Facility agriculture internet of things environment control method and device based on big data algorithm model
By collecting growth environment and physiological data in facility agriculture and using pre-trained models to generate equipment control strategies, the problem of AI algorithms lacking accurate quantitative modeling is solved, and precise environmental control and energy conservation are achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- NONGAN INFORMATION TECHNOLOGY (BEIJING) CO LTD
- Filing Date
- 2026-03-19
- Publication Date
- 2026-06-19
Smart Images
Figure CN122243678A_ABST
Abstract
Description
Technical Field
[0001] The embodiments disclosed herein relate to the field of computer technology, and more specifically to a method and apparatus for environmental control of facility agriculture based on big data algorithm models using the Internet of Things. Background Technology
[0002] Global agriculture faces resource constraints and climate challenges, and traditional experience-based planting methods can no longer meet the demands of precision production. The integration of IoT sensors, AI algorithms, and edge computing to build a closed-loop "perception-decision-execution" system has become a core path for upgrading modern agriculture. For example, combining agricultural expert experience with AI algorithms has led to the development of user-friendly crop growth models, lowering the technical threshold for producers. Furthermore, IoT-based environmental control in facility agriculture is achieved through human intervention.
[0003] However, when using the above method, the following technical problems often arise: Existing AI algorithms and human intervention lack precise quantitative modeling of crop internal physiological processes (such as photosynthesis and transpiration), making it impossible to regulate the environment according to the actual needs of crops. This leads to significant errors in the regulation results and disrupts the crop growth cycle. Furthermore, the inability to adjust equipment parameters in a timely manner due to environmental changes results in energy waste. Summary of the Invention
[0004] The summary portion of this disclosure is intended to provide a brief overview of the concepts, which will be described in detail in the detailed description portion. This summary portion is not intended to identify key or essential features of the claimed technical solutions, nor is it intended to limit the scope of the claimed technical solutions.
[0005] Some embodiments of this disclosure propose a method and apparatus for environmental control of facility agriculture based on big data algorithm models using the Internet of Things, in order to solve the technical problems mentioned in the background section above.
[0006] In a first aspect, some embodiments of this disclosure provide a facility agriculture IoT environmental control method based on a big data algorithm model. The method includes: collecting a set of growth environment parameters and a physiological dataset of crops within a facility agriculture area, wherein the growth environment parameters are collected by multiple different sensors, and the physiological dataset is collected by pre-deployed acquisition devices; inputting the growth environment parameters and the physiological dataset into a pre-trained crop growth detection model to obtain a crop growth status profile and a dynamic demand threshold; inputting the crop growth status profile and the growth environment parameters into a pre-trained edge prediction model to obtain predicted environmental parameter change information; generating a multi-objective optimization equipment control strategy set based on the predicted environmental parameter change information and the dynamic demand threshold; driving an environmental control execution device to perform actions based on the multi-objective optimization equipment control strategy set to obtain post-execution feedback data; and updating the crop growth detection model and the edge prediction model based on the post-execution feedback data.
[0007] Secondly, some embodiments of this disclosure provide a facility agriculture IoT environmental control device based on a big data algorithm model. The device includes: a data acquisition unit configured to acquire a set of growth environment parameters and a set of physiological data of crops within a facility agriculture area, wherein the set of growth environment parameters is acquired through multiple different sensors, and the set of physiological data is acquired through pre-deployed acquisition devices; a first input unit configured to input the set of growth environment parameters and the set of physiological data into a pre-trained crop growth detection model to obtain a crop growth status profile and a dynamic demand threshold; a second input unit configured to input the crop growth status profile and the set of growth environment parameters into a pre-trained edge prediction model to obtain predicted environmental parameter change information; a generation unit configured to generate a multi-objective optimization equipment control strategy set based on the predicted environmental parameter change information and the dynamic demand threshold; a driving unit configured to drive an environmental control execution device to perform actions based on the multi-objective optimization equipment control strategy set to obtain post-execution feedback data; and an update unit configured to update the crop growth detection model and the edge prediction model based on the post-execution feedback data.
[0008] Thirdly, some embodiments of this disclosure provide an electronic device, including: one or more processors; and a storage device having one or more programs stored thereon, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the method described in any implementation of the first aspect above.
[0009] Fourthly, some embodiments of this disclosure provide a computer-readable medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the method described in any of the implementations of the first aspect above.
