Intelligent greenhouse environment control method and system

Abstract

The present invention discloses an intelligent greenhouse environment control method and system, which specifically relates to the field of greenhouse control technology. The method and system are used to solve the problem that existing greenhouse carbon dioxide concentration monitoring is difficult to identify stratification phenomena. The method determines whether carbon dioxide stratification exists by modeling and identifying carbon dioxide concentration and airflow distribution data in the greenhouse based on a cluster analysis algorithm. When stratification exists, the optical flow method is used to extract airflow velocity field characteristics and the velocity field image is analyzed in combination with a convolutional neural network to identify the airflow vortex morphological characteristics in the local area of ​​the greenhouse. The Shannon entropy is used to calculate the information entropy of the concentration change and the dynamic programming algorithm is used to analyze the diffusion path to evaluate the diffusion path complexity of the stratified area. The vortex risk area is further identified. The spectral analysis algorithm is used to analyze the chlorophyll distribution characteristics of the crop and the carbon dioxide concentration data are used to evaluate growth anomalies. The anomalies are used as auxiliary conditions to accurately control ventilation and carbon dioxide replenishment.

Description

Technical Field

[0001] The present invention relates to the technical field of greenhouse control, and more particularly, to an intelligent greenhouse environment control method and system. Background Art

[0002] In large or high greenhouses, if the ventilation is uneven or the air inlet is not positioned properly, the carbon dioxide concentration may be stratified in the upper or lower layers, forming high-concentration or low-concentration air masses; if the turbulent pattern of the internal airflow has vortices or dead corners, it will cause long-term stagnation or depletion areas of carbon dioxide.

[0003] In existing technologies, common arrangements may only monitor the average carbon dioxide concentration, which is difficult to reflect the actual conditions at different heights and locations. For crops that require precise carbon dioxide supplementation during their growth stages, if these eddies or blind spots cannot be discovered and regulated in a timely manner, it may cause overall crop growth differences and even lead to concentrated outbreaks of pests and diseases in specific microenvironments. Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art, embodiments of the present invention provide an intelligent greenhouse environment control method and system to solve the problems raised in the above-mentioned background technology.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] An intelligent greenhouse environment control method comprises the following steps:

[0007] Model and identify carbon dioxide concentration and airflow distribution data based on cluster analysis algorithms to determine whether there is carbon dioxide stratification;

[0008] When carbon dioxide stratification exists, the optical flow method is used to extract the airflow velocity field characteristics and the velocity field image is analyzed with a convolutional neural network to identify the vortex morphology characteristics of the airflow in the local area of ​​the greenhouse. The Shannon entropy is used to calculate the information entropy of the concentration change and the dynamic programming algorithm is used to analyze the diffusion path to evaluate the diffusion path complexity of the carbon dioxide stratified area in the greenhouse.

[0009] Identify vortex risk areas within the greenhouse based on the vortex morphology characteristics of local airflow areas and the complexity of the diffusion path of carbon dioxide stratification within the greenhouse.

[0010] The spectral analysis algorithm was used to analyze the chlorophyll distribution characteristics of crop leaves in the vortex risk area, and the abnormal distribution of crop growth conditions was evaluated in combination with regional carbon dioxide concentration data;

[0011] The abnormal distribution of crop growth conditions is used as an auxiliary condition to perform local ventilation adjustment operations in vortex risk areas, and local carbon dioxide supplementation operations are enabled when predetermined conditions are met.

[0012] In a preferred embodiment, modeling and identifying the carbon dioxide concentration and airflow distribution data based on a cluster analysis algorithm to determine whether carbon dioxide stratification occurs includes:

[0013] By arranging carbon dioxide concentration sensors and airflow monitoring devices at different heights and positions in the greenhouse, carbon dioxide concentration data and airflow distribution data at different points in the greenhouse are collected;

[0014] Preprocessing of carbon dioxide concentration data and airflow distribution data, including data denoising and time series alignment;

[0015] Based on the pre-processed carbon dioxide concentration data and airflow distribution data, a cluster analysis algorithm is used for grouping processing. The cluster analysis algorithm divides the data into multiple regions according to different spatial locations by setting grouping parameters that adapt to the environmental characteristics of the greenhouse.

[0016] Modeling the carbon dioxide concentration distribution and airflow characteristics in each area to generate a hierarchical model of carbon dioxide concentration and airflow distribution within the greenhouse;

[0017] Based on the stratification model, the carbon dioxide concentration gradient and airflow intensity difference in each area are calculated to determine whether there is carbon dioxide stratification in the greenhouse.

[0018] In a preferred embodiment, the airflow velocity field characteristics are extracted by optical flow method and analyzed by convolutional neural network on the velocity field image to identify the airflow vortex morphological characteristics in the local area of ​​the greenhouse, including:

[0019] Collect air velocity data at different locations in the greenhouse to form a multi-point air velocity distribution sequence;

[0020] Based on the airflow velocity distribution sequence, the optical flow method is used to extract the airflow velocity field characteristics. The optical flow method generates a dynamic characteristic map of the airflow velocity field by calculating the change in airflow velocity at adjacent time points.

[0021] The dynamic feature map of the airflow velocity field is used as input data to the convolutional neural network model for feature extraction and analysis. The convolutional neural network extracts the local vortex morphological features in the dynamic feature map of the airflow velocity field through multi-layer convolution and pooling operations.

[0022] The output layer of the convolutional neural network classifies and labels the extracted local vortex morphological features, and generates a vortex morphological distribution model based on the vortex morphological characteristics at different locations in the dynamic characteristic map of the airflow velocity field;

[0023] The airflow vortex position and vortex morphological characteristics in the local area of ​​the greenhouse are identified based on the vortex morphological distribution model. The airflow vortex morphological characteristics include vortex intensity, vortex radius and rotation direction.

[0024] In a preferred embodiment, the information entropy of concentration changes is calculated based on Shannon entropy and the diffusion path is analyzed in combination with a dynamic programming algorithm to evaluate the complexity of the diffusion path of the carbon dioxide stratified area in the greenhouse, including:

[0025] Collect carbon dioxide concentration data from different areas of the greenhouse and construct a concentration distribution matrix in spatial and temporal dimensions;

[0026] Calculate the carbon dioxide concentration gradient in each area based on the constructed concentration distribution matrix and generate a concentration gradient change map;

[0027] Based on the concentration gradient change graph, the Shannon entropy formula is used to calculate the concentration change information entropy of each region to express the complexity of the concentration change;

[0028] According to the spatial distribution results of concentration change information entropy, the diffusion path is optimized and modeled in combination with the dynamic programming algorithm to generate a diffusion path complexity model;

[0029] The diffusion path complexity index was calculated using the diffusion path complexity model to evaluate the diffusion path complexity of the carbon dioxide stratification region.

