AloT-based container modular sprout vegetable intelligent cultivation device

Abstract

This invention discloses an AIoT-based containerized modular intelligent sprout cultivation device. The device collects environmental data in real time through multiple sensors deployed within the container, which is then uploaded to the platform layer via an edge computing gateway. The platform layer uses an LSTM predictive control module to analyze the data, predict environmental trends, and generate optimized environmental control and growth agent dispensing strategies, which are then sent to the edge gateway to drive actuators. Simultaneously, an intelligent circulating water system integrates multi-stage treatment and recycling devices to achieve efficient water resource recycling. This invention solves the problems of lagging environmental control, low resource utilization efficiency, and poor coordination among subsystems in existing plant factory systems through integrated closed-loop intelligent regulation of sensing, prediction, and control, achieving precise and automated management of the entire plant growth cycle.

Description

Technical Field

[0001] This invention relates to the field of intelligent agricultural cultivation technology, specifically to an AIoT-based containerized modular intelligent sprout cultivation device. Background Technology

[0002] Plant factories, as a highly efficient agricultural model for year-round continuous crop production through precise control of environmental factors, have received widespread attention in recent years. Among them, modular plant factories utilizing converted shipping containers have shown unique advantages in specific scenarios due to their flexible deployment and lack of geographical and climatic limitations. However, existing container-based plant factory systems still have significant shortcomings in terms of intelligence and system integration. The core issues lie in the lag in environmental control and the isolated operation of various subsystems.

[0003] Specifically, existing control systems mostly employ control logic based on preset thresholds or simple feedback. For example, when a sensor detects that the temperature exceeds a set value, a fan is activated to cool the plant. This reactive control mode has inherent delays and cannot cope with the time-varying, nonlinear, and high-inertia characteristics of environmental parameters. This results in the environment inside the enclosure constantly being in an unstable state of "fluctuation-regulation-re-fluctuation," making it difficult to maintain it within the optimal range for crop growth. Furthermore, subsystems such as environmental regulation, water and fertilizer management, and supplemental lighting are often managed by independent controllers, lacking coordination. For example, the water and fertilizer irrigation system fails to correlate with changes in environmental temperature and humidity, or the efficiency of plant transpiration, leading to low efficiency in water and nutrient utilization. This "information silo" phenomenon prevents the entire system from functioning as a unified organism for coordinated optimization, limiting crop yield and quality improvements and causing energy and resource waste. Therefore, there is an urgent need for an intelligent control scheme capable of forward-looking prediction and coordinated optimization of multiple subsystems to fundamentally solve the problems of lagging environmental regulation and poor system coordination, thereby achieving efficient, precise, and automated plant production. Summary of the Invention

[0004] To address the shortcomings of existing technologies, the purpose of this invention is to provide an AIoT-based containerized modular intelligent sprout cultivation device. Through integrated closed-loop intelligent regulation of sensing, prediction, and control, it solves the problems of lagging environmental regulation, low resource utilization efficiency, and poor coordination among subsystems in existing plant factory systems, and achieves precise and automated management of the entire plant growth cycle.

[0005] To achieve the above objectives, the embodiments of this invention provide the following technical solutions:

