Rain-sheltering cultivation greenhouse control method and system based on intelligent perception

Abstract

The invention relates to the technical field of agricultural equipment control, in particular to a rain-sheltering cultivation greenhouse control method and system based on intelligent perception. Importing the real-time growth state model diagram into a dictionary search tree for search matching, and obtaining growth moisture requirement range data of the pepper seedlings in the current growth stage in the target greenhouse; collecting real-time moisture data of soil in the target greenhouse at a plurality of preset time nodes, and analyzing according to the collected real-time moisture data and the growth moisture requirement range data in combination with a Bezier curve algorithm to obtain a soil moisture suitability deviation coefficient of the pepper seedlings in the target greenhouse; the moisture condition of soil in the target greenhouse is analyzed according to the soil moisture suitability deviation coefficient, the target greenhouse is regulated and controlled according to the analysis result, the greenhouse can be intelligently regulated and controlled according to the soil moisture condition, it is ensured that pepper seedlings grow under the appropriate soil moisture condition, and the yield of the pepper seedlings is improved. Therefore, the growth quality and yield of pepper seedlings are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of agricultural equipment control, and in particular to a control method and system for a rain shelter cultivation greenhouse based on intelligent perception. Background Art

[0002] In recent years, with the continuous development of agricultural science and technology, intelligent greenhouses have been widely used in pepper cultivation, effectively improving the yield and quality of peppers and reducing production costs. However, the control method of the traditional pepper cultivation rain shelter mainly relies on simple sensor data, which has the following shortcomings: First, it is impossible to perceive the growth status of pepper seedlings in real time, and it is difficult to accurately grasp the demand for water of pepper seedlings, resulting in unreasonable irrigation strategies, which easily cause excessive or low soil moisture and affect the growth and development of pepper seedlings; second, there is a lack of accurate judgment of the growth stage of pepper seedlings, and it is impossible to formulate corresponding control strategies according to the differentiated needs of different growth stages, resulting in slow growth of pepper seedlings or diseases; third, there is a lack of intelligent decision-making mechanism, and it is difficult to achieve refined management of the greenhouse environment. At the same time, in the prior art, although some studies have tried to use image recognition technology to identify the growth status of pepper seedlings, most of them lack accurate judgment of the growth stage of pepper seedlings, and it is difficult to combine with the management of soil moisture. There is a lack of accurate grasp of the growth needs of pepper seedlings, and it is difficult to achieve precise irrigation control. Therefore, there is an urgent need for an intelligent control method that can combine multiple factors such as the growth status, growth stage and soil moisture of pepper seedlings to achieve refined management of pepper cultivation rain shelters, improve pepper yield and quality, and reduce production costs. Summary of the invention

[0003] The present invention overcomes the shortcomings of the prior art and provides a method and system for controlling a rain shelter for pepper cultivation based on intelligent perception.

[0004] To achieve the above-mentioned purpose, the technical solution adopted by the present invention is:

[0005] The present invention discloses a rain shelter cultivation greenhouse control method based on intelligent perception, comprising the following steps:

[0006] Acquire a real-time growth status image of the pepper seedlings in the target greenhouse, and construct a real-time growth status model diagram of the pepper seedlings in the target greenhouse according to the real-time growth status image;

[0007] The growth characteristic images and growth water requirement range data corresponding to the pepper seedlings at different growth stages are obtained through the big data network, and a dictionary search tree is constructed according to the growth characteristic images and growth water requirement range data corresponding to the pepper seedlings at different growth stages;

[0008] Importing the real-time growth state model diagram into the dictionary search tree for search and matching, and obtaining the growth water requirement range data of the pepper seedlings in the target greenhouse at the current growth stage;

[0009] At a number of preset time points, real-time moisture data of the soil in the target greenhouse is collected, and the soil moisture suitability deviation coefficient of the pepper seedlings in the target greenhouse is obtained by combining the collected real-time moisture data with the growth moisture requirement range data and the Bezier curve algorithm analysis;

[0010] The moisture condition of the soil in the target greenhouse is analyzed according to the soil moisture suitability deviation coefficient, and the target greenhouse is regulated and processed according to the analysis result.