[0010] The various embodiments of this disclosure have the following beneficial effects: The facility agriculture IoT environmental control method based on big data algorithm models in some embodiments of this disclosure reduces errors in the control results and reduces energy waste. Specifically, the reason why large errors in the control results lead to crop growth cycle disorders and energy waste is that existing AI algorithms and manual intervention lack precise quantitative modeling of the internal physiological processes of crops (such as photosynthesis and transpiration), and cannot control the environment according to the actual needs of the crops, resulting in large errors in the control results and crop growth cycle disorders. Furthermore, the inability to adjust equipment parameters in a timely manner due to environmental changes leads to energy waste. Based on this, the facility agriculture IoT environmental control method based on big data algorithm models in some embodiments of this disclosure first collects the crop growth environment parameter set and physiological dataset within the facility agriculture area. The growth environment parameter set is collected through multiple different sensors, and the physiological dataset is collected through pre-deployed acquisition equipment. This provides convenience for subsequent processing and solves the problems of high data transmission latency and poor reliability of traditional methods. Then, the aforementioned set of growth environment parameters and physiological datasets are input into a pre-trained crop growth detection model to obtain a crop growth status profile and dynamic demand thresholds. This enables rapid localized identification and prediction of crop growth and sudden weather changes, improving the accuracy of crop growth detection. Next, the aforementioned crop growth status profile and set of growth environment parameters are input into a pre-trained edge prediction model to obtain information on predicted environmental parameter changes. This allows for precise quantitative modeling of internal crop physiological processes (e.g., photosynthesis and transpiration). Then, based on the predicted environmental parameter changes and the dynamic demand thresholds, a multi-objective optimization equipment control strategy set is generated. This allows for environmental control according to the actual needs of the crop, reducing errors in the control results and stabilizing the crop growth cycle. Based on the multi-objective optimization equipment control strategy set, the environmental control execution equipment is driven to perform actions, obtaining post-execution feedback data. This allows for timely adjustment of equipment parameters based on environmental changes, reducing energy waste. Based on the post-execution feedback data, the aforementioned crop growth detection model and edge prediction model are updated. Therefore, environmental control can be performed according to the actual needs of the crop, reducing errors in the control results and stabilizing the crop growth cycle. Timely adjustment of equipment parameters due to environmental changes reduces energy waste. Attached Figure Description
[0011] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent from the accompanying drawings and the following detailed description. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic, and elements are not necessarily drawn to scale.
[0012] Figure 1 This is a flowchart of some embodiments of the facility agriculture Internet of Things environmental control method based on big data algorithm model according to this disclosure; Figure 2 This is a schematic diagram of the structure of some embodiments of the facility agriculture Internet of Things environmental control device based on a big data algorithm model according to the present disclosure; Figure 3 This is a schematic diagram of the structure of an electronic device suitable for implementing some embodiments of the present disclosure; Figure 4 This is a structural diagram of a crop growth detection model based on some embodiments of the IoT-based environmental control method for facility agriculture according to the present disclosure. Detailed Implementation
[0013] Embodiments of this disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this disclosure. It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure.
[0014] It should also be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings. Unless otherwise specified, the embodiments and features described in this disclosure can be combined with each other.
[0015] It should be noted that the concepts of "first" and "second" mentioned in this disclosure are used only to distinguish different devices, modules or units, and are not used to limit the order of functions performed by these devices, modules or units or their interdependencies.
[0016] It should be noted that the terms "a" and "a plurality of" used in this disclosure are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".
[0017] The names of messages or information exchanged between multiple devices in the embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of such messages or information.
[0018] This disclosure will now be described in detail with reference to the accompanying drawings and embodiments.
[0019] Figure 1 A flowchart 100 is shown, illustrating some embodiments of the IoT-based environmental control method for facility agriculture according to the present disclosure. This IoT-based environmental control method for facility agriculture includes the following steps: Step 101: Collect the set of growth environment parameters and physiological datasets of crops within the facility agriculture area.
[0020] In some embodiments, the implementing entity (e.g., a computing device) of the facility agriculture IoT environmental control method based on a big data algorithm model can collect a set of growth environment parameters and a set of physiological data of crops within the facility agriculture area. The growth environment parameter set is collected through multiple different sensors. The physiological data is collected through pre-deployed acquisition devices.
[0021] Here, the aforementioned set of growth environment parameters can be a collection of parameter data of the physical environment in which the crop exists. For example, the aforementioned set of growth environment parameters may include: {Temperature: 25.3 degrees Celsius, Humidity: 65%, CO2 concentration: 850 ppm, Soil volumetric water content: 22%}. The aforementioned different sensors may include, but are not limited to: temperature sensors, humidity sensors, CO2 sensors, and soil moisture sensors. The aforementioned physiological dataset can be a collection of data describing the crop's own life activities. For example, the aforementioned physiological dataset may include: {NDVI index: 0.72, Canopy temperature: 26.5 degrees Celsius, Net photosynthetic rate: 12.5 μmol CO2 / m³}. 2 / s, porosity: 0.25mol / m 2 The aforementioned data collection equipment may include, but is not limited to, infrared thermal imagers and multispectral cameras. The NDVI index can be the Normalized Difference Vegetation Index (NDVI). The aforementioned facility agriculture area can be the environmental area where crops grow. For example, the aforementioned facility agriculture area can be inside a facility agriculture greenhouse.
[0022] As an example, the aforementioned execution entity can measure ambient temperature and humidity data with an accuracy of ±0.2 degrees Celsius using a humidity sensor. CO2 concentration data is obtained via a CO2 sensor using non-dispersive infrared (NDIR) technology. Multispectral images in the red-edge band (e.g., 730nm ± 10nm) and near-infrared band (e.g., 780nm) are acquired using an infrared thermal imager to calculate the Normalized Difference Vegetation Index (NDVI). Furthermore, this data, along with real-time environmental parameters (such as photosynthetically active radiation, air temperature and humidity, and CO2 concentration), are input into a built-in crop physiological inversion model (a machine learning model trained on a large amount of sample data using the FvCB photosynthetic biochemical model and the Jarvis stomatal conductance model as core algorithms, with auxiliary corrections using a photosynthesis measuring instrument), thereby estimating key physiological parameters such as the crop's net photosynthetic rate and stomatal conductance.
[0023] Step 102: Input the above-mentioned growth environment parameter set and the above-mentioned physiological dataset into the pre-trained crop growth detection model to obtain a crop growth status profile and dynamic demand threshold.