[0006] This application provides an AIoT-based modular intelligent sprout cultivation device for containers, comprising: a modular container unit, the inner wall of which is provided with an insulation layer and a moisture-proof coating, and the internal space is divided into multiple cultivation areas by a detachable layered loading rack; an environmental sensing and execution unit, including at least one temperature sensor, humidity sensor, CO2 sensor, EC sensor, and pH sensor disposed in the cultivation area, as well as a variable frequency fan, humidifier, LED supplemental lighting device, and growth agent dispensing device for controlled environmental regulation; an edge computing gateway, deployed inside the modular container unit, communicating with the environmental sensing and execution unit, for collecting sensor data and issuing control commands; and an intelligent circulating water system, including a main circulation pipeline and a condensate recovery branch connected to the main circulation pipeline, wherein a sand filter unit, a UV disinfection unit, and a reverse osmosis unit are sequentially connected in series on the main circulation pipeline. The pipeline is equipped with flow sensors and electric valves. A condensate collection device located on the top of the container is installed at the inlet of the condensate recovery branch. The outlet of the condensate collection device is connected to the main circulation pipeline upstream of the reverse osmosis unit. The outlet of the growth agent dispensing device is connected to the main circulation pipeline downstream of the reverse osmosis unit via a peristaltic pump. The platform layer, communicatively connected to the edge computing gateway, includes: a data storage module for storing time-series environmental data and device operation logs uploaded by the edge computing gateway; an LSTM predictive control module for receiving the time-series environmental data, predicting future plant growth environment trends, and generating an optimized control strategy based on the prediction results and a preset growth model, including temperature setpoints, humidity setpoints, and growth ratio parameters; and a device control module for distributing the optimized control strategy to the edge computing gateway to drive the environmental sensing and execution unit and the peristaltic pump.

[0007] Furthermore, the optimized control strategy for generating the growth agent ratio parameters specifically includes: inputting historical and current temperature, humidity, and CO2 concentration data into a trained LSTM network, and outputting predicted values ​​of environmental parameters for the future time period; comparing the predicted values ​​with a preset optimal parameter range for plant growth; if the predicted values ​​deviate from the optimal parameter range, calculating adjustment commands for driving the variable frequency fan or humidifier using a PID algorithm, which constitute the temperature setpoint and humidity setpoint; and simultaneously, dynamically adjusting the growth agent ratio parameters based on the predicted plant growth environment trend and plant growth cycle for the future time period.

[0008] Furthermore, the growth agent dispensing device includes multiple peristaltic pumps connected in parallel, each peristaltic pump being connected to an independent growth agent stock solution container. The device control module controls the start, stop, and rotation speed of the different peristaltic pumps to mix and form a nutrient solution with a target EC / pH value.

[0009] Furthermore, each layer of the layered loading rack is equipped with an independent water pipe interface and an electrical interface, and the water interface is connected to the intelligent circulating water system via a quick connector.

[0010] Furthermore, the edge computing gateway communicates with the platform layer via Modbus or MQTT protocols.

[0011] Furthermore, the device also includes a traceability management module, which is configured to generate a unique traceability code for each container and associate the traceability code with the full-cycle environmental data, growth agent dosing records and quality inspection records of the corresponding container.

[0012] Furthermore, the insulation layer is a polyurethane insulation layer formed by on-site foaming.

[0013] Accordingly, this application also provides an AIoT-based intelligent cultivation method for modular containerized sprouts. The method includes: continuously collecting environmental data of each cultivation area within the modular container unit using sensors in the environmental sensing and execution unit; uploading the environmental data to the platform layer via the edge computing gateway; processing the environmental data through the LSTM predictive control module of the platform layer, predicting environmental trends, and generating optimized control strategies and growth agent dispensing strategies; distributing the environmental control strategies and growth agent dispensing strategies to the edge computing gateway via the device control module in the platform layer; and driving the variable frequency fan, humidifier, LED supplemental lighting device in the environmental sensing and execution unit, and the peristaltic pump in the intelligent circulating water system via the edge computing gateway to perform corresponding control actions.

[0014] The beneficial effects of this invention are as follows: By integrating multifunctional units into modular containers, the advantages of standardized transportation and rapid deployment are combined to achieve the integration of environmental perception, intelligent control, and recycled water utilization; the intelligent recycled water system 4 greatly improves water resource utilization; LSTM prediction combined with precise control allows environmental parameters to be adapted to the needs of seedlings; the traceability module optimizes the growth status of seedlings; and the full-cycle data recording enables traceable management of the cultivation process. Attached Figure Description

[0015] Figure 1 This application provides a schematic diagram of the module connection of a containerized modular intelligent sprout cultivation device based on AIoT.

[0016] Figure 2 This application provides a structural cross-sectional schematic diagram of an AIoT-based containerized modular intelligent sprout cultivation device.

