 series main control unit as its core and is equipped with a lightweight AI model for agarwood-specific light-temperature-water-fertilizer coupling control with less than 1MB of parameters to complete local data preprocessing, real-time decision-making, and closed-loop control. The cloud platform enables big data storage, AI model iterative training, and blockchain-based full-process traceability, while the application interface enables human-computer interaction and remote management. This allows for precise multi-parameter collaborative control of the agarwood seedling growth environment, full-link automated control, autonomous operation even when offline, and full-process product traceability.

[0024] like Figures 1-2As shown, the perception layer also includes a wireless transmission module. The environmental parameter acquisition sensor array includes air temperature and humidity sensors, soil moisture sensors, soil conductivity sensors, light intensity sensors, and carbon dioxide concentration sensors, used to collect environmental parameter data inside the smart greenhouse in real time. The wireless transmission module uses an MX02 Bluetooth module, based on BLE 5.0 technology, to transmit the environmental parameter data inside the smart greenhouse to the transmission and edge computing layer, and to receive control commands from the transmission and edge computing layer. The environmental parameter control equipment array includes a shade net, ventilation windows, irrigation devices, supplemental LED lights, a water and fertilizer integrated machine, and heating and cooling equipment that work together. Among them, the supplemental LED lights are stepless dimming LED lights with PWM pulse width modulation, the irrigation devices are water and fertilizer supply devices that combine drip irrigation and micro-sprinkler modes, and the water and fertilizer integrated machine is compatible with the irrigation devices. The fertilizer dilution ratio of the water and fertilizer integrated machine can be adjusted within a range of 1:50 to 1:500. In this embodiment, the sensing layer includes an air temperature sensor, an air humidity sensor, a soil moisture sensor, a soil electrical conductivity sensor (EC sensor, used to measure the concentration of soluble salt ions in the soil to reflect soil nutrient status), a light intensity sensor (such as a BH1750 digital ambient light sensor), and a carbon dioxide concentration sensor (JW01 module). Each sensor module transmits data wirelessly via an MX02 Bluetooth module (based on BLE 5.0 technology) and a transmission and edge computing gateway. The MX02 Bluetooth module features low data rate, low power consumption, and support for multi-slave self-organizing networks, making it ideal for close-range, dense deployment of dozens of sensor nodes within a greenhouse.

[0025] like Figures 1-2 As shown, the transmission and edge computing layer also includes a local data caching module, a WiFi communication module, and an alarm push module that work together. The edge computing gateway is a main control unit using an STM32F103C8T6 microprocessor, serving as the core processing unit of the transmission and edge computing layer. The artificial intelligence model is a lightweight AI inference engine, which incorporates a light-temperature-water-fertilizer coupling control model specifically for agarwood. This model has fewer than 1MB of parameters and a single inference latency of less than 50ms. The local data caching module is a non-volatile storage unit used to cache environmental parameter data and device operation logs during network interruptions. The transmission and edge computing layer can operate autonomously for at least 72 hours without a network connection and automatically synchronizes the cached data to the cloud platform and application layer after network recovery. The WiFi communication module is an ESP01S communication module, conforming to the IEEE 802.11b / g / n standard and the MQTT IoT communication protocol, used for data interaction between the transmission and edge computing layer and the cloud platform and application layer. In this embodiment, the transmission and edge computing layer is the core of achieving real-time performance, reliability, and intelligent decision-making in the system.

[0026] This layer includes an edge computing gateway, centered on an STM32F103C8T6 (a 32-bit microcontroller based on the ARM Cortex-M3 core, a product of STMicroelectronics, hereinafter referred to as the "main control unit"), or other similar microcontrollers may be used as alternatives. The main control unit connects to the MX02 Bluetooth module via a serial communication interface, receives environmental data collected by the Bluetooth slave sensors, and sends control commands to the environmental parameter control device array.