[0024] In some embodiments, the execution entity can input the above-mentioned growth environment parameter set and the above-mentioned physiological dataset into a pre-trained crop growth detection model to obtain a crop growth status profile and dynamic demand threshold.
[0025] Here, the crop growth detection model described above can be trained using the aforementioned set of growth environment parameters and physiological datasets as input, and crop growth state profile and dynamic demand threshold as output. The crop growth detection model can include an input layer, a photosynthesis layer, a water stress assessment layer, and an output layer. The photosynthesis layer is used to calculate multiple state parameters during photosynthesis, i.e., a photosynthetic state parameter set. For example, the photosynthetic state parameter set includes, but is not limited to, net photosynthetic rate and stomatal conductance. The water stress assessment layer is used to calculate the evaporation rate and actual evaporation rate of crop water during transpiration. The evaporation rate can be the evaporation rate under stomatal conductance. The actual evaporation rate can be the true evaporation rate of the crop. For example, the photosynthesis layer can include an A-photosynthesis model. The water stress assessment layer can include a Penman-Monteith model. The crop growth state profile can include: growth stage labels, photosynthetic state parameters, and water evaporation index. For example, the growth stage label can be: "Tomato - Peak Flowering Stage". The above photosynthetic state parameters can be: "Net photosynthetic rate: 12.5 μmol CO2 / m 2 / s". The above water evaporation index can be: "CWSI: 0.3". The above water evaporation index can be the Crop Water Stress Index (CWSI). The above dynamic demand threshold can include: the temperature range of the current growth stage and the carbon dioxide concentration range. For example, the temperature range of the current growth stage can be [20 degrees Celsius, 25 degrees Celsius]. The above carbon dioxide concentration range can be [750 ppm, 850 ppm].
[0026] Optionally, the aforementioned execution entity can input the aforementioned growth environment parameter set and physiological dataset into a pre-trained crop growth detection model through the following steps to obtain a crop growth status profile and dynamic demand threshold: The first step is to align the above-mentioned growth environment parameter set and the above-mentioned physiological dataset with timestamps to generate an aligned growth environment parameter set and an aligned physiological dataset.
[0027] As an example, the aforementioned execution entity can use a unified timestamp to align the aforementioned growth environment parameter set and physiological dataset to generate an aligned growth environment parameter set and an aligned physiological dataset. For example, the unified timestamp could be a Unix timestamp.
[0028] The second step is to perform spatial region association on the above-mentioned aligned growth environment parameter set and the above-mentioned aligned physiological dataset to generate a registered growth environment parameter set and a registered physiological dataset.
[0029] As an example, the aforementioned execution entity can spatially correlate the aligned growth environment parameter set and the aligned physiological dataset using unified geographic coordinate information to generate a registered growth environment parameter set and a registered physiological dataset. For example, the unified geographic coordinate information can be WGS84 coordinates.
[0030] The third step is to fuse the above-mentioned registration growth environment parameter set and the above-mentioned registration physiological dataset to generate a multi-dimensional feature vector set.
[0031] As an example, the aforementioned execution entity can combine each registration growth environment parameter in the aforementioned registration growth environment parameter set with the corresponding registration physiological data in the aforementioned registration physiological dataset to generate a multidimensional feature vector, thereby obtaining a multidimensional feature vector set.
[0032] The fourth step is to input the above multidimensional feature vector set into the photosynthesis layer in the crop growth detection model to obtain the photosynthesis state parameter set.
[0033] As an example, the aforementioned execution entity can input each multidimensional feature vector in the multidimensional feature vector set into the photosynthesis layer in the crop growth detection model to generate photosynthesis state parameters and obtain a set of photosynthesis state parameters.
[0034] The fifth step involves inputting the aforementioned multidimensional feature vector set and photosynthetic state parameter set into the water stress assessment layer of the crop growth detection model to obtain the water evaporation index set.
[0035] As an example, the aforementioned implementing entity can input the multidimensional feature vector set (including net radiation, soil heat flux, air temperature, humidity, wind speed, and canopy temperature) into the water stress assessment layer of the crop growth monitoring model. This layer, based on the Penman-Monteith evapotranspiration model and the canopy energy balance principle, calculates the actual canopy resistance through inversion and ultimately calculates the crop water stress index (CWSI), with the formula CWSI = 1 - ETa / ETp. Here, ETa is the actual transpiration rate based on the inverted canopy resistance, and ETp is the potential transpiration rate based on the crop's specific minimum canopy resistance. The CWSI value range is [0, 1], with values closer to 1 indicating more severe water stress. Figure 4 As shown, the above multidimensional feature vector set is input into the input layer, then into the photosynthesis layer to obtain the photosynthesis state parameter set, and then the photosynthesis state parameter set and the multidimensional feature vector set are input together into the water stress assessment layer to obtain the water evaporation index set.
[0036] The sixth step is to determine the growth stage of the above physiological dataset in order to generate growth stage identifiers.
[0037] As an example, the aforementioned execution entity can determine the growth stage of the physiological dataset based on the temporal variation curve of the NDVI index and a preset crop calendar (e.g., seedling stage, vegetative growth stage, flowering stage, fruit enlargement stage, and maturity stage) to generate growth stage identifiers. The NDVI index can be the Normalized Difference Vegetation Index (NDVI).