[0017] Reference numerals: 1-Modular container unit, 11-Insulation layer, 12-Moisture-proof coating, 13-Layered loading rack, 2-Environmental sensing and execution unit, 3-Edge computing gateway, 4-Intelligent circulating water system, 41-Main circulation pipeline, 42-Condensate recovery branch, 43-Sand filter unit, 44-UV disinfection unit, 45-Reverse osmosis unit, 46-Condensate collection device, 47-Flow sensor, 48-Electric valve, 5-Platform layer, 6-Data storage module, 7-LSTM predictive control module, 8-Equipment control module. Detailed Implementation

[0018] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for illustration and explanation only and are not intended to limit the scope of the present invention.

[0019] In this invention, the terms "system" and "network" are used interchangeably. "Multiple" refers to two or more; therefore, in this invention, "multiple" can also be understood as "at least two." "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone. Additionally, the character " / ", unless otherwise specified, generally indicates that the preceding and following related objects have an "or" relationship. Furthermore, it should be understood that in the description of this invention, terms such as "first" and "second" are used only for descriptive purposes and should not be construed as indicating or implying relative importance or order.

[0020] like Figure 1-2As shown, this application provides an AIoT-based containerized modular intelligent sprout cultivation device, comprising: a modular container unit 1, the inner wall of which is provided with an insulation layer 11 and a moisture-proof coating 12, and the internal space is divided into multiple cultivation areas by a detachable layered loading rack 13; an environmental sensing and execution unit 2, including at least one temperature sensor, humidity sensor, CO2 sensor, EC sensor and pH sensor installed in the cultivation area, as well as a variable frequency fan, humidifier, LED supplemental lighting device and growth agent dispensing device for controlled environmental regulation; an edge computing gateway 3, deployed inside the modular container unit 1, communicating with the environmental sensing and execution unit 2, for collecting sensor data and issuing control commands; and an intelligent circulating water system 4, including a main circulation pipeline 41 and a condensate recovery branch 42 connected to the main circulation pipeline 41, wherein a sand filter unit 43, a UV disinfection unit 44 and a reverse osmosis unit 45 are connected in series on the main circulation pipeline 41. A flow sensor 47 and an electric valve 48 are installed on the duct 41. A condensate collection device 46 located on the top of the container is installed at the inlet end of the condensate recovery branch 42. The outlet end of the condensate collection device 46 is connected to the main circulation pipeline 41 upstream of the reverse osmosis unit 45. The outlet of the growth agent dispensing device is connected to the main circulation pipeline 41 downstream of the reverse osmosis unit 45 via a peristaltic pump. The platform layer 5 is communicatively connected to the edge computing gateway 3 and includes: a data storage module 6 for storing time-series environmental data and device operation logs uploaded by the edge computing gateway 3; an LSTM predictive control module 7 for receiving the time-series environmental data, predicting the plant growth environment trend in the future time period, and generating an optimized control strategy including temperature setpoint, humidity setpoint, and growth ratio parameters based on the prediction results and a preset growth model; and a device control module 8 for sending the optimized control strategy to the edge computing gateway 3 to drive the environmental sensing and execution unit 2 and the peristaltic pump.

[0021] In another possible embodiment, a standard shipping container is first selected as the main body. A polyurethane insulation layer 11 is foamed on its inner wall and a moisture-proof coating 12 is sprayed on. Then, a detachable layered loading rack 13 divides the interior into multiple cultivation areas, each with its own independent water and electricity interface. Subsequently, temperature, humidity, CO2, EC, and pH sensors are deployed in each cultivation area, along with variable frequency fans, humidifiers, LED supplemental lighting equipment, and growth agent dispensing devices, forming an environmental sensing and execution unit 2. Next, an edge computing gateway 3 is deployed inside the container to establish communication connections with the aforementioned sensors and execution devices, responsible for collecting and temporarily storing sensor data. Then, an intelligent circulating water system 4 is constructed, with the main circulation pipeline 41 sequentially connecting a sand filter unit 43, a UV disinfection unit 44, and a reverse osmosis unit 45, and a flow meter is installed on the main pipeline. Sensor 47 and electric valve 48 are installed, and a condensate collection device 46 is installed on the top of the container. The collected condensate is connected to the main circulation pipeline 41 upstream of the reverse osmosis unit 45. The outlet of the growth agent dispensing device is connected to the main circulation pipeline 41 downstream of the reverse osmosis unit 45 through a peristaltic pump. Finally, a platform layer 5 is built, which communicates with the edge computing gateway 3 through Modbus or MQTT protocol. The data stream storage module of the platform layer 5 receives and stores the time-series environmental data and equipment operation logs uploaded by the edge gateway in real time. The LSTM predictive control module 7 predicts future environmental trends based on these data and generates temperature setpoints, humidity setpoints and growth agent ratio parameters by combining the preset optimal parameters for seedling growth. The equipment control module 8 sends these optimization strategies to the edge computing gateway 3, and the gateway drives the execution unit to complete the environmental regulation and growth agent dispensing actions.