[0027] A lightweight AI inference engine, with an artificial intelligence model deployed in the main control unit's internal memory, has a parameter size of less than 1MB (megabytes) and is used to calculate the optimal control strategy in real time based on current environmental parameters. This model is a "light-temperature-water-fertilizer coupled control model" trained on the growth characteristics of agarwood seedlings. It can comprehensively analyze multiple parameters such as light intensity, air temperature, air humidity, soil moisture, carbon dioxide, and soil conductivity, and output coordinated control commands for the environmental parameter control equipment array. The model uses quantization compression technology, with an inference latency of less than 50ms, which meets real-time control requirements.

[0028] The local data caching module, equipped with non-volatile storage units (such as SD memory cards or Flash memory chips), continuously caches sensing data and control logs during network interruptions. Once the network is restored, the cached data is automatically synchronized to the cloud platform. The system can maintain autonomous operation for at least 72 hours in a network-off state. The WiFi module (ESP01S, a low-power wireless fidelity communication module supporting the IEEE 802.11 b / g / n standard) is responsible for data communication between the edge computing gateway and the cloud platform. The alarm push module pushes alarm information to the user terminal via SMS gateway or mobile application when environmental parameters exceed preset thresholds or when device anomalies occur.

[0029] like Figures 1-2As shown, the cloud platform includes a cloud data server, a deep learning model training platform, and a blockchain traceability service module. The cloud data server is equipped with a MySQL relational database and an InfluxDB time-series database. The MySQL relational database is used to store business log data, and the InfluxDB time-series database is used to store environmental parameter data streams from high-frequency sensors. The deep learning model training platform, based on a GPU cluster, can utilize long-term accumulated agarwood growth data to retrain and optimize the parameters of a dedicated agarwood light-temperature-water-fertilizer coupling control model. The updated model is then distributed to the transmission and edge computing layers. The blockchain traceability service module uses the FISCO BCOS consortium blockchain framework. It can perform hash calculations on environmental data and agricultural operation records throughout the entire growth cycle of agarwood seedlings, store them on the blockchain, and generate a unique traceability code for each batch of agarwood. The application interface includes a computer web interface, a mobile app, and a WeChat mini-program. These interfaces are used for real-time data display, historical data curve query, manual equipment control, anomaly alarm reception, greenhouse distribution visualization, and traceability code scanning. In this embodiment, the cloud data server is deployed on Alibaba Cloud or other public cloud platforms, using a MySQL relational database to store historical environmental data, equipment operation logs, user operation records, etc.; and a time-series database (such as InfluxDB) to store environmental parameter data streams from high-frequency sensors, supporting historical data curve queries and report generation. Ordinary users can register and log in via a browser or app to view greenhouse environmental monitoring data, historical trend charts, and abnormal alarm records in real time, and can manually set thresholds. System administrators can manage users, greenhouses, and sensors through the backend management interface. The application interface supports map mode display of greenhouse distribution across various bases; clicking on a greenhouse will take you to its details page.

[0030] like Figures 1-2 As shown, the cloud perception layer, transmission and edge computing layer, cloud platform and application layer together constitute a closed-loop control process of perception-decision-execution-feedback. The perception layer collects environmental parameter data at a preset frequency and uploads it. The transmission and edge computing layer receives the environmental parameter data and generates control commands to send to the perception layer. The perception layer continuously transmits environmental parameter data back until the environmental parameter data stabilizes within the target range set by the agarwood-specific light-temperature-water-fertilizer coupling control model. The overall closed-loop response time is less than 30 seconds.

[0031] In this embodiment, the entire workflow of the intelligent greenhouse monitoring and control system is as follows: (1) Each sensor in the perception layer sends environmental parameters such as air temperature, air humidity, soil temperature, soil humidity, light intensity, carbon dioxide concentration, and soil conductivity to the edge computing gateway via the MX02 Bluetooth module at a preset acquisition frequency (e.g., once every 30 seconds).

[0032] (2) After receiving the data, the edge computing gateway performs data verification and preprocessing (including outlier removal and sliding window mean filtering), and uses the preprocessed data as input to the lightweight AI inference engine.

[0033] (3) The lightweight AI inference engine is based on the light-temperature-water-fertilizer coupling control model for agarwood. Combined with the current growth stage of the agarwood seedling, it calculates the deviation between the target value and the current value of each environmental parameter in real time and outputs the control instructions of the environmental parameter control equipment array (such as "open the ventilation window to 30% opening" and "start drip irrigation for 120 seconds").