[0011] Preferably, a real-time growth status model diagram of pepper seedlings in the target greenhouse is constructed according to the real-time growth status image, specifically:

[0012] A Canny edge detection algorithm is introduced to perform feature extraction on the real-time growth status image based on the Canny edge detection algorithm to obtain contour feature information of each part of the pepper seedling;

[0013] Convert the real-time growth status image from the RGB color space to the HSV color space, and extract the color feature information of each part of the pepper seedling in the HSV space;

[0014] A gray-level co-occurrence matrix extraction algorithm is introduced, and the gray-level co-occurrence frequencies of pixel pairs in each preset direction in the real-time growth status image are extracted based on the gray-level co-occurrence matrix extraction algorithm, and the texture feature information of each part of the pepper seedling is determined according to the obtained gray-level co-occurrence frequencies;

[0015] According to the contour feature information, color feature information and texture feature information of each part of the pepper seedlings and combined with three-dimensional software, a real-time growth status model diagram of the pepper seedlings in the target greenhouse is constructed.

[0016] Preferably, the growth characteristic images and growth water requirement range data corresponding to the pepper seedlings at different growth stages are obtained through a big data network, and a dictionary search tree is constructed according to the growth characteristic images and growth water requirement range data corresponding to the pepper seedlings at different growth stages, specifically:

[0017] Acquire growth characteristic images and growth water requirement range data corresponding to pepper seedlings at different growth stages through a big data network, and construct growth characteristic model diagrams corresponding to pepper seedlings at different growth stages according to the growth characteristic images;

[0018] Constructing a dictionary tree, and dividing a number of dictionary branches in the dictionary tree according to different growth stages of peppers;

[0019] Obtain the growth water requirement range data corresponding to the pepper seedlings at different growth stages, and convert the growth water requirement range data corresponding to the pepper seedlings at different growth stages into specific string data;

[0020] Map each string data to the root node of the corresponding dictionary branch, and complete the coding operation on each dictionary branch;

[0021] The growth characteristic model graphs corresponding to the pepper seedlings at different growth stages are stored in the leaf nodes of the corresponding dictionary fork branches to obtain a dictionary search tree.

[0022] Preferably, the real-time growth state model diagram is imported into the dictionary search tree for search and matching, and the growth water requirement range data of the pepper seedlings in the target greenhouse at the current growth stage is obtained, specifically:

[0023] Obtain the real-time growth status model diagram of the pepper seedlings in the target greenhouse, and obtain the growth characteristic model diagram on each dictionary branch in the dictionary search tree;

[0024] Performing similarity analysis on the real-time growth state model graph and the growth characteristic model graphs on each dictionary fork branch to obtain the similarity between the real-time growth state model graph and the growth characteristic model graphs on each dictionary fork branch;

[0025] Sorting the similarities between the real-time growth state model graph and the growth feature model graphs on each dictionary branch to obtain the maximum similarity;

[0026] The growth characteristic model diagram corresponding to the maximum similarity is marked, the character string data of the dictionary fork of the marked growth characteristic model diagram is obtained, and the obtained character string data is interpreted to obtain the growth water requirement range data of the pepper seedlings in the target greenhouse at the current growth stage.

[0027] Preferably, real-time moisture data of the soil in the target greenhouse is collected at several preset time nodes, and the soil moisture suitability deviation coefficient of the pepper seedlings in the target greenhouse is obtained based on the collected real-time moisture data and the growth moisture requirement range data combined with the Bezier curve algorithm analysis, specifically:

[0028] The real-time moisture data of the soil in the target greenhouse is collected at several preset time nodes, the Bezier curve algorithm is introduced, and the real-time moisture data collected at each preset time node is used as a discrete point;

[0029] Connect each discrete point in sequence according to each acquisition time node to construct the Bezier curve of soil moisture dynamics in the target greenhouse;

[0030] Determine an upper constraint boundary and a lower constraint boundary according to the growth water requirement range data of the pepper seedlings in the target greenhouse at the current growth stage, and divide a constraint area in the water dynamic Bezier curve according to the upper constraint boundary and the lower constraint boundary;

[0031] Calculating the total curve length of the moisture dynamic Bezier curve outside the constraint area, and calculating the total curve length of the moisture dynamic Bezier curve within the constraint area;

[0032] The total curve length of the moisture dynamic Bezier curve outside the constraint area is processed by ratio with the total curve length of the moisture dynamic Bezier curve within the constraint area to obtain the soil moisture suitability deviation coefficient of the pepper seedlings in the target greenhouse.