[0022] By integrating multifunctional units into modular containers, the system combines the advantages of standardized transportation and rapid deployment to achieve integrated environmental sensing, intelligent control, and water recycling. The intelligent water recycling system significantly improves water resource utilization. LSTM prediction combined with precise control allows environmental parameters to be adapted to the needs of seedlings. The traceability module optimizes the growth status of seedlings, while full-cycle data recording enables traceable management of the cultivation process.

[0023] Traditional seedling cultivation environmental control mostly uses a fixed threshold trigger mode, which can only passively adjust parameters and cannot predict environmental changes, resulting in large parameter fluctuations, unstable seedling growth, and poor consistency in cultivation quality.

[0024] In the embodiments of this application, the optimized control strategy for generating the growth agent ratio parameters specifically includes: inputting historical and current temperature, humidity, and CO2 concentration data into a trained LSTM network, and outputting predicted values ​​of environmental parameters for future time periods; comparing the predicted values ​​with a preset optimal parameter range for plant growth; if the predicted values ​​deviate from the optimal parameter range, calculating adjustment commands for driving the variable frequency fan or humidifier using a PID algorithm, thus forming the temperature setpoint and humidity setpoint; and simultaneously, dynamically adjusting the growth agent ratio parameters based on the predicted future plant growth environment trend and plant growth cycle.

[0025] In another possible embodiment, temperature, humidity, and CO2 concentration data from the historical cultivation process are first input into the LSTM network for training to obtain a model with environmental trend prediction capabilities. During the cultivation process, the edge computing gateway 3 continuously uploads real-time collected temperature, humidity, and CO2 concentration data to the platform layer 5. After receiving this data, the LSTM prediction control module 7 outputs predicted environmental parameters for the next 1-2 hours. Subsequently, the predicted value is compared with the preset optimal parameter range for seedling growth. If the predicted value deviates from the optimal range, the PID algorithm is immediately used to calculate the speed adjustment command of the variable frequency fan or the start / stop duration of the humidifier, generating the corresponding temperature and humidity setpoints. At the same time, the LSTM prediction control module 7 combines the environmental trend prediction results with the current growth cycle stage of the seedlings to dynamically adjust the ratio parameters of each component in the growth agent to ensure that the nutrient supply is accurately matched with the seedling growth needs.

[0026] By using an LSTM prediction model to anticipate environmental changes and combining it with a PID algorithm to actively regulate environmental parameters, the temperature, humidity, and other parameters are kept stable within the optimal range for seedling growth. At the same time, the growth agent ratio is dynamically adjusted according to the growth cycle, further improving the uniformity of seedling growth and the stability of seedling quality.

[0027] Traditional growth promoters are often applied using a single stock solution dilution method, which cannot flexibly adjust the proportion of various nutrients according to the real-time EC / pH value of the seedlings. This can easily lead to nutrient imbalance, resulting in slow seedling growth or seedling burn.

[0028] In an embodiment of this application, the growth agent dispensing device includes multiple peristaltic pumps connected in parallel, each of which is connected to an independent growth agent stock solution container. The device control module 8 controls the start, stop, and rotation speed of the different peristaltic pumps to mix and form a nutrient solution with a target EC / pH value.