[0038] The seventh step is to integrate the above growth stage identifiers, the above photosynthetic state parameter set, and the above water evaporation index set to obtain a crop growth state profile.
[0039] As an example, the aforementioned executing entity can merge the aforementioned growth stage identifiers, photosynthetic state parameter set, and water evaporation index set to obtain a crop growth state profile. That is, crop growth state profile = {growth stage identifiers, photosynthetic state parameter set, water evaporation index set}.
[0040] Step 8: Based on the above set of photosynthetic state parameters and the above set of water evaporation indexes, the preset environmental parameter threshold range is offset and adjusted to obtain the dynamic demand threshold.
[0041] As an example, the aforementioned implementing entity can adaptively adjust the preset environmental parameter threshold range based on the aforementioned photosynthetic state parameter set (photosynthetic state parameters such as photosynthetic limitation type) and the aforementioned water evaporation index set (for example, when the water evaporation index exceeds the preset environmental parameter threshold range, the upper limit of the suitable temperature is lowered by 1-2℃), to obtain a dynamic demand threshold adapted to the current real-time physiological state of the crop. For example: "When the water evaporation index (CWSI) is higher than the threshold X, the upper limit of the suitable temperature is lowered by Y℃." The threshold X is the critical point for determining whether the crop has entered a state of water stress requiring artificial intervention. It is obtained through historical data regression analysis (using "ambient temperature" and "CWSI" as independent variables and "photosynthetic rate" or "growth rate" as dependent variables in a multiple regression analysis) or expert experience value. The downward adjustment range Y is to balance the goals of "alleviating stress" and "maintaining growth." Under water stress, stomata close, and the transpiration cooling effect weakens. In order to maintain the canopy temperature below a certain dangerous value, it is necessary to lower the air temperature to compensate. The estimated downward adjustment Y can be roughly correlated with the increase in the crown temperature difference.
[0042] Step 103: Input the above crop growth status profile and the above growth environment parameter set into the pre-trained edge prediction model to obtain the predicted environmental parameter change information.
[0043] In some embodiments, the execution entity can input the crop growth status profile and the growth environment parameter set into a pre-trained edge prediction model to obtain information on predicted environmental parameter changes.
[0044] Here, the predicted environmental parameter changes mentioned above can refer to environmental conditions over a future period. For example, the predicted environmental parameter changes could include, but are not limited to, a temperature drop from 25 degrees Celsius to 22 degrees Celsius. Taking 7:00 AM as the current time, the future period could be from 8:00 AM to 9:00 AM. The aforementioned edge prediction model is trained using a crop growth status profile and a set of growth environment parameters as input, and the predicted environmental parameter changes as output. The aforementioned edge prediction model is used to predict environmental changes over a future period. The aforementioned edge prediction model includes an input layer, an LSTM encoder layer, and an output layer. The aforementioned LSTM encoder layer includes a Long Short-Term Memory (LSTM) network.
[0045] Optionally, the aforementioned execution entity can input the crop growth status profile and the growth environment parameter set into a pre-trained edge prediction model through the following steps to obtain information on predicted environmental parameter changes: The first step is to clean the time-series data in the above-mentioned growth environment parameter set to generate a time-series dataset of environmental parameters.
[0046] As an example, the aforementioned execution entity can perform linear imputation of missing values on the time-series data in the aforementioned growth environment parameter set to generate an imputed environment parameter time-series dataset. Then, the Z-score normalization method is used to normalize the imputed environment parameter time-series dataset to generate an environment parameter time-series dataset.
[0047] The second step is to extract dynamic indicators from the above crop growth status profile to generate a dynamic time-series dataset of crop physiology.
[0048] Here, the aforementioned dynamic indicators may include, but are not limited to: net photosynthetic rate and water evaporation index.
[0049] As an example, the aforementioned execution entity can extract the net photosynthetic rate and water evaporation index from the aforementioned photosynthetic state parameter set and transpiration state parameter set from the aforementioned crop growth state profile to generate a crop physiological dynamic time series dataset.
[0050] The third step involves performing feature stitching and sliding window sampling on the aforementioned environmental parameter time-series dataset and the aforementioned crop physiological dynamic time-series dataset to generate a multivariate feature time-series window set.
[0051] As an example, the aforementioned execution entity can set a fixed window length (e.g., 24 time steps, representing the past 2 hours) and a prediction step size (e.g., 12 time steps, representing the future 1 hour), perform sliding window sampling on the aforementioned environmental parameter time-series dataset and the aforementioned crop physiological dynamics time-series dataset, and concatenate the sampled sliding windows to generate a multivariate feature time-series window set. The aforementioned environmental parameter time-series dataset and the aforementioned crop physiological dynamics time-series dataset are spatiotemporally aligned, and sliding window sampling can simultaneously sample both datasets.
[0052] The fourth step is to input the above multi-feature time series window set into the data prediction layer of the above edge prediction model to obtain a standardized prediction value sequence.
[0053] Here, the aforementioned data prediction layer can be an LSTM encoder layer. This LSTM encoder layer includes a Long Short-Term Memory (LSTM) network.
[0054] The fifth step is to perform curve fitting on the above standardized predicted value sequence to generate information on predicted changes in environmental parameters.
[0055] Optionally, the aforementioned implementing entity can perform curve fitting on the standardized predicted value sequence through the following steps to generate information on predicted changes in environmental parameters: The first step is to destandardize the above standardized predicted value sequence to obtain the environmental parameter predicted value sequence.