[0030] In a preferred embodiment, the diffusion path complexity index is calculated by a diffusion path complexity model to evaluate the diffusion path complexity of the carbon dioxide stratification region, specifically:

[0031] The diffusion path complexity index is calculated according to the diffusion path complexity model: ;in, represents the diffusion path complexity index, is the total number of path points, Indicates the path The diffusion cost of a point, Indicates the index of the waypoint.

[0032] In a preferred embodiment, based on the airflow vortex morphology characteristics of a local area of ​​the greenhouse and the complexity of the diffusion path of the carbon dioxide stratified area in the greenhouse, identifying the vortex risk area in the greenhouse includes:

[0033] Extract the vortex morphological characteristics of the airflow, including vortex intensity, vortex radius and rotation direction;

[0034] Obtain the diffusion path complexity index of the carbon dioxide stratification area in the greenhouse and construct the diffusion complexity distribution map of the carbon dioxide stratification area;

[0035] Based on the vortex morphological characteristics of the airflow and the diffusion complexity distribution map, a spatial correlation analysis was performed to determine the intersection area between the vortex area and the stratified area, and the intersection area was marked as the vortex risk area.

[0036] In a preferred embodiment, the chlorophyll distribution characteristics of crop leaves in the vortex risk area are analyzed by a spectral analysis algorithm, and the abnormal distribution of crop growth conditions is evaluated in combination with regional carbon dioxide concentration data, including:

[0037] Collect real-time image data of crop leaves in the eddy current risk area, including the reflectance spectrum information of the crop leaves;

[0038] Preprocessing the reflectance spectrum information of crop leaves, including spectrum correction and background noise removal;

[0039] The chlorophyll content distribution of crop leaves is calculated based on the spectral analysis algorithm, and the chlorophyll content distribution data corresponds to the spatial position of the crop leaves;

[0040] Obtain carbon dioxide concentration data at different locations within the eddy risk area and match the carbon dioxide concentration data with chlorophyll content distribution data by spatial location;

[0041] Based on the matching results, the correlation between the carbon dioxide concentration data and the chlorophyll content distribution data in the eddy risk area was analyzed, the abnormal distribution of crop growth conditions was evaluated, and the abnormal growth distribution areas were marked.

[0042] In a preferred embodiment, the abnormal distribution of crop growth conditions is used as an auxiliary condition to perform local ventilation adjustment operations in the vortex risk area, and the local carbon dioxide replenishment operation is enabled when the predetermined conditions are met, including:

[0043] Determine the specific spatial location of the abnormality based on the identified distribution area of ​​growth abnormalities;

[0044] Control the operating range and air volume parameters of local ventilation equipment in the vortex risk area based on the specific spatial location of the anomaly;

[0045] During local ventilation operation, real-time monitoring of carbon dioxide concentration and airflow velocity data in the abnormal growth distribution area is performed to determine whether the carbon dioxide concentration has reached the preset replenishment condition;

[0046] When the carbon dioxide concentration in the abnormal growth distribution area reaches the preset supplementary conditions, the local carbon dioxide supplementary equipment is activated to release carbon dioxide to the abnormal growth distribution area through the precise gas supply port.

[0047] On the other hand, the present invention provides an intelligent greenhouse environment control system, including a stratification phenomenon recognition module, a vortex characteristic analysis module, a diffusion complexity assessment module, an eddy current risk recognition module, an abnormal distribution assessment module, and an environmental precision control module;

[0048] Stratification phenomenon identification module: Based on the cluster analysis algorithm, it models and identifies the carbon dioxide concentration and airflow distribution data to determine whether there is carbon dioxide stratification phenomenon;

[0049] Vortex characteristic analysis module: When carbon dioxide stratification exists, the optical flow method is used to extract the airflow velocity field characteristics and the convolutional neural network is combined with the velocity field image to analyze the velocity field image to identify the airflow vortex morphological characteristics in the local area of ​​the greenhouse;

[0050] Diffusion Complexity Assessment Module: When carbon dioxide stratification occurs, the information entropy of concentration changes is calculated based on Shannon entropy and combined with a dynamic programming algorithm to analyze the diffusion path to assess the complexity of the diffusion path of carbon dioxide stratified areas within the greenhouse;

[0051] Vortex risk identification module: Identifies vortex risk areas within the greenhouse based on the vortex morphology characteristics of the airflow in local areas of the greenhouse and the complexity of the diffusion path of the carbon dioxide stratification area within the greenhouse;

[0052] Abnormal distribution assessment module: Analyzes the chlorophyll distribution characteristics of crop leaves in the vortex risk area through spectral analysis algorithms, and combines regional carbon dioxide concentration data to assess the abnormal distribution of crop growth conditions;

[0053] Environmental precision control module: uses the abnormal distribution of crop growth conditions as an auxiliary condition to perform local ventilation adjustment operations in vortex risk areas, and enables local carbon dioxide replenishment operations when predetermined conditions are met.

[0054] The technical effects and advantages of the intelligent greenhouse environment control method and system of the present invention are as follows:

[0055] 1. Modeling and identifying carbon dioxide concentration and airflow distribution data through a cluster analysis algorithm can not only reflect the carbon dioxide distribution at different heights and locations, but also effectively detect stratification phenomena that may be caused by high-concentration or low-concentration areas, solving the defect of existing technologies that rely solely on average concentration monitoring to find stratification problems. In addition, through the combination of optical flow method and convolutional neural network, the morphological characteristics of airflow vortices in local areas can be accurately extracted, providing technical support for the identification of vortex risk areas and avoiding the long-term depletion or retention of carbon dioxide caused by airflow dead corners or vortex areas.