[0029] In another possible embodiment, multiple peristaltic pumps are configured in parallel in the growth agent dispensing device. Each peristaltic pump corresponds to an independent growth agent stock solution container (such as nitrogen solution, phosphorus solution, potassium solution, etc.). After receiving the growth agent ratio parameters generated by the LSTM predictive control module 7, the device control module 8 of the platform layer 5 calculates the start-stop time and operating speed of each peristaltic pump according to the parameters. Then, it sends control commands to the corresponding peristaltic pumps. Each peristaltic pump extracts the corresponding amount of stock solution according to the commands and injects it into the main circulation pipeline 41. The solution is mixed in the pipeline to form a nutrient solution that meets the target EC / pH value and is finally delivered to the seedling cultivation device in each cultivation area.

[0030] By controlling the dosage of different growth agent solutions in parallel using multiple peristaltic pumps, precise mixing of various nutrients is achieved. The ratio of each solution can be dynamically adjusted according to the real-time EC / pH value of the seedlings, ensuring that the EC / pH value of the nutrient solution always meets the growth needs of the seedlings and avoiding nutrient imbalance.

[0031] In the embodiments of this application, each layer of the layered loading rack 13 is provided with an independent water pipe interface and an electrical interface, and the water interface is connected to the intelligent circulating water system 4 through a quick connector.

[0032] In another possible embodiment, each layer of the detachable layered filling rack 13 is provided with an independent water pipe interface and a circuit interface. The water pipe interface adopts a quick-connect connector to connect with the branch pipe of the intelligent circulating water system 4, and the circuit interface adopts a snap-fit ​​structure to connect with the power cord of the LED supplemental lighting device. When it is necessary to adjust the layout of the cultivation area, simply unplug the quick-connect water pipe connector and the snap-fit ​​circuit interface to disassemble or move the layer of the filling rack. After the adjustment is completed, the interface can be reconnected to restore its use.

[0033] By setting independent quick water and electricity interfaces on each layer of the layered filling rack 13, the cultivation rack can be quickly disassembled and its layout adjusted, which greatly improves the modular adaptability of the device and can flexibly meet the cultivation space requirements of different seedlings.

[0034] In the embodiments of this application, the edge computing gateway 3 communicates with the platform layer 5 via Modbus or MQTT protocols.

[0035] In another possible embodiment, Modbus and MQTT protocols are integrated into the communication module of the edge computing gateway 3. When the platform layer 5 and the edge gateway are in the same local area network, the Modbus protocol is used to establish a communication connection; when the platform layer 5 and the edge gateway are in a remote network environment, the connection is automatically switched to the MQTT protocol. During the nurturing process, the edge gateway uploads sensor data in real time through the selected protocol, and the platform layer 5 also sends control commands to the edge gateway through the same protocol, realizing bidirectional and stable transmission of data and commands.

[0036] In embodiments of this application, the device further includes a traceability management module, which is configured to generate a unique traceability code for each container and associate the traceability code with the full-cycle environmental data, growth agent dosing records, and quality inspection records of the corresponding container.

[0037] In another possible embodiment, a traceability management module is deployed at platform layer 5. When a container begins cultivation, the traceability management module automatically generates a unique traceability code and binds it to the container. During the cultivation process, the traceability management module associates and stores the container's full-cycle environmental data (including temperature, humidity, etc.), growth agent application records (including growth agent application time, ratio, etc.), and quality inspection records (including seedling growth status, quality grade, etc.) with the traceability code in real time. After the seedling cultivation is completed, the traceability code is marked on the product packaging, and users can scan the code to query the complete cultivation information of the corresponding batch of seedlings.

[0038] By linking the entire cultivation data with a unique traceability code, the entire process from seeding and environmental control to quality inspection is traceable. When quality problems occur, the problematic links can be quickly located and targeted rectification can be carried out, thereby enhancing the market trust in the product.

[0039] In the embodiments of this application, the insulation layer 11 is a polyurethane insulation layer 11 formed by on-site foaming.