[0056] As an example, the aforementioned implementing entity can perform destandardization on the above standardized prediction value sequence using the formula: Environmental parameter prediction value = standardized prediction value × standard deviation σ + mean μ, to obtain the environmental parameter prediction value sequence.
[0057] The second step is to perform trend aggregation on the above environmental parameter prediction value series to obtain a description of the environmental change trend.
[0058] As an example, the aforementioned implementing entity can perform first-order difference calculations on the predicted environmental parameter sequence, and then obtain the slope through linear fitting, which serves as a description of the environmental change trend. This slope can characterize the strength of the environmental change trend. The environmental change trend description could be: "The temperature will continue to drop in the next 2 hours, with an average rate of decrease of approximately 0.5 degrees Celsius per half hour."
[0059] The third step is to compare each predicted environmental parameter value in the above environmental parameter prediction value sequence with the dynamic demand threshold point by point to generate prediction value comparison results and obtain a prediction value comparison result set.
[0060] As an example, the aforementioned executing entity can compare each predicted environmental parameter value (e.g., the predicted temperature of 23 degrees Celsius at t+30min) in the above environmental parameter prediction value sequence with the dynamic demand threshold (e.g., temperature thresholds [24 degrees Celsius, 28 degrees Celsius]) point by point in chronological order to generate prediction value comparison results, thus obtaining a prediction value comparison result set. The prediction value comparison results in the prediction value comparison result set can be Boolean values, with True indicating "compliant" and False indicating "out of bounds".
[0061] The fourth step is to determine that there are prediction comparison results in the above prediction comparison result set that indicate prediction out of bounds, and to determine the environmental parameter prediction values corresponding to each prediction comparison result that indicates prediction out of bounds as the prediction out of bounds event set.
[0062] Here, the above-mentioned predicted out-of-bounds event set includes, but is not limited to: occurrence time, parameter type, predicted value, threshold boundary, and out-of-bounds magnitude. For example, the above-mentioned predicted out-of-bounds event set can be {time: t+45min, parameter type: temperature, predicted value: 22.8 degrees Celsius, lower threshold: 24.0 degrees Celsius, out-of-bounds magnitude: 1.2 degrees Celsius}.
[0063] The fifth step is to encapsulate the above environmental parameter prediction value sequence into prediction data blocks.
[0064] As an example, the aforementioned execution entity can package the aforementioned environmental parameter prediction value sequence into prediction data blocks according to a preset data structure (such as a JSON object).
[0065] The sixth step is to add the above description of environmental change trends to the above prediction data block to obtain enhanced prediction information.
[0066] As an example, the aforementioned implementing entity can fill the aforementioned environmental change trend description into the aforementioned prediction data block to obtain enhanced prediction information.
[0067] The seventh step is to convert the predicted out-of-bounds events in the above predicted out-of-bounds event set into instructions, and bind the converted out-of-bounds event instructions with the above enhanced prediction information to obtain the predicted environmental parameter change information.
[0068] Here, the above binding can be a combination.
[0069] As an example, the aforementioned executing entity can convert the predicted out-of-bounds events in the aforementioned predicted out-of-bounds event set into early warning commands in chronological order, and inject the early warning commands into the aforementioned enhanced prediction information to obtain information on changes in predicted environmental parameters. This conversion can be used to adjust the prediction 15 minutes before the out-of-bounds event occurs. For example, the predicted out-of-bounds event: {Time: t+45min, Parameter type: Temperature, Exceedance range: -1.2 degrees Celsius} can be converted into an early warning command: "It is recommended to implement a temperature increase at t+30min".
[0070] The content in steps one through seven above constitutes an inventive point of this disclosure, solving the technical problem of "low reliability of predicted environmental parameter changes leading to crop growth cycle disruption." Factors contributing to this low reliability and disruption of crop growth cycles often include: poor usability and operability of prediction results using AI models, and significant errors in prediction curves due to human intervention, all of which lower the reliability of predicted environmental parameter changes and disrupt crop growth cycles. Solving these factors improves the reliability of predicted environmental parameter changes and stabilizes crop growth cycles. To achieve this, firstly, the standardized predicted value sequence is de-standardized to obtain a sequence of predicted environmental parameters. This makes the prediction results more understandable and measurable, improving their usability and operability. Secondly, trend aggregation is performed on the predicted environmental parameter value sequence to obtain a description of environmental change trends. This reduces information complexity and allows for a rapid understanding of future environmental trends. Each predicted environmental parameter value in the above environmental parameter prediction sequence is compared point-by-point with the dynamic demand threshold to generate a prediction comparison result set. In response to the determination that there are prediction comparison results indicating prediction out-of-bounds errors in the prediction comparison result set, the environmental parameter prediction values corresponding to each prediction comparison result indicating prediction out-of-bounds errors are defined as the prediction out-of-bounds event set. This solves the problem of large errors in prediction curves caused by manual intervention, thus improving the reliability of predicted environmental parameter change information. The above environmental parameter prediction sequence is encapsulated into prediction data blocks. The above description of environmental change trends is added to the above prediction data blocks to obtain enhanced prediction information. The prediction out-of-bounds events in the above prediction out-of-bounds event set are converted into instructions, and the converted out-of-bounds event instructions are bound to the above enhanced prediction information to obtain predicted environmental parameter change information. Therefore, by improving the reliability of predicted environmental parameter change information, the crop growth cycle is stabilized.