[0056] 2. The abnormal distribution of crop growth conditions in the vortex risk area is evaluated through spectral analysis algorithms, and combined with real-time monitoring data of carbon dioxide concentration in the area, a decision-making basis is provided for subsequent regulation. By using the abnormal distribution of crop growth as an auxiliary condition, precise control of ventilation and carbon dioxide replenishment in the vortex risk area can be achieved, optimizing the regional environment. This intelligent control method avoids crop growth differentiation caused by long-term imbalance of local carbon dioxide concentration and reduces the possibility of concentrated outbreaks of pests and diseases, thereby significantly improving the growth efficiency and yield stability of crops in the greenhouse, while reducing energy waste and operating costs, with significant economic and environmental value. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 This is a schematic diagram of an intelligent greenhouse environment control method of the present invention;

[0058] Figure 2 This is a structural schematic diagram of an intelligent greenhouse environment control system of the present invention. DETAILED DESCRIPTION

[0059] The following will provide a clear and complete description of the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0060] Example 1: Figure 1 The present invention provides an intelligent greenhouse environment control method, comprising the following steps:

[0061] Based on the cluster analysis algorithm, the carbon dioxide concentration and airflow distribution data are modeled and identified to determine whether there is carbon dioxide stratification.

[0062] When carbon dioxide stratification exists, the optical flow method is used to extract the airflow velocity field characteristics and the convolutional neural network is used to analyze the velocity field image to identify the airflow vortex morphological characteristics in the local area of ​​the greenhouse. The information entropy of the concentration change is calculated based on the Shannon entropy and the diffusion path is analyzed in combination with the dynamic programming algorithm to evaluate the diffusion path complexity of the carbon dioxide stratified area in the greenhouse.

[0063] Based on the vortex morphological characteristics of the airflow in local areas of the greenhouse and the complexity of the diffusion path of the carbon dioxide stratified areas in the greenhouse, the vortex risk areas in the greenhouse are identified.

[0064] The chlorophyll distribution characteristics of crop leaves in the vortex risk area were analyzed using a spectral analysis algorithm, and the abnormal distribution of crop growth conditions was evaluated in combination with regional carbon dioxide concentration data.

[0065] The abnormal distribution of crop growth conditions is used as an auxiliary condition to perform local ventilation adjustment operations in vortex risk areas, and local carbon dioxide supplementation operations are enabled when predetermined conditions are met.

[0066] Modeling and identifying CO2 concentration and airflow distribution data based on cluster analysis algorithms to determine whether CO2 stratification exists, including:

[0067] By arranging carbon dioxide concentration sensors and airflow monitoring devices at different heights and positions in the greenhouse, carbon dioxide concentration data and airflow distribution data at different points in the greenhouse are collected:

[0068] Equipment for monitoring CO2 concentration and airflow distribution is installed within the greenhouse. CO2 concentration sensors are installed at various heights and locations within the greenhouse to collect data on CO2 concentrations within the greenhouse. Airflow monitoring devices utilize multiple wind speed sensors placed at key locations within the greenhouse (such as near vents, above crops, and between crops) to capture data on the speed and direction of airflow within the greenhouse.

[0069] Each CO2 sensor and airflow monitoring device is connected to the data acquisition system via wireless or wired communication, uploading the collected CO2 concentration and airflow velocity values ​​in real time. Data collection is set to once per minute to ensure sufficient temporal and spatial resolution while avoiding storage and computational overhead caused by excessive redundant data.

[0070] Preprocess the carbon dioxide concentration data and airflow distribution data, including data denoising and time series alignment:

[0071] Data denoising: A denoising method based on wavelet transform is used to decompose the collected data into multi-scale frequency components. By removing high-frequency noise components, the signal is smoothed, thereby eliminating the interference of instantaneous abnormal data.

[0072] Time series alignment: Since the data collected by the carbon dioxide concentration sensor and the airflow monitoring device may have timestamp errors, the dynamic time warping (DTW) algorithm is used to time align different data streams to ensure that the carbon dioxide concentration data and the airflow distribution data have the same time basis.

[0073] Based on the pre-processed carbon dioxide concentration data and airflow distribution data, a cluster analysis algorithm is used for grouping processing. The cluster analysis algorithm divides the data into multiple regions according to different spatial locations by setting grouping parameters that adapt to the environmental characteristics of the greenhouse:

[0074] A dynamic clustering algorithm based on K-means is selected. This algorithm can extract spatial location-related clustering features in high-dimensional data by iteratively updating cluster centers and cluster member distributions.

[0075] According to the structural characteristics of the greenhouse, the initial grouping parameters, including the number of groups and the cluster distance threshold, are set to ensure that the grouping results can adapt to the dynamic changes of the greenhouse environment.

[0076] The preprocessed carbon dioxide concentration data and airflow distribution data are used as algorithm inputs, and multiple iterative calculations are performed to generate spatial clustering results.

[0077] The carbon dioxide concentration distribution and airflow characteristics of each area are modeled to generate a hierarchical model of carbon dioxide concentration and airflow distribution in the greenhouse:

[0078] For each cluster area, the mean carbon dioxide concentration, concentration change gradient value, and distribution characteristics of airflow speed and direction in the area are calculated respectively.

[0079] Using the grid interpolation method, the carbon dioxide concentration distribution and airflow velocity field in the area are reconstructed into a three-dimensional layered model. The three-dimensional layered model includes the carbon dioxide concentration value, concentration change gradient and airflow velocity vector at each spatial location.

[0080] Based on the stratification model, the presence of carbon dioxide stratification in the greenhouse is determined by calculating the carbon dioxide concentration gradient and airflow intensity differences in each area:

[0081] If the carbon dioxide concentration gradient in a certain area exceeds a set threshold (such as 10 ppm / m), or the airflow intensity difference exceeds a set value (such as 1 m / s), it is determined that carbon dioxide stratification exists in the area.

[0082] Output the identification results of the stratification phenomenon, including the location of the specific stratification area and its corresponding concentration gradient and airflow intensity information.

[0083] The optical flow method is used to extract the airflow velocity field features and the convolutional neural network is used to analyze the velocity field image to identify the airflow vortex morphological characteristics in the local area of ​​the greenhouse, including:

[0084] Collect air velocity data at different locations in the greenhouse to form a multi-point air velocity distribution sequence:

[0085] Airflow monitoring devices are arranged at different heights inside the greenhouse, in edge areas, central areas, and near ventilation openings and air supply openings to ensure a comprehensive spatial distribution of airflow data. Each airflow monitoring device contains a sensor that can record the airflow speed and direction in real time.