[0034] (4) The main control unit sends control commands to the environmental parameter control equipment array through the serial interface, driving the environmental parameter control equipment array to perform actions.

[0035] (5) After execution, the sensor continuously collects and transmits environmental parameter changes, forming a closed-loop control of "perception-decision-execution-feedback" until the environmental parameters stabilize within the target range.

[0036] The workflow of the intelligent greenhouse monitoring and control system when the network is interrupted is as follows: (1) The WiFi module of the edge computing gateway continuously monitors the network connection status with the cloud platform.

[0037] (2) Once a network interruption is detected, the main control unit automatically switches to “autonomous mode”, the lightweight AI inference engine continues to use the local model to make control decisions, and all sensor data and control logs are written to the non-volatile storage unit of the local data cache module.

[0038] (3) During network interruption, the alarm push module cannot send remote alarms through the mobile network, but the system can alert on-site management personnel through local sound and light alarm devices.

[0039] (4) When the WiFi module detects that the network has been restored, the system will automatically exit the autonomous mode, upload all data during the cache period to the cloud data server in batches, and restore the remote alarm push function.

[0040] The blockchain traceability process is as follows: (1) When agarwood seedlings are transplanted into the greenhouse, the system generates a unique batch number for the batch, records the start time and initializes the traceability data record.

[0041] (2) During the entire growth cycle, at preset intervals (e.g., every hour), the system packages the current average environmental parameters, cumulative irrigation amount, fertilization records and other key data and calculates their hash values.

[0042] (3) Every preset period (e.g., daily), the set of hash values ​​in that period is used to generate a Merkle root hash. The root hash value is written into the blockchain through a smart contract to obtain the corresponding transaction hash and block height, thus completing the on-chain notarization.

[0043] (4) When agarwood products are harvested, the system encodes the complete growth data and blockchain transaction hash corresponding to the batch number into a unique traceability code, which is printed on the product packaging in the form of a QR code.

[0044] (5) After the consumer scans the QR code, the front-end application queries the cloud for the complete growth data corresponding to the traceability code and compares and verifies it with the evidence hash value on the blockchain. After confirming that the data has not been tampered with, the visualized growth data is displayed to the consumer.

[0045] As a second embodiment of the present invention, the intelligent greenhouse for agarwood seedlings in Maoming, Guangdong Province, is deployed in a standard agarwood seedling greenhouse in Maoming City, Guangdong Province. The greenhouse area is approximately 667 square meters (1 mu), cultivating approximately 2000 agarwood seedlings. The system deployment is as follows: 1. Perception Layer Configuration Six sets of air temperature and humidity sensors are evenly distributed and suspended inside the shed at heights of 1.2 meters and 2.0 meters above the ground. They employ DHT11 digital temperature and humidity sensors, with a temperature measurement accuracy of ±0.3℃ and a humidity measurement accuracy of ±2%RH (relative humidity).

[0046] Twelve soil temperature, humidity, and conductivity sensors were installed at depths of 10 cm and 20 cm in the cultivation substrate. YL-69 soil sensors and TDS01 conductivity modules were used, with a measurement range of 0-10000 μS / cm (micro-Siemens per centimeter).

[0047] Four light intensity sensors are installed on the sunny side of the ceiling below the roof. They utilize BH1750FVI digital light sensors with a measurement range of 1-65535 lux.

[0048] Two sets of carbon dioxide concentration sensors are installed in the ventilated area of ​​the shed. They employ JW01 analog carbon dioxide sensors with a measurement range of 0-5000 ppm.

[0049] 2. Configuration of Transmission and Edge Computing Layers The main control unit uses an STM32F103C8T6 microcontroller with a main frequency of 72MHz, 64KB of SRAM (static random access memory), and 512KB of Flash memory.

[0050] The Bluetooth module uses the MX02 chip (ASR Technology Co., Ltd.) to build a Bluetooth network with a capacity of up to 20 network nodes.