[0071] Step 104: Based on the predicted environmental parameter change information and the dynamic demand threshold, generate a multi-objective optimization equipment control strategy set.
[0072] In some embodiments, the execution entity may generate a set of multi-objective optimization equipment control strategies based on the predicted environmental parameter change information and the dynamic demand threshold.
[0073] Here, the aforementioned multi-objective optimization equipment control strategy set can be a set of schemes for adjusting and controlling crop equipment after considering multiple optimization objectives. For example, the aforementioned multi-objective optimization equipment control strategy set may include, but is not limited to: starting the heat pump to low power 30 minutes in advance, and closing the roller shutter to 80%.
[0074] Optionally, the aforementioned implementing entity can generate a multi-objective optimization equipment control strategy set based on the predicted environmental parameter changes and the dynamic demand thresholds through the following steps: The first step is to determine the aforementioned dynamic demand threshold as the primary optimization objective.
[0075] The second step is to determine the first optimization objective and the preset optimization objective set as the optimization objective set.
[0076] Here, the aforementioned preset optimization target set may include, but is not limited to: minimum energy consumption, minimum temperature fluctuation, and minimum equipment start-up and shutdown losses.
[0077] The third step is to perform multi-objective optimization on the predicted environmental parameter change information based on the above-mentioned optimization objective set, so as to generate a multi-objective optimization equipment control strategy set.
[0078] As an example, the aforementioned execution entity can employ a genetic algorithm (e.g., NSGA-II) to perform multi-objective Pareto optimization on the control parameters of multiple devices corresponding to the predicted environmental parameter changes such as heat pump start-up and shutdown status, shutter opening degree, and fan speed, through the aforementioned optimization objective set, to obtain the globally optimal control strategy set, which serves as the multi-objective optimized device regulation strategy set.
[0079] Step 105: Based on the above multi-objective optimization equipment control strategy set, drive the environmental control execution equipment to perform actions and obtain post-execution feedback data.
[0080] In some embodiments, the aforementioned execution entity can drive the environmental control execution device to perform actions based on the aforementioned multi-objective optimization device control strategy set, and obtain post-execution feedback data.
[0081] Here, the aforementioned environmental control execution device can be an environmental control execution device. This device is used to alter the physical environment of the crop. The post-execution feedback data can be environmental parameters and crop data collected again after the action is performed. This feedback data can be used to verify the effectiveness of the action. For example, the post-execution feedback data could be {temperature: 23.5℃, humidity: 68%}. The action execution can be the environmental control execution device receiving instructions and performing physical actions. For example, the roller shutter motor receives the instruction "adjust the opening to 70%" and executes it.
[0082] Optionally, the aforementioned executing entity can drive the environmental control execution equipment to perform actions based on the above multi-objective optimization equipment control strategy set through the following steps, and obtain post-execution feedback data: The first step is to perform a simulation and evaluation of the above-mentioned multi-objective optimization equipment control strategy set to obtain the equipment control simulation result set.
[0083] As an example, the aforementioned implementing entity can input each multi-objective optimization equipment control strategy in the multi-objective optimization equipment control strategy set into the digital twin simulation model for simulation and evaluation, so as to generate equipment control simulation results and obtain the equipment control simulation result set.
[0084] The second step is to determine the equipment control simulation result with the highest comprehensive score from the above equipment control simulation results as the final execution strategy.
[0085] As an example, the aforementioned implementing entity can use the Monte Carlo method to introduce random fluctuations (e.g., simulated sensor errors, external weather disturbances) into the equipment control simulation result set to evaluate the comprehensive score of each equipment control simulation result in the set, and determine the equipment control simulation result with the highest comprehensive score as the final implementation strategy. The comprehensive score includes, but is not limited to: temperature uniformity score and strategy robustness score. The temperature uniformity score indicates that the smaller the standard deviation of the temperature, the more uniform the temperature, and the higher the score. The strategy robustness score indicates that the lower the failure rate (number of simulation failures divided by the total number of simulations), the higher the score. Simulation failure indicates that the temperature at the detection point is below the crop's tolerance limit (e.g., 18 degrees Celsius).
[0086] The third step is to drive the environmental control execution equipment to perform actions according to the above final execution strategy and obtain feedback data after execution.
[0087] As an example, the aforementioned executing entity can respond to the determination that the environmental control execution device has received the final execution strategy, drive the environmental control execution device to perform actions, and obtain post-execution feedback data. For example, the roller shutter motor receives the instruction "adjust the opening to 70%" and executes it. 30 minutes after execution, the sensor returns {temperature: 23.5℃, humidity: 68%}, and these data are the post-execution feedback data. For example, the environmental control execution device may include, but is not limited to: roller shutter motor, circulating fan, supplemental lighting, integrated water and fertilizer machine, and heat pump unit.
[0088] Step 106: Update the crop growth detection model and the edge prediction model based on the feedback data after execution.
[0089] In some embodiments, the execution entity may update the crop growth detection model and the edge prediction model based on the feedback data after execution.
[0090] As an example, the aforementioned execution entity can again collect new sets of growth environment parameters and new physiological datasets using different sensors and acquisition devices from step 101. Then, the new set of growth environment parameters and the new physiological dataset are merged into a feedback dataset. Next, using the training sample set comprised of the feedback dataset, incremental training or parameter fine-tuning is performed on the crop growth detection model and the edge prediction model. This can improve the accuracy of model predictions and the ability to generate optimization strategies. Here, the aforementioned update can be used to fine-tune model parameters using new data.