[0040] In another possible embodiment, after the inner wall of the container is cleaned, polyurethane foam material is evenly sprayed onto the inner wall of the container using on-site foaming equipment. The foam material expands and cures on-site, forming a completely bonded insulation layer 11 with the inner wall of the container. Then, a moisture-proof coating 12 is sprayed onto the surface of the insulation layer 11 to prevent the high humidity environment from damaging the insulation layer 11.

[0041] The polyurethane insulation layer 11, formed by on-site foaming, can completely adhere to the inner wall of the container, greatly improving the insulation effect and thus ensuring the stability of the seedling growth environment.

[0042] Accordingly, this application also provides an AIoT-based intelligent cultivation method for modular container sprouts. The method includes: continuously collecting environmental data of each cultivation area within the modular container unit 1 through sensors in the environmental sensing and execution unit 2; uploading the environmental data to the platform layer 5 through the edge computing gateway 3; processing the environmental data, predicting environmental trends, and generating optimized control strategies and growth agent dispensing strategies through the LSTM predictive control module 7 of the platform layer 5; distributing the environmental control strategies and growth agent dispensing strategies to the edge computing gateway 3 through the device control module 8 in the platform layer 5; and driving the variable frequency fan, humidifier, LED supplemental lighting device in the environmental sensing and execution unit 2, and the peristaltic pump in the intelligent circulating water system 4 through the edge computing gateway 3 to perform corresponding control actions.

[0043] In another possible embodiment, all sensors in the environmental perception and execution unit 2 are first activated to continuously collect environmental data such as temperature, humidity, CO2 concentration, EC value, and pH value in each cultivation area. Then, the edge computing gateway 3 receives these sensor data in real time and uploads the data to the platform layer 5 via Modbus or MQTT protocol. After receiving the data, the LSTM predictive control module 7 of the platform layer 5 predicts future environmental trends and generates environmental optimization control strategies and growth agent dispensing strategies by combining preset optimal seedling growth parameters. Subsequently, the device control module 8 of the platform layer 5 converts these strategies into specific control commands and sends them to the edge computing gateway 3. Finally, according to the received commands, the edge computing gateway 3 drives the variable frequency fan to adjust its speed, starts and stops the humidifier, adjusts the brightness of the LED supplemental lighting equipment, and controls the peristaltic pump in the intelligent circulating water system 4 to adjust the amount of growth agent dispensing, thus completing the corresponding environmental regulation and growth agent dispensing actions. The entire process is repeated cyclically to achieve fully automated management of the cultivation process.

[0044] Through a fully automated closed-loop process of "sensing-uploading-analysis-decision-execution", manual operation has been replaced, realizing unmanned management of the cultivation process, greatly improving cultivation efficiency, while the accuracy of parameter control has been greatly improved, and the consistency of seedling quality has been significantly enhanced.

[0045] The optional embodiments of the present invention have been described in detail above with reference to the accompanying drawings. However, the embodiments of the present invention are not limited to the specific details in the above embodiments. Within the scope of the technical concept of the embodiments of the present invention, various simple modifications can be made to the technical solutions of the embodiments of the present invention, and these simple modifications all fall within the protection scope of the embodiments of the present invention.

[0046] It should also be noted that the various specific technical features described in the above embodiments can be combined in any suitable manner without contradiction. To avoid unnecessary repetition, the embodiments of the present invention will not describe the various possible combinations separately.

[0047] Furthermore, various different implementations of the present invention can be combined arbitrarily, as long as they do not violate the spirit of the present invention, they should also be regarded as the content disclosed in the present invention.