[0086] The airflow monitoring device records data at fixed time intervals. Each record includes the magnitude and direction of the airflow velocity. All data is stored in the form of a time series. The data storage format includes location index, timestamp, airflow velocity magnitude and direction information.

[0087] Based on the airflow velocity distribution sequence, the optical flow method is used to extract the airflow velocity field characteristics. The optical flow method calculates the change in airflow velocity at adjacent time points to generate a dynamic feature map of the airflow velocity field:

[0088] Optical flow is a method that calculates the pixel motion vector field by analyzing image changes between adjacent time points. In this scenario, optical flow is used to calculate the dynamic changes in the airflow velocity field.

[0089] Based on the airflow velocity data, a time-series image sequence of the velocity field is established, and each frame of the image represents the spatial distribution of the airflow velocity in the greenhouse.

[0090] Assuming that the pixel motion of the airflow velocity field satisfies the optical flow constraint equation, the formula is as follows: ;in, Represents the horizontal gradient of the image, Represents the gradient of the image in the vertical direction, represents the gradient of the image in time, and Represent the components of optical flow in the horizontal and vertical directions respectively.

[0091] By solving the above equations, the motion vector of the airflow velocity field at each pixel position is obtained.

[0092] The output of the optical flow method is a set of vector field images, which represent the dynamic changes of the airflow velocity field at each position, namely the dynamic feature map of the airflow velocity field. This result is used as the input of the subsequent convolutional neural network.

[0093] The dynamic feature map of the airflow velocity field is used as input data to the convolutional neural network model for feature extraction and analysis. The convolutional neural network extracts the local vortex morphological features in the dynamic feature map of the airflow velocity field through multi-layer convolution and pooling operations:

[0094] The dynamic feature map of the airflow velocity field generated by the optical flow method is used as input and imported into the pre-trained convolutional neural network model for feature extraction and analysis.

[0095] The convolutional neural network model includes multiple convolutional layers, pooling layers, and fully connected layers. The convolutional layer is used to extract local features of the dynamic feature map of the airflow velocity field, and the size of the convolution kernel is set to 3×3 based on the characteristics of greenhouse data. The pooling layer is used to reduce the dimensionality of the feature map and retain the main features. The fully connected layer is used to integrate the features extracted by each layer to generate the vortex morphology classification results.

[0096] The first convolution layer performs a convolution operation on the input dynamic feature map of the airflow velocity field. The formula is as follows: ;in, Indicates that in the convolutional neural network In the layer, the output feature map is at position Pixel value of Represents the row index of the feature map; Represents the column index of the feature map; Indicates the current layer number of the convolutional neural network, counting from the first layer; Indicates the horizontal offset of the convolution kernel in the input feature map; Indicates the vertical offset of the convolution kernel in the input feature map; Indicates the The convolution kernel of the layer is at position The weight value of is the row index of the convolution kernel, which is used to indicate the row position in the convolution kernel matrix; Is the column index of the convolution kernel, used to indicate the column position in the convolution kernel matrix; Indicates the The layer input feature map is at position Pixel value of Indicates the first Row plus convolution kernel horizontal offset The pixel position of Indicates the first Column plus the vertical offset of the convolution kernel The pixel position of Represents the feature map of the previous layer; Indicates the The bias term of the layer convolution operation is a learnable parameter used to offset the convolution calculation results.

[0097] in, and The value range of is determined by the size of the convolution kernel (for example, the value range is {−1, 0, 1} for a 3×3 convolution kernel). The feature map is an abbreviation of the dynamic feature map of the airflow velocity field.

[0098] The last output layer classifies the extracted vortex morphological features and marks the vortex intensity, radius and rotation direction at different locations.

[0099] The output layer of the convolutional neural network classifies and labels the extracted local vortex morphological features, and generates a vortex morphological distribution model based on the vortex morphological characteristics at different locations in the dynamic characteristic map of the airflow velocity field:

[0100] The output layer of the convolutional neural network classifies the vortex morphological features extracted by the previous convolution and pooling layers through a fully connected layer. The classification criteria include the intensity, radius and rotation direction of the vortex.

[0101] The output classification labels include: vortex intensity level (such as weak, medium, strong); vortex radius range (divided in meters, divided into small vortex, medium vortex and large vortex); vortex rotation direction (divided into clockwise and counterclockwise).

[0102] The classification results are spatially labeled, and the morphological characteristics of the vortex are matched with its specific position in the dynamic characteristic diagram of the airflow velocity field.

[0103] The annotation results include: the spatial coordinates of the vortex center (determined by the point of maximum rotation intensity); the set of boundary points corresponding to the vortex radius; and the rotation direction of the vortex.

[0104] Based on the classification and labeling results of all vortices, a three-dimensional vortex morphology distribution model is generated: the model is based on the three-dimensional spatial coordinates of the greenhouse; each vortex is described by its center coordinates, radius and rotation direction.

[0105] The vortex morphology distribution model is stored in a matrix form. Each unit of the matrix corresponds to a small area in the greenhouse space, marking whether the area belongs to a vortex and the morphological characteristics of the vortex.

[0106] Based on the vortex shape distribution model, the airflow vortex position and vortex shape characteristics in the local area of ​​the greenhouse are identified. The airflow vortex shape characteristics include vortex intensity, vortex radius and rotation direction:

[0107] The specific location of airflow vortices in the greenhouse is identified through the vortex morphology distribution model: the location of the vortex center is determined by the point of maximum rotation intensity in the airflow velocity field; the vortex boundary is defined by the location where the rotation intensity gradient is lower than the preset threshold.

[0108] Vortex intensity is calculated as the difference in airflow velocity between the vortex center and its edges, indicating the severity of velocity fluctuations within the vortex. Vortex radius is calculated as the maximum distance between the vortex center and its edges, measured in meters. Rotation direction is determined by analyzing the rotation pattern of the airflow vector within the vortex region to determine whether the rotation direction is clockwise or counterclockwise.

[0109] The identification results are output in the form of a list, including the center coordinates, intensity, radius and rotation direction of each vortex.

[0110] When carbon dioxide stratification exists, the information entropy of concentration changes is calculated based on Shannon entropy and combined with a dynamic programming algorithm to analyze the diffusion path. This evaluates the complexity of the diffusion path in the carbon dioxide stratification area within the greenhouse, including:

[0111] Collect carbon dioxide concentration data from different areas of the greenhouse and construct a concentration distribution matrix in spatial and temporal dimensions:

[0112] Carbon dioxide concentration sensors are arranged at different heights and areas inside the greenhouse. The sensors are placed in key locations, including air supply inlets, ventilation openings, the center area of ​​the greenhouse, and the edge areas.