[0051] The WiFi module uses the ESP01S module, which communicates with the main control unit via serial port and supports MQTT protocol (Message Queuing Telemetry Transport Protocol, a lightweight IoT communication protocol based on publish / subscribe model) and cloud integration.

[0052] This lightweight AI model employs a quantized micro-neural network. The input layer receives six types of environmental parameters and one growth stage identifier, while the output layer generates control signals for five types of actuators. The model firmware is approximately 850KB in size, and a single inference operation takes about 38ms. The model training dataset is derived from historical data on agarwood seedling cultivation in Maoming and experimental data from the Tropical Crops Research Institute, with a total sample size of approximately 50,000 groups.

[0053] 3. Configuration of environmental parameter control equipment array: Shading net, driven by an electric film roller, supports 0%-100% opening adjustment.

[0054] Ventilation window, electrically driven, supports opening / closing and proportional opening control.

[0055] The irrigation system combines drip irrigation and micro-sprinkler irrigation, with solenoid valves controlling the on / off state and duration, and a water pump with a rated flow rate of 3 m³ / h.

[0056] LED supplemental lighting, full-spectrum LED plant supplemental lighting, 100W / lamp, 16 lamps in total, supports PWM (pulse width modulation) stepless dimming.

[0057] The water and fertilizer integrated machine features a proportional fertilizer injection pump that supports an adjustable dilution ratio from 1:50 to 1:500 and can be linked with the irrigation system.

[0058] 4. Cloud Platform and Application Layer Configuration Cloud server configuration: Alibaba Cloud ECS server, 4 cores, 8GB memory, 100GB cloud storage, 10Mbps bandwidth.

[0059] The blockchain platform uses the FISCO BCOS 2.0 consortium blockchain, deploying 4 consensus nodes (1 node owned by the project team and 3 nodes held by partner institutions). The consensus algorithm used is PBFT (Practical Byzantine Fault Tolerance), and the single-chain transaction throughput is approximately 2000 TPS (transactions per second).

[0060] It supports Android and iOS systems and is developed using the Android Studio platform and WeChat developer tools. The interface includes four core modules: real-time data dashboard, historical curve query, remote device control, and alarm information list.