Claims

1. An AIoT-based containerized modular intelligent sprout cultivation device, characterized in that, include: Modular container units, with insulation and moisture-proof coatings on the inner walls, and the interior space is divided into multiple cultivation areas by detachable layered loading racks; The environmental sensing and execution unit includes at least one temperature sensor, humidity sensor, CO2 sensor, EC sensor and pH sensor installed in the cultivation area, as well as a variable frequency fan, humidifier, LED supplemental lighting device and growth agent dispensing device that are controlled to perform environmental regulation. An edge computing gateway, deployed inside the modular container unit, is communicatively connected to the environmental perception and execution unit and is used to collect sensor data and issue control commands. The intelligent circulating water system includes a main circulation pipeline and a condensate recovery branch connected to the main circulation pipeline. A sand filter unit, a UV disinfection unit, and a reverse osmosis unit are connected in series on the main circulation pipeline. A flow sensor and an electric valve are installed on the main circulation pipeline. A condensate collection device located on the top of the container is installed at the inlet end of the condensate recovery branch. The outlet end of the condensate collection device is connected to the main circulation pipeline upstream of the reverse osmosis unit. The outlet of the growth agent dosing device is connected to the main circulation pipeline downstream of the reverse osmosis unit via a peristaltic pump. The platform layer, which communicates with the edge computing gateway, includes: The data storage module is used to store time-series environmental data and device operation logs uploaded by the edge computing gateway; The LSTM predictive control module is used to receive the time-series environmental data, predict the plant growth environment trend in the future time period, and generate an optimized control strategy including temperature setpoint, humidity setpoint and growth ratio parameters based on the prediction results and the preset growth model. The device control module is used to send the optimized control strategy to the edge computing gateway to drive the environment perception and execution unit and the peristaltic pump.

2. The containerized modular intelligent sprout cultivation device based on AIoT according to claim 1, characterized in that, The optimized control strategy for generating the blow ratio parameter specifically includes: Historical and current temperature, humidity, and CO2 concentration data are input into a trained LSTM network, which outputs predicted values ​​of environmental parameters for future time periods. The predicted value is compared with the preset optimal parameter range for plant growth. If the predicted value deviates from the optimal parameter range, the PID algorithm is used to calculate the adjustment command for driving the variable frequency fan or humidifier, which constitutes the temperature setpoint and humidity setpoint. Meanwhile, the growth agent ratio parameters are dynamically adjusted based on the predicted future plant growth environment trends and plant growth cycles.

3. The containerized modular intelligent sprout cultivation device based on AIoT according to claim 1, characterized in that, The growth agent dispensing device includes multiple peristaltic pumps connected in parallel, each of which is connected to an independent growth agent stock solution container. The device control module controls the start, stop, and rotation speed of the different peristaltic pumps to mix and form a nutrient solution with a target EC / pH value.

4. The containerized modular intelligent sprout cultivation device based on AIoT according to claim 1, characterized in that, Each layer of the layered loading rack is equipped with an independent water pipe interface and an electrical interface. The water interface is connected to the intelligent circulating water system via a quick connector.

5. The containerized modular intelligent sprout cultivation device based on AIoT according to claim 1, characterized in that, The edge computing gateway communicates with the platform layer via Modbus or MQTT protocols.

6. The containerized modular intelligent sprout cultivation device based on AIoT according to claim 1, characterized in that, The device also includes a traceability management module, which is configured to generate a unique traceability code for each container and associate the traceability code with the full-cycle environmental data, growth agent dosing records and quality inspection records of the corresponding container.

7. The containerized modular intelligent sprout cultivation device based on AIoT according to claim 1, characterized in that, The insulation layer is a polyurethane insulation layer formed on-site by foaming.

8. A containerized modular intelligent cultivation method for sprouts based on AIoT, characterized in that, Applied to the system as described in any one of claims 1-7, the method comprises: The environmental data of each cultivation area within the modular container unit is continuously collected through the sensors in the environmental sensing and execution unit. The environmental data is uploaded to the platform layer via the edge computing gateway; The environmental data is processed by the LSTM predictive control module of the platform layer to predict environmental trends and generate optimized control strategies and growth agent delivery strategies. The environmental control strategy and growth agent delivery strategy are distributed to the edge computing gateway through the device control module in the platform layer. The edge computing gateway drives the variable frequency fan, humidifier, LED supplementary lighting device in the environmental perception and execution unit, as well as the peristaltic pump in the intelligent circulating water system, to perform corresponding control actions.

Images