[0113] The carbon dioxide concentration data were collected at fixed time intervals (the sampling interval was set to 5 seconds), and the concentration values ​​and their corresponding spatial locations and timestamps were recorded.

[0114] Based on the collected carbon dioxide concentration data, a three-dimensional concentration distribution matrix is ​​constructed. The rows and columns of the matrix correspond to the spatial position index, and the third dimension corresponds to the time series.

[0115] Based on the constructed concentration distribution matrix, the carbon dioxide concentration gradient in each area is calculated to generate a concentration gradient change map:

[0116] According to the concentration distribution matrix, the concentration gradient in space and time dimensions is calculated: , ;in, Indicates location In time Horizontal carbon dioxide concentration gradient; Yes The partial derivative of the coordinate represents the rate of change of the lateral concentration; Indicates location In time The vertical carbon dioxide concentration gradient; represents the three-dimensional concentration distribution matrix; Yes The partial derivative of the coordinate represents the rate of change of the longitudinal concentration; Represents the horizontal position coordinate in space; Represents the vertical position coordinate in space; and represent the partial derivatives of the three-dimensional concentration distribution matrix in the horizontal and vertical directions, respectively.

[0117] The concentration gradient values ​​at all positions are expressed in image form to generate a concentration gradient change map, which serves as the basis for subsequent information entropy calculation.

[0118] Based on the concentration gradient change graph, the Shannon entropy formula is used to calculate the concentration change information entropy of each region to express the complexity of the concentration change:

[0119] For each region, the probability distribution of its concentration change is calculated, and the information entropy is calculated based on the following formula: ;in, It represents information entropy and is used to quantify the complexity of changes in carbon dioxide concentration in a certain area; Indicates the concentration change state in a certain area Probability of occurrence; An index indicating the possible change state of carbon dioxide concentration in a certain area; For a certain state The ratio of the number of occurrences to the total number of occurrences of all states in the region.

[0120] The information entropy results are used to quantify the complexity of the concentration changes in each area. The results are stored in a table format, including the location index and its corresponding information entropy value.

[0121] According to the spatial distribution results of the concentration change information entropy, the diffusion path is optimized and modeled in combination with the dynamic programming algorithm to generate a diffusion path complexity model:

[0122] Based on the spatial distribution of information entropy, a diffusion path optimization model is established, and the path selection must meet the condition of minimizing the entropy value.

[0123] The state transition equation of dynamic programming is: ;in, Indicates that the diffusion path reaches the path point The minimum cost; represents the minimum cost of the diffusion path to reach the path point bbb; Indicates a waypoint and waypoints Diffusion costs between Indicates that at all possible preceding path points Select a state to reach the waypoint The total cost is minimal; Indicates the index of the current target path point; Indicates the pre-order index of the path point; Represents the set of all path points that can be selected for the current diffusion path.

[0124] The dynamic programming algorithm outputs a path complexity model, which includes the spatial coordinates of the diffusion path and the corresponding cost.

[0125] The diffusion path complexity index is calculated using the diffusion path complexity model to evaluate the diffusion path complexity of the carbon dioxide stratification region:

[0126] According to the diffusion path complexity model, the diffusion path complexity index is calculated: ;in, represents the diffusion path complexity index, which is used to quantify the overall complexity of the diffusion path; is the total number of path points, indicating the length of the diffusion path; Indicates the path The diffusion cost of each point; Indicates the index of the path point, from 1 to .

[0127] The output diffusion path complexity index and its corresponding stratified area position are used to characterize the complexity of the diffusion path in the area. The larger the diffusion path complexity index, the higher the diffusion path complexity of the carbon dioxide stratified area, reflecting that the path of carbon dioxide is more tortuous or blocked when it diffuses from the stratified area to the surrounding area. This may cause carbon dioxide to stay in the stratified area for a longer time, resulting in uneven local concentration changes, affecting the effective absorption of carbon dioxide by crops. Furthermore, areas with excessively high or low concentrations may lead to differentiated crop growth, and in severe cases may even induce concentrated outbreaks of pests and diseases in specific microenvironments, reducing the overall crop growth efficiency and yield stability of the greenhouse.

[0128] Based on the vortex morphology characteristics of the airflow in the local area of ​​the greenhouse and the complexity of the diffusion path of the carbon dioxide stratification area in the greenhouse, the vortex risk areas in the greenhouse are identified, including:

[0129] Extract the vortex morphological characteristics of the airflow, including vortex intensity, vortex radius and rotation direction.

[0130] Obtain the diffusion path complexity index of the carbon dioxide stratification area in the greenhouse and construct the diffusion complexity distribution map of the carbon dioxide stratification area:

[0131] According to the diffusion path complexity index, the complexity of the carbon dioxide stratification area is constructed in the form of spatial distribution as a diffusion complexity distribution map.

[0132] Each pixel point in the diffusion complexity distribution map represents the diffusion path complexity index of a specific location, and a color gradient is used to indicate the high and low indexes, for example, red represents high complexity and blue represents low complexity.

[0133] Based on the spatial correlation analysis of the airflow vortex morphological characteristics and the diffusion complexity distribution map, the intersection area of ​​the vortex area and the stratified area is determined; the intersection area is marked as the vortex risk area:

[0134] The extracted airflow vortex morphological characteristic data are matched with the diffusion path complexity index data of the carbon dioxide stratification area according to spatial location.

[0135] The matched data were spatially analyzed to determine the overlapping areas between the regions with significant eddy characteristics and the regions with high diffusion path complexity index.

[0136] The intersection area is a spatial location that meets the following conditions: the area has significant vortex intensity, a large vortex radius or a clear rotation direction; at the same time, the diffusion path complexity index of the area is higher than the preset threshold.

[0137] The specific location of the intersection area and its corresponding vortex morphological characteristics and diffusion path complexity index are output in the form of a list for use in subsequent steps.

[0138] The intersection area is marked as a vortex risk area, and detailed information including spatial location, vortex intensity, vortex radius, rotation direction and diffusion path complexity index is output.