[0033] Preferably, the moisture condition of the soil in the target greenhouse is analyzed according to the soil moisture suitability deviation coefficient, and the target greenhouse is regulated and processed according to the analysis result, specifically:

[0034] The soil moisture suitability deviation coefficient of the pepper seedlings in the target greenhouse is compared with the preset coefficient threshold;

[0035] If the soil moisture suitability deviation coefficient of the pepper seedlings in the target greenhouse is not greater than the preset coefficient threshold, it means that the moisture conditions of the soil in the target greenhouse meet the growth requirements of the pepper seedlings in the current growth stage, and the target greenhouse will not be regulated.

[0036] Preferably, the moisture condition of the soil in the target greenhouse is analyzed according to the soil moisture suitability deviation coefficient, and the target greenhouse is regulated and processed according to the analysis result, specifically:

[0037] If the soil moisture suitability deviation coefficient of the pepper seedlings in the target greenhouse is greater than the preset coefficient threshold, it means that the moisture condition of the soil in the target greenhouse does not meet the growth requirements of the pepper seedlings in the current growth stage, then the average value of the real-time moisture data collected at several preset time nodes is calculated to obtain the average moisture data;

[0038] The average value of the moisture data is compared with a preset threshold value. If the average value of the moisture data is greater than the preset threshold value, the soil in the target greenhouse is determined to be in a drought state; otherwise, the soil in the target greenhouse is determined to be in an over-humid state.

[0039] Preferably, the moisture condition of the soil in the target greenhouse is analyzed according to the soil moisture suitability deviation coefficient, and the target greenhouse is regulated according to the analysis result, further comprising the following steps:

[0040] If the soil in the target greenhouse is too wet, the real-time rainfall in the preset range of the target greenhouse is obtained. If the real-time rainfall in the preset range of the target greenhouse is greater than the preset rainfall, the roof of the target greenhouse is controlled to be closed to achieve the rain-proof function;

[0041] If the real-time rainfall within the preset range of the target greenhouse is not greater than the preset rainfall, the real-time light intensity within the preset range of the target greenhouse is obtained; if the real-time light intensity within the preset range of the target greenhouse is greater than the preset intensity value, the roof of the target greenhouse is controlled to open to realize the natural sunlight dehumidification function.

[0042] Preferably, the moisture condition of the soil in the target greenhouse is analyzed according to the soil moisture suitability deviation coefficient, and the target greenhouse is regulated according to the analysis result, further comprising the following steps:

[0043] If the soil in the target greenhouse is in a drought condition, the real-time rainfall in the preset range of the target greenhouse is obtained. If the real-time rainfall in the preset range of the target greenhouse is greater than the preset rainfall, the roof of the target greenhouse is controlled to open to realize the natural rainwater irrigation function;

[0044] If the real-time rainfall within the preset range of the target greenhouse is not greater than the preset rainfall, the real-time light intensity within the preset range of the target greenhouse is obtained; if the real-time light intensity within the preset range of the target greenhouse is greater than the preset intensity value, the roof of the target greenhouse is controlled to be closed to enhance the moisturizing function of the target greenhouse.

[0045] The present invention also discloses a rain shelter cultivation greenhouse control system based on intelligent perception, the rain shelter cultivation greenhouse control system includes a memory and a processor, the memory stores a rain shelter cultivation greenhouse control method program, when the rain shelter cultivation greenhouse control method program is executed by the processor, any one of the rain shelter cultivation greenhouse control method steps is implemented.

[0046] The present invention solves the technical defects existing in the background technology, and has the following beneficial effects: This method of controlling a rain shelter greenhouse for pepper cultivation based on intelligent perception realizes accurate monitoring and judgment of the growth status of pepper seedlings and soil moisture conditions by integrating image analysis, big data processing, algorithm calculation and environmental factor considerations. It can accurately obtain the water demand of the pepper seedlings at the current growth stage, quantify the suitability of soil moisture, and intelligently regulate the greenhouse environment according to the analysis results to ensure that the pepper seedlings grow under suitable soil moisture conditions, thereby improving the growth quality and yield of the pepper seedlings, while reducing the blindness of manual intervention and improving the efficiency and scientificity of greenhouse cultivation management. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, drawings of other embodiments can be obtained based on these drawings without paying creative work.