[0091] Further reference Figure 2 As an implementation of the methods shown in the above figures, this disclosure provides some embodiments of an IoT-based environmental control device for facility agriculture based on a big data algorithm model. These device embodiments are similar to... Figure 1 Corresponding to the method embodiments shown, this facility agriculture IoT environmental control device based on big data algorithm model can be specifically applied to various electronic devices.
[0092] like Figure 2 As shown, some embodiments of the facility agriculture Internet of Things environmental control device 200 based on big data algorithm model include: a data acquisition unit 201, a first input unit 202, a second input unit 203, a generation unit 204, a driving unit 205, and an update unit 206. The system includes: a data acquisition unit 201, configured to acquire a set of environmental growth parameters and a set of physiological data for crops within a facility agriculture area; wherein the environmental growth parameters are acquired through multiple sensors and the physiological data is acquired through pre-deployed acquisition devices; a first input unit 202, configured to input the environmental growth parameters and the physiological data into a pre-trained crop growth detection model to obtain a crop growth status profile and a dynamic demand threshold; a second input unit 203, configured to input the crop growth status profile and the environmental growth parameters into a pre-trained edge prediction model to obtain information on predicted environmental parameter changes; a generation unit 204, configured to generate a set of multi-objective optimization equipment control strategies based on the predicted environmental parameter changes and the dynamic demand threshold; a driving unit 205, configured to drive the environmental control execution device to perform actions based on the set of multi-objective optimization equipment control strategies to obtain post-execution feedback data; and an update unit 206, configured to update the crop growth detection model and the edge prediction model based on the post-execution feedback data.
[0093] It is understandable that the units recorded in the facility agriculture IoT environmental control device 200 based on big data algorithm models are similar to those in the reference system. Figure 1The steps described in the method correspond to each other. Therefore, the operations, features, and beneficial effects described above for the method are also applicable to the facility agriculture IoT environmental control device 200 based on a big data algorithm model and the units contained therein, and will not be repeated here.
[0094] The following is for reference. Figure 3 It shows a schematic diagram of the structure of an electronic device 300 (e.g., a computing device) suitable for implementing some embodiments of the present disclosure. Figure 3 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments of this disclosure.
[0095] like Figure 3 As shown, the electronic device 300 may include a processing unit 301 (e.g., a central processing unit, a graphics processor, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 302 or a program loaded from a storage device 308 into a random access memory (RAM) 303. The RAM 303 also stores various programs and data required for the operation of the electronic device 300. The processing unit 301, ROM 302, and RAM 303 are interconnected via a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.
[0096] Typically, the following devices can be connected to I / O interface 305: input devices 306 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 307 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 308 including, for example, magnetic tapes, hard disks, etc.; and communication devices 309. Communication device 309 allows electronic device 300 to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 3 An electronic device 300 with various devices is shown; however, it should be understood that it is not required to implement or possess all of the devices shown. More or fewer devices may be implemented or possessed alternatively. Figure 3 Each box shown can represent a device or multiple devices as needed.
[0097] In particular, according to some embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, some embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication device 309, or installed from storage device 308, or installed from ROM 302. When the computer program is executed by processing device 301, it performs the functions defined in the methods of some embodiments of this disclosure.
[0098] It should be noted that, in some embodiments of this disclosure, the computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium may be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In some embodiments of this disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In some embodiments of this disclosure, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wires, optical fibers, RF (radio frequency), etc., or any suitable combination thereof.
[0099] In some implementations, clients and servers can communicate using any currently known or future-developed network protocol such as HTTP (Hypertext Transfer Protocol) and can interconnect with digital data communication (e.g., communication networks) of any form or medium. Examples of communication networks include local area networks (“LANs”), wide area networks (“WANs”), the Internet (e.g., the Internet of Things), and peer-to-peer networks (e.g., ad hoc peer-to-peer networks), as well as any currently known or future-developed networks.
[0100] The aforementioned computer-readable medium may be included in the aforementioned electronic device; or it may exist independently and not assembled into the electronic device. The aforementioned computer-readable medium carries one or more programs, which, when executed by the electronic device, cause the electronic device to: collect a set of growth environment parameters and a physiological dataset of crops within a facility agriculture area, wherein the set of growth environment parameters is collected through multiple different sensors, and the physiological dataset is collected through pre-deployed acquisition devices; input the set of growth environment parameters and the physiological dataset into a pre-trained crop growth detection model to obtain a crop growth status profile and a dynamic demand threshold; input the crop growth status profile and the set of growth environment parameters into a pre-trained edge prediction model to obtain predicted environmental parameter change information; generate a multi-objective optimization equipment control strategy set based on the predicted environmental parameter change information and the dynamic demand threshold; drive the environmental control execution device to perform actions based on the multi-objective optimization equipment control strategy set, and obtain post-execution feedback data; and update the crop growth detection model and the edge prediction model based on the post-execution feedback data.
[0101] Computer program code for performing operations of some embodiments of this disclosure can be written in one or more programming languages or a combination thereof, including object-oriented programming languages such as Java, Python, and C++, and conventional procedural programming languages such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).
[0102] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.
[0103] The functions described above in this document can be performed at least in part by one or more hardware logic components. For example, exemplary types of hardware logic components that can be used, without limitation, include: field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), system-on-a-chip (SoCs), complex programmable logic devices (CPLDs), and so on.