[0061] As can be seen from the above, this invention provides an intelligent greenhouse monitoring and control system for agarwood seedlings. It adopts a three-level collaborative architecture of cloud-edge-device, consisting of a perception layer, a transmission and edge computing layer, and a cloud platform and application layer. The perception layer deploys an array of environmental parameter acquisition sensors and an array of environmental parameter control devices to achieve multi-dimensional data acquisition and device control of air temperature and humidity, soil parameters, light intensity, and carbon dioxide concentration within the greenhouse. The transmission and edge computing layer uses an STM32 series main control unit as its core, equipped with a lightweight AI model for agarwood-specific light-temperature-water-fertilizer coupling control with parameters less than 1MB. This model completes local data preprocessing, real-time decision-making, and closed-loop control, while also possessing autonomous operation capability during network outages, allowing it to work independently for 7 days when the network is interrupted. Over 2 hours; the cloud platform enables big data storage, AI model iterative training, and blockchain full-process traceability. Combined with the application interface, it allows for human-computer interaction and remote control, achieving precise multi-parameter coordinated regulation of the agarwood seedling growth environment, full-chain automated control, autonomous operation even when offline, and full-process product traceability. This improves seedling survival rate and product quality, reduces manual management and equipment investment costs, and promotes the intelligent, standardized, and large-scale development of the agarwood seedling industry. Furthermore, the high precision of environmental regulation significantly enhances seedling growth. It establishes a dedicated light-temperature-water-fertilizer coupled AI model for agarwood, abandoning the traditional single-threshold control mode to achieve dynamic coordinated regulation of multiple environmental parameters, keeping temperature fluctuations within ±1℃ and improving water and fertilizer utilization. The survival rate of agarwood seedlings has been increased to over 85%, and the yield per unit area has increased by 35% to 45%, completely solving the problem of extensive environmental management in traditional greenhouses. The system features full-chain automated control with fast response times. It constructs a closed-loop control chain of "perception-decision-execution-feedback," relying on edge computing to achieve rapid local decision-making. The overall closed-loop response time is controlled within 30 seconds. The perception, decision-making, and execution modules are deeply interconnected, requiring no manual intervention and significantly reducing the workload and error rate of manual management. The system also boasts strong self-regulation capabilities during network outages and high operational reliability. Lightweight AI models and local storage units are deployed at edge nodes, eliminating strong dependence on the cloud and network. The system can operate autonomously for over 72 hours after a network outage. After network recovery, offline data is automatically re-uploaded, perfectly adapting to application scenarios with unstable networks in remote planting bases; blockchain traceability empowers and enhances product added value by storing the entire agarwood seedling cultivation data on the blockchain, generating a unique traceability QR code, ensuring data immutability, and realizing a visual link between the planting process and product quality. Certified products can achieve a premium rate of 50%~80%, contributing to agarwood brand building and quality protection; low cost and ease of use facilitate large-scale promotion. The entire system is built with domestically produced hardware, with a single greenhouse construction cost of only 120,000~180,000 yuan, reducing costs by more than 70% compared to imported high-end systems; it is equipped with lightweight interactive interfaces such as WeChat mini-programs, reducing the learning cost for farmers by 70% and adapting to the usage habits of middle-aged and elderly planting groups;Green and energy-saving with outstanding ecological benefits, the system supports dual power supply modes of photovoltaics and municipal grid power. Combined with precision drip irrigation technology, overall energy consumption is reduced by 30%, water savings reach 40%, and a single greenhouse can reduce wastewater discharge by 120 tons annually. The equipment shell uses biodegradable boards and recycled plastics, significantly reducing the carbon footprint and meeting the requirements of green and sustainable agricultural development. It solves the problems of low environmental control precision, insufficient system intelligence, high network dependence, lack of quality traceability, and high promotion costs in existing agarwood cultivation greenhouses.

[0062] The specific embodiments described above are preferred embodiments of the present invention and are not intended to limit the specific scope of the present invention. The scope of the present invention includes, but is not limited to, these specific embodiments. All equivalent changes made in accordance with the present invention are within the protection scope of the present invention.

Claims

1. A smart greenhouse monitoring and control system for agarwood seedlings, characterized in that: The system comprises a sensing layer, a transmission and edge computing layer, and a cloud platform and application layer that work together. The sensing layer is wirelessly connected to the transmission and edge computing layer, and the transmission and edge computing layer is wirelessly connected to the cloud platform and application layer. The sensing layer includes an array of environmental parameter acquisition sensors and an array of environmental parameter control devices. The environmental parameter acquisition sensor array is used to collect multiple environmental parameters within the agarwood seedling greenhouse. The sensing layer also receives control commands to drive the environmental parameter control device array within the greenhouse to perform environmental control actions. The transmission and edge computing layer includes an edge computing gateway with a built-in artificial intelligence model. The transmission and edge computing layer receives environmental parameter data collected by the perception layer and generates control commands through the edge computing gateway, which are then sent to the perception layer. It can also automatically cache local data and operate autonomously when disconnected from the cloud platform and application layer. The cloud platform and application layer includes a cloud platform terminal and multiple application operation interfaces. It stores all historical data, trains and iterates the artificial intelligence model of the edge computing gateway offline, and constructs a blockchain traceability system. The application operation interfaces provide users with data viewing, remote control, alarm notifications, and traceability query services.

2. The intelligent greenhouse monitoring and control system for agarwood seedlings according to claim 1, characterized in that: The sensing layer also includes a wireless transmission module. The environmental parameter acquisition sensor array includes an air temperature and humidity sensor, a soil moisture sensor, a soil conductivity sensor, a light intensity sensor, and a carbon dioxide concentration sensor, used to collect environmental parameter data inside the smart greenhouse in real time. The wireless transmission module uses an MX02 Bluetooth module, based on BLE 5.0 technology, to transmit the environmental parameter data inside the smart greenhouse to the transmission and edge computing layer, and to receive control commands from the transmission and edge computing layer. The environmental parameter control device array includes mutually cooperating shading nets, ventilation windows, irrigation devices, supplemental LED lights, integrated water and fertilizer machines, and heating and cooling equipment.