[0139] Significant vortex intensity is determined by calculating the magnitude of the change in the airflow velocity gradient and determining if it exceeds the average velocity gradient by a multiple. A larger vortex radius is determined by measuring the distance from the vortex center to the edge and setting it as larger if it exceeds the average of a preset radius range. A clear rotation direction is determined by analyzing the rotation pattern of the airflow vector and calculating the positive and negative values ​​of the vector field's curl to determine whether it is clockwise or counterclockwise. The judgment criteria are based on upper and lower limits determined by statistical analysis.

[0140] The preset threshold for the diffusion path complexity index is determined based on the complexity statistics under normal diffusion conditions within the greenhouse. The threshold is the mean plus a certain multiple of the standard deviation. For example, the diffusion index distribution range is calculated using historical data, and the value above the 70th percentile is used as the risk warning threshold for the complexity index.

[0141] The spectral analysis algorithm is used to analyze the chlorophyll distribution characteristics of crop leaves in the vortex risk area. Combined with regional carbon dioxide concentration data, the abnormal distribution of crop growth conditions is evaluated, including:

[0142] Collect real-time image data of crop leaves in the eddy current risk area. The image data includes the reflectance spectrum information of the crop leaves:

[0143] Multispectral imaging equipment is deployed within eddy current risk areas to collect real-time image data of crop leaves. The equipment is installed at key growing points within the area and is adjustable to accommodate varying crop heights.

[0144] The collected image data contains the reflectance spectrum information of crop leaves in different bands (such as red light, green light, and near-infrared light) for subsequent analysis.

[0145] The multispectral imaging equipment collects image data at fixed time intervals, and the sampling frequency is set to once every 10 minutes to ensure that the spectral changes of crop leaves are dynamically reflected.

[0146] The reflectance spectrum information of crop leaves is preprocessed, including spectrum correction and background noise removal:

[0147] A standard white plate is used for spectral correction to eliminate the influence of changes in ambient light intensity on the reflected spectrum data and ensure the accuracy of the data.

[0148] A filtering algorithm is used to remove background noise from the image data, including excess spectral signals outside the crop leaves. Each frame of image data is spatially filtered to retain the effective spectral information of the crop leaf area.

[0149] The chlorophyll content distribution of crop leaves is calculated based on the spectral analysis algorithm. The chlorophyll content distribution data corresponds to the spatial position of the crop leaves:

[0150] The chlorophyll content of crop leaves was calculated using the Normalized Difference Vegetation Index (NDVI) based on the corrected reflectance spectral information. Based on the chlorophyll index value for each pixel in the image data, a two-dimensional distribution map of chlorophyll content within the eddy current risk area was generated. Each pixel in the distribution map corresponds to the chlorophyll content at a specific spatial location.

[0151] Obtain carbon dioxide concentration data at different locations within the eddy risk area and match the carbon dioxide concentration data with the chlorophyll content distribution data by spatial location:

[0152] CO2 concentration sensors are deployed in vortex risk areas, installed at key locations, including the air inlet, vortex center, and area boundaries. CO2 concentration data is collected at regular intervals, with each collected concentration value associated with a spatial location and timestamp.

[0153] The carbon dioxide concentration data are mapped one-to-one to the spatial coordinates of the eddy risk area and stored in a three-dimensional matrix format, where the matrix dimensions are spatial position and time.

[0154] Based on the matching results, the correlation between the carbon dioxide concentration data and the chlorophyll content distribution data in the eddy risk area is analyzed to evaluate the abnormal distribution of crop growth conditions and mark the abnormal growth distribution areas:

[0155] The chlorophyll content distribution map is matched with the carbon dioxide concentration data based on spatial location to ensure that the chlorophyll content at each location is associated with the corresponding carbon dioxide concentration data.

[0156] Statistical regression analysis methods were used to calculate the correlation strength between carbon dioxide concentration and chlorophyll content, and to determine the impact pattern of high-concentration areas on chlorophyll distribution.

[0157] Based on the results of the association analysis, areas with significantly lower chlorophyll content than expected were identified and marked as areas of abnormal growth distribution.

[0158] Taking the abnormal distribution of crop growth conditions as an auxiliary condition, local ventilation adjustment operations are performed in the vortex risk area. When the predetermined conditions are met, local carbon dioxide supplementation operations are enabled, including:

[0159] Determine the specific spatial location of the abnormality based on the identified growth abnormality distribution area:

[0160] Extract the specific spatial location index of the growth anomaly distribution area, each location includes three-dimensional coordinates and the corresponding anomaly type label (such as low chlorophyll content, low carbon dioxide concentration, etc.).

[0161] Generate a three-dimensional position table of growth abnormality distribution areas as input for subsequent regulation.

[0162] Control the operating range and air volume parameters of local ventilation equipment in the vortex risk area based on the specific spatial location of the abnormality:

[0163] Based on the three-dimensional location table of the abnormal growth distribution area, determine the local ventilation equipment that needs to be operated and its coverage range.

[0164] The coverage of each ventilation device is set by the device installation position and operation angle parameters to ensure that the influence area of ​​the ventilation device overlaps with the growth abnormality distribution area.

[0165] Set ventilation air volume parameters based on the size and type of areas with abnormal growth distribution. For areas with abnormal chlorophyll content, set moderate air volumes to optimize airflow distribution. For areas with abnormal CO2 concentrations, set higher air volumes to promote balanced CO2 concentrations.

[0166] The air volume parameters are controlled by the fan speed and outlet air speed of the ventilation equipment. The specific parameter range is pre-calibrated according to the equipment model.

[0167] During local ventilation operation, the carbon dioxide concentration and airflow velocity data in the abnormal growth distribution area are monitored in real time to determine whether the carbon dioxide concentration reaches the preset supplementary conditions:

[0168] Carbon dioxide concentration sensors and airflow velocity sensors are placed within the growth anomaly distribution area to collect real-time concentration and velocity values ​​at each location. The sampling frequency can be set to once every 10 seconds, and the data is stored in a time series format to ensure the continuity of monitoring data.

[0169] The preset replenishment condition for CO2 concentration is determined based on the crop growth needs within the area, typically ranging from 400 to 800 ppm (specific values ​​are set based on crop type and growth stage). The replenishment condition is triggered when the monitored CO2 concentration in the area falls below the preset threshold and the airflow rate remains stable.