[0104] The above description is merely a selection of preferred embodiments of this disclosure and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in the embodiments of this disclosure is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described inventive concept. For example, technical solutions formed by substituting the above-described features with (but not limited to) technical features with similar functions disclosed in the embodiments of this disclosure.
Claims
1. A method for environmental control of facility agriculture based on a big data algorithm model using the Internet of Things, characterized in that, include: Collect a set of growth environment parameters and a set of physiological data of crops in the facility agriculture area. The set of growth environment parameters is collected by multiple different sensors, and the set of physiological data is collected by pre-deployed collection equipment. The growth environment parameter set and the physiological dataset are input into a pre-trained crop growth detection model to obtain a crop growth status profile and dynamic demand threshold. The crop growth status profile and the growth environment parameter set are input into a pre-trained edge prediction model to obtain information on predicted changes in environmental parameters. Based on the predicted environmental parameter change information and the dynamic demand threshold, a multi-objective optimization equipment control strategy set is generated. Based on the multi-objective optimization equipment control strategy set, drive the environmental control execution equipment to perform actions and obtain post-execution feedback data; Based on the feedback data after execution, the crop growth detection model and the edge prediction model are updated.
2. The method according to claim 1, characterized in that, The step of driving the environmental control execution device to perform actions according to the multi-objective optimization equipment control strategy set, and obtaining post-execution feedback data, includes: The set of multi-objective optimized equipment control strategies is simulated and evaluated to obtain a set of equipment control simulation results. The equipment control simulation result with the highest comprehensive score is determined as the final execution strategy. According to the final execution strategy, the environmental control execution device is driven to perform actions and obtain post-execution feedback data.
3. The method according to claim 1, characterized in that, The step of inputting the crop growth status profile and the set of growth environment parameters into a pre-trained edge prediction model to obtain information on predicted changes in environmental parameters includes: Data cleaning is performed on the time-series data in the growth environment parameter set to generate a time-series dataset of environment parameters; Dynamic indicators are extracted from the crop growth status profile to generate a crop physiological dynamic time-series dataset; The environmental parameter time series dataset and the crop physiological dynamic time series dataset are subjected to feature concatenation and sliding window sampling to generate a multivariate feature time series window set; The multivariate feature time series window set is input into the data prediction layer of the edge prediction model to obtain a standardized prediction value sequence; Curve fitting is performed on the standardized predicted value sequence to generate information on predicted changes in environmental parameters.
4. The method according to claim 1, characterized in that, The step of generating a multi-objective optimization equipment control strategy set based on the predicted environmental parameter change information and the dynamic demand threshold includes: The dynamic demand threshold is determined as the first optimization objective; The first optimization objective and the preset optimization objective set are determined as the optimization objective set; Based on the optimization objective set, the predicted environmental parameter change information is subjected to multi-objective optimization to generate a multi-objective optimized equipment control strategy set.
5. The method according to claim 1, characterized in that, The step of inputting the growth environment parameter set and the physiological dataset into a pre-trained crop growth detection model to obtain a crop growth status profile and dynamic demand threshold includes: The growth environment parameter set and the physiological dataset are timestamped to generate an aligned growth environment parameter set and an aligned physiological dataset. Spatial region association is performed on the aligned growth environment parameter set and the aligned physiological dataset to generate a registered growth environment parameter set and a registered physiological dataset; The registered growth environment parameter set and the registered physiological dataset are fused to generate a multidimensional feature vector set; The multidimensional feature vector set is input into the photosynthesis layer in the crop growth detection model to obtain the photosynthesis state parameter set. The multidimensional feature vector set and the photosynthetic state parameter set are input into the water stress assessment layer of the crop growth detection model to obtain the water evaporation index set. The physiological dataset is subjected to growth stage determination to generate growth stage identifiers; By integrating the growth stage identifier, the photosynthetic state parameter set, and the water evaporation index set, a crop growth state profile is obtained. Based on the set of photosynthetic state parameters and the set of water evaporation indexes, the preset environmental parameter threshold range is offset and adjusted to obtain the dynamic demand threshold.
6. A facility agriculture IoT environmental control device based on a big data algorithm model, characterized in that, include: The data acquisition unit is configured to collect a set of growth environment parameters and a set of physiological data of crops within the facility agriculture area. The set of growth environment parameters is collected by multiple different sensors, and the set of physiological data is collected by pre-deployed acquisition devices. The first input unit is configured to input the set of growth environment parameters and the physiological dataset into a pre-trained crop growth detection model to obtain a crop growth status profile and dynamic demand threshold. The second input unit is configured to input the crop growth status image and the growth environment parameter set into a pre-trained edge prediction model to obtain information on the predicted changes in environmental parameters. The generation unit is configured to generate a set of multi-objective optimization equipment control strategies based on the predicted environmental parameter change information and the dynamic demand threshold. The drive unit is configured to drive the environmental control execution device to perform actions according to the multi-objective optimization device control strategy set, and obtain feedback data after execution; The update unit is configured to update the crop growth detection model and the edge prediction model based on the feedback data after execution.
7. An electronic device, characterized in that, include: One or more processors; A storage device on which one or more programs are stored; When the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in any one of claims 1 to 5.
8. A computer-readable medium, characterized in that, It stores a computer program thereon, wherein the computer program, when executed by a processor, implements the method as described in any one of claims 1 to 5.