3. The intelligent greenhouse monitoring and control system for agarwood seedlings according to claim 2, characterized in that: The transmission and edge computing layer also includes a local data caching module, a WiFi communication module, and an alarm push module that work together. The edge computing gateway is a main control unit using an STM32F103C8T6 microprocessor, serving as the core processing unit of the transmission and edge computing layer. The artificial intelligence model is a lightweight AI inference engine. The lightweight AI inference engine has a built-in light-temperature-water-fertilizer coupling control model specifically for agarwood. The parameters of the light-temperature-water-fertilizer coupling control model for agarwood are less than 1MB, and the latency of a single inference is less than 50ms.

4. The intelligent greenhouse monitoring and control system for agarwood seedlings according to claim 3, characterized in that: The local data caching module is a non-volatile storage unit used to cache environmental parameter data and device operation logs when the network is interrupted. The transmission and edge computing layer runs autonomously for no less than 72 hours when the network is down, and automatically synchronizes the cached data to the cloud platform and application layer after the network is restored.

5. The intelligent greenhouse monitoring and control system for agarwood seedlings according to claim 4, characterized in that: The cloud platform includes a cloud data server, a deep learning model training platform, and a blockchain traceability service module. The cloud data server is equipped with a MySQL relational database and an InfluxDB time-series database. The MySQL relational database is used to store business log data, and the InfluxDB time-series database is used to store environmental parameter data streams from high-frequency sensors.

6. The intelligent greenhouse monitoring and control system for agarwood seedlings according to claim 5, characterized in that: The deep learning model training platform is based on a GPU cluster and can use long-term accumulated agarwood growth data to complete the retraining and parameter optimization of the agarwood-specific light-temperature-water-fertilizer coupling regulation model. The updated agarwood-specific light-temperature-water-fertilizer coupling regulation model is then distributed to the transmission and edge computing layer. The blockchain traceability service module is based on the FISCO BCOS consortium blockchain framework, which can perform hash calculations on the environmental data and agricultural operation records of the entire growth cycle of agarwood seedlings and store them on the blockchain for evidence, and generate a unique traceability code for each batch of agarwood.

7. The intelligent greenhouse monitoring and control system for agarwood seedlings according to claim 6, characterized in that: The application operation interface includes a computer web interface, a mobile terminal APP, and a WeChat mini program. The application operation interface is used for real-time data display, historical data curve query, manual equipment control, abnormal alarm reception, greenhouse distribution visualization, and traceability code scanning query.

8. The intelligent greenhouse monitoring and control system for agarwood seedlings according to claim 7, characterized in that: The perception layer, the transmission and edge computing layer, and the cloud platform and application layer together constitute a closed-loop control process of perception-decision-execution-feedback. The perception layer collects and uploads environmental parameter data at a preset frequency. The transmission and edge computing layer receives the environmental parameter data and generates control commands to send to the perception layer. The perception layer continuously transmits environmental parameter data back until the environmental parameter data stabilizes within the target range set by the agarwood-specific light-temperature-water-fertilizer coupling control model. The overall closed-loop response time is less than 30 seconds.

9. The intelligent greenhouse monitoring and control system for agarwood seedlings according to claim 8, characterized in that: The supplemental LED light is a stepless dimming LED light with PWM pulse width modulation. The irrigation device is a water and fertilizer supply device that combines drip irrigation and micro-spraying modes. The integrated water and fertilizer machine is compatible with the irrigation device. The fertilizer dilution ratio of the integrated water and fertilizer machine is adjustable from 1:50 to 1:

500.

10. The intelligent greenhouse monitoring and control system for agarwood seedlings according to claim 9, characterized in that: The WiFi communication module is an ESP01S communication module, which conforms to the IEEE 802.11b / g / n standard and the MQTT IoT communication protocol, and is used for data interaction between the transmission and edge computing layer and the cloud platform and application layer.

Images