[0170] When the carbon dioxide concentration in the abnormal growth distribution area reaches the preset replenishment condition, the local carbon dioxide replenishment equipment is activated to release carbon dioxide to the abnormal growth distribution area through the precise gas replenishment port:

[0171] Activate CO2 replenishment devices installed near areas of abnormal growth distribution. Each device's release direction is precisely aligned with the abnormal area by adjusting the angle of the replenishment port to ensure that the CO2 directly covers the target area.

[0172] The amount of carbon dioxide released is calculated based on the area and concentration requirements of the abnormal growth distribution area. The specific formula is: ;in, Indicates the amount of carbon dioxide released (unit: liter); Indicates the area of ​​abnormal growth distribution (unit: square meters); Indicates the target concentration (unit: ppm); Indicates the current concentration (unit: ppm).

[0173] Example 2: Figure 2 A structural schematic diagram of an intelligent greenhouse environmental control system of the present invention is given, which includes a stratification phenomenon recognition module, a vortex characteristic analysis module, a diffusion complexity assessment module, an eddy current risk identification module, an abnormal distribution assessment module and an environmental precision control module.

[0174] Stratification phenomenon identification module: Based on the cluster analysis algorithm, the carbon dioxide concentration and airflow distribution data are modeled and identified to determine whether carbon dioxide stratification phenomenon exists.

[0175] Vortex characteristic analysis module: When carbon dioxide stratification exists, the airflow velocity field characteristics are extracted through the optical flow method and the velocity field image is analyzed in combination with the convolutional neural network to identify the airflow vortex morphological characteristics in the local area of ​​the greenhouse.

[0176] Diffusion complexity assessment module: When carbon dioxide stratification exists, the information entropy of concentration changes is calculated based on Shannon entropy and the diffusion path is analyzed in combination with a dynamic programming algorithm to evaluate the diffusion path complexity of the carbon dioxide stratification area in the greenhouse.

[0177] Vortex risk identification module: Identifies vortex risk areas within the greenhouse based on the airflow vortex morphology characteristics in local areas of the greenhouse and the complexity of the diffusion path of the carbon dioxide stratification area within the greenhouse.

[0178] Abnormal distribution assessment module: Analyzes the chlorophyll distribution characteristics of crop leaves in the vortex risk area through spectral analysis algorithms, and combines regional carbon dioxide concentration data to assess the abnormal distribution of crop growth conditions.

[0179] Environmental precision control module: uses the abnormal distribution of crop growth conditions as an auxiliary condition to perform local ventilation adjustment operations in vortex risk areas, and enables local carbon dioxide replenishment operations when predetermined conditions are met.

[0180] The above formulas are all dimensionless and numerical calculations. The formulas are obtained by collecting a large amount of data and performing software simulation to obtain the most recent real situation. The preset parameters and thresholds in the formulas are set by technicians in this field according to actual conditions.

[0181] The above embodiments can be implemented in whole or in part via software, hardware, firmware, or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product comprises one or more computer instructions or computer programs. When loaded or executed on a computer, the processes or functions described in the embodiments of this application are fully or partially performed. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired means (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium accessible by a computer or a data storage device such as a server or data center that contains a collection of one or more available media. The available medium can be magnetic media (e.g., floppy disks, hard disks, tapes), optical media (e.g., DVDs), or semiconductor media. The semiconductor media can be a solid-state drive.

[0182] Those skilled in the art will appreciate that the modules and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0183] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and modules described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0184] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of the modules is only a logical function division. In actual implementation, there may be other division methods, such as multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or modules, which can be electrical, mechanical or other forms.

[0185] The modules described as separate components may or may not be physically separate, and the components shown as modules may or may not be physical modules, and may be located in one place or distributed across multiple network modules. Some or all of the modules may be selected to achieve the purpose of this embodiment according to actual needs.

[0186] In addition, each functional module in each embodiment of the present application may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.

[0187] If the functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0188] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

[0189] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A smart greenhouse environment control method, characterized in that: The steps include: Model and identify carbon dioxide concentration and airflow distribution data based on cluster analysis algorithms to determine whether there is carbon dioxide stratification; When carbon dioxide stratification exists, the optical flow method is used to extract the airflow velocity field characteristics and the convolutional neural network is used to analyze the velocity field image to identify the airflow vortex morphological characteristics in the local area of ​​the greenhouse, including: Collect air velocity data at different locations in the greenhouse to form a multi-point air velocity distribution sequence; Based on the airflow velocity distribution sequence, the optical flow method is used to extract the airflow velocity field characteristics. The optical flow method generates a dynamic characteristic map of the airflow velocity field by calculating the change in airflow velocity at adjacent time points. The dynamic feature map of the airflow velocity field is imported as input data into the convolutional neural network model for feature extraction and analysis. The convolutional neural network extracts the local vortex morphological features in the dynamic feature map of the airflow velocity field through multi-layer convolution and pooling operations. The output layer of the convolutional neural network classifies and labels the extracted local vortex morphological features, and generates a vortex morphological distribution model based on the vortex morphological characteristics at different locations in the dynamic characteristic map of the airflow velocity field; Identify the airflow vortex location and vortex morphology characteristics in a local area of ​​the greenhouse based on the vortex morphology distribution model. The vortex morphology characteristics include vortex intensity, vortex radius, and rotation direction. The information entropy of concentration changes is calculated based on Shannon entropy and combined with a dynamic programming algorithm to analyze the diffusion path and evaluate the complexity of the diffusion path of carbon dioxide stratification in the greenhouse, including: Collect carbon dioxide concentration data from different areas of the greenhouse and construct a concentration distribution matrix in spatial and temporal dimensions; Calculate the carbon dioxide concentration gradient in each area based on the constructed concentration distribution matrix and generate a concentration gradient change map; Based on the concentration gradient change graph, the Shannon entropy formula is used to calculate the concentration change information entropy of each region to express the complexity of the concentration change; According to the spatial distribution results of concentration change information entropy, the diffusion path is optimized and modeled in combination with the dynamic programming algorithm to generate a diffusion path complexity model; The diffusion path complexity index was calculated using the diffusion path complexity model to evaluate the diffusion path complexity of the carbon dioxide stratification region. Identify vortex risk areas within the greenhouse based on the vortex morphology characteristics of local airflow areas and the complexity of the diffusion path of carbon dioxide stratification within the greenhouse. The spectral analysis algorithm was used to analyze the chlorophyll distribution characteristics of crop leaves in the vortex risk area, and the abnormal distribution of crop growth conditions was evaluated in combination with regional carbon dioxide concentration data; The abnormal distribution of crop growth conditions is used as an auxiliary condition to perform local ventilation adjustment operations in vortex risk areas, and local carbon dioxide supplementation operations are enabled when predetermined conditions are met.

2. The intelligent greenhouse environment control method according to claim 1, characterized in that: Modeling and identifying CO2 concentration and airflow distribution data based on cluster analysis algorithms to determine whether CO2 stratification exists, including: By arranging carbon dioxide concentration sensors and airflow monitoring devices at different heights and positions in the greenhouse, carbon dioxide concentration data and airflow distribution data at different points in the greenhouse are collected; Preprocessing of carbon dioxide concentration data and airflow distribution data, including data denoising and time series alignment; Based on the pre-processed carbon dioxide concentration data and airflow distribution data, a cluster analysis algorithm is used for grouping processing. The cluster analysis algorithm divides the data into multiple regions according to different spatial locations by setting grouping parameters that adapt to the environmental characteristics of the greenhouse. Modeling the carbon dioxide concentration distribution and airflow characteristics in each area to generate a hierarchical model of carbon dioxide concentration and airflow distribution within the greenhouse; Based on the stratification model, the carbon dioxide concentration gradient and airflow intensity difference in each area are calculated to determine whether there is carbon dioxide stratification in the greenhouse.

3. The intelligent greenhouse environment control method according to claim 1, characterized in that: The diffusion path complexity index is calculated using the diffusion path complexity model to evaluate the diffusion path complexity of the carbon dioxide stratification region. Specifically, The diffusion path complexity index is calculated according to the diffusion path complexity model: ;in, represents the diffusion path complexity index, is the total number of path points, Indicates the path The diffusion cost of a point, Indicates the index of the waypoint.

4. The intelligent greenhouse environment control method according to claim 1, characterized in that: Based on the vortex morphology characteristics of the airflow in the local area of ​​the greenhouse and the complexity of the diffusion path of the carbon dioxide stratification area in the greenhouse, the vortex risk areas in the greenhouse are identified, including: Extract the vortex morphological characteristics of the airflow, including vortex intensity, vortex radius and rotation direction; Obtain the diffusion path complexity index of the carbon dioxide stratification area in the greenhouse and construct the diffusion complexity distribution map of the carbon dioxide stratification area; Based on the vortex morphological characteristics of the airflow and the diffusion complexity distribution map, a spatial correlation analysis was performed to determine the intersection area between the vortex area and the stratified area, and the intersection area was marked as the vortex risk area.

5. The intelligent greenhouse environment control method according to claim 1, characterized in that: The spectral analysis algorithm is used to analyze the chlorophyll distribution characteristics of crop leaves in the vortex risk area. Combined with regional carbon dioxide concentration data, the abnormal distribution of crop growth conditions is evaluated, including: Collect real-time image data of crop leaves in the eddy current risk area, including the reflectance spectrum information of the crop leaves; Preprocessing the reflectance spectrum information of crop leaves, including spectrum correction and background noise removal; The chlorophyll content distribution of crop leaves is calculated based on the spectral analysis algorithm, and the chlorophyll content distribution data corresponds to the spatial position of the crop leaves; Obtain carbon dioxide concentration data at different locations within the eddy risk area and match the carbon dioxide concentration data with chlorophyll content distribution data by spatial location; Based on the matching results, the correlation between the carbon dioxide concentration data and the chlorophyll content distribution data in the eddy risk area was analyzed, the abnormal distribution of crop growth conditions was evaluated, and the abnormal growth distribution areas were marked.

6. The intelligent greenhouse environment control method according to claim 1, characterized in that: Taking the abnormal distribution of crop growth conditions as an auxiliary condition, local ventilation adjustment operations are performed in the vortex risk area. When the predetermined conditions are met, local carbon dioxide supplementation operations are enabled, including: Determine the specific spatial location of the abnormality based on the identified distribution area of ​​growth abnormalities; Control the operating range and air volume parameters of local ventilation equipment in the vortex risk area based on the specific spatial location of the anomaly; During local ventilation operation, real-time monitoring of carbon dioxide concentration and airflow velocity data in the abnormal growth distribution area is performed to determine whether the carbon dioxide concentration has reached the preset replenishment condition; When the carbon dioxide concentration in the abnormal growth distribution area reaches the preset supplementary conditions, the local carbon dioxide supplementary equipment is activated to release carbon dioxide to the abnormal growth distribution area through the precise gas supply port.

7. An intelligent greenhouse environment control system, used to implement the intelligent greenhouse environment control method according to any one of claims 1 to 6, characterized in that: It includes a layered phenomenon identification module, a vortex characteristic analysis module, a diffusion complexity assessment module, an eddy current risk identification module, an abnormal distribution assessment module, and an environmental precision control module; Stratification phenomenon identification module: Based on the cluster analysis algorithm, it models and identifies the carbon dioxide concentration and airflow distribution data to determine whether there is carbon dioxide stratification phenomenon; Vortex characteristic analysis module: When carbon dioxide stratification exists, the optical flow method is used to extract the airflow velocity field characteristics and the convolutional neural network is combined with the velocity field image to analyze the velocity field image to identify the airflow vortex morphological characteristics in the local area of ​​the greenhouse; Diffusion Complexity Assessment Module: When carbon dioxide stratification occurs, the information entropy of concentration changes is calculated based on Shannon entropy and combined with a dynamic programming algorithm to analyze the diffusion path to assess the complexity of the diffusion path of carbon dioxide stratified areas within the greenhouse; Vortex risk identification module: Identifies vortex risk areas within the greenhouse based on the vortex morphology characteristics of the airflow in local areas of the greenhouse and the complexity of the diffusion path of the carbon dioxide stratification area within the greenhouse; Abnormal distribution assessment module: Analyzes the chlorophyll distribution characteristics of crop leaves in the vortex risk area through spectral analysis algorithms, and combines regional carbon dioxide concentration data to assess the abnormal distribution of crop growth conditions; Environmental precision control module: uses the abnormal distribution of crop growth conditions as an auxiliary condition to perform local ventilation adjustment operations in vortex risk areas, and enables local carbon dioxide replenishment operations when predetermined conditions are met.

Images