[0048] Figure 1 This is the overall method flow chart of the pepper cultivation rain shelter greenhouse control method;

[0049] Figure 2 This is a partial method flow chart of the control method of the rain shelter greenhouse for pepper cultivation;

[0050] Figure 3 This is the system block diagram of the control system for the pepper cultivation rain shelter greenhouse. DETAILED DESCRIPTION

[0051] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0052] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0053] The present invention discloses a rain shelter cultivation greenhouse control method based on intelligent perception, such as Figure 1 As shown, the following steps are included:

[0054] S102: Acquire a real-time growth status image of the pepper seedlings in the target greenhouse, and construct a real-time growth status model diagram of the pepper seedlings in the target greenhouse according to the real-time growth status image;

[0055] S104: acquiring growth characteristic images and growth water requirement range data corresponding to pepper seedlings at different growth stages through a big data network, and constructing a dictionary search tree according to the growth characteristic images and growth water requirement range data corresponding to pepper seedlings at different growth stages;

[0056] S106: Importing the real-time growth state model diagram into the dictionary search tree for search and matching, and obtaining the growth water requirement range data of the pepper seedlings in the target greenhouse at the current growth stage;

[0057] S108: collecting real-time moisture data of the soil in the target greenhouse at a number of preset time nodes, and obtaining a soil moisture suitability deviation coefficient of the pepper seedlings in the target greenhouse based on the collected real-time moisture data and the growth moisture requirement range data combined with a Bezier curve algorithm analysis;

[0058] S110: analyzing the moisture condition of the soil in the target greenhouse according to the soil moisture suitability deviation coefficient, and performing control processing on the target greenhouse according to the analysis result.

[0059] This method of controlling a rain shelter greenhouse for pepper cultivation based on intelligent perception integrates image analysis, big data processing, algorithm calculation, and environmental factors to achieve accurate monitoring and judgment of the growth status of pepper seedlings and soil moisture conditions. It can accurately obtain the water demand of pepper seedlings at the current growth stage, quantify the suitability of soil moisture, and intelligently regulate the greenhouse environment based on the analysis results to ensure that the pepper seedlings grow under suitable soil moisture conditions, thereby improving the growth quality and yield of pepper seedlings, while reducing the blindness of manual intervention and improving the efficiency and scientificity of greenhouse cultivation management.

[0060] Preferably, a real-time growth status model diagram of pepper seedlings in the target greenhouse is constructed according to the real-time growth status image, specifically:

[0061] A Canny edge detection algorithm is introduced to perform feature extraction on the real-time growth status image based on the Canny edge detection algorithm to obtain contour feature information of each part of the pepper seedling;

[0062] Convert the real-time growth status image from the RGB color space to the HSV color space, and extract the color feature information of each part of the pepper seedling in the HSV space;

[0063] A gray-level co-occurrence matrix extraction algorithm is introduced, and the gray-level co-occurrence frequencies of pixel pairs in each preset direction in the real-time growth status image are extracted based on the gray-level co-occurrence matrix extraction algorithm, and the texture feature information of each part of the pepper seedling is determined according to the obtained gray-level co-occurrence frequencies;

[0064] According to the contour feature information, color feature information and texture feature information of each part of the pepper seedlings and combined with three-dimensional software, a real-time growth status model diagram of the pepper seedlings in the target greenhouse is constructed.

[0065] It should be noted that the Canny edge detection algorithm is a classic edge detection algorithm. When processing the real-time growth status image of pepper seedlings, it can accurately identify the edges of various parts of the pepper seedlings, thereby obtaining contour feature information. For example, it can clearly distinguish the edge shapes of the stems, leaves, fruits and other parts of the pepper seedlings, which helps to determine the overall morphological structure of the pepper seedlings and provide a basic shape framework for subsequent model construction.

[0066] It is meaningful to convert the image from RGB color space to HSV color space. RGB color space is sensitive to changes in lighting, while HSV color space is more consistent with human perception of color. Extracting color feature information of each part of pepper seedlings in HSV space can more accurately describe the color state of pepper seedlings. For example, the color of pepper seedlings at different growth stages may change from light green to dark green, and the color of fruits may change from green to red. The color feature extraction of HSV space can capture these changes well.

[0067] The gray-level co-occurrence matrix extraction algorithm is used to extract the gray-level co-occurrence frequency of pixel pairs in each preset direction in the real-time growth status image. The texture feature information of each part of the pepper seedling can be determined through this frequency. The leaves, stems and other parts of the pepper seedling have different textures. For example, the surface of the leaves may have subtle vein textures, and the stems may have a certain roughness texture. These texture features help to further distinguish different parts of the pepper seedlings and can reflect the health status and growth status of the pepper seedlings.

[0068] Based on the previously acquired contour feature information, color feature information, and texture feature information of each part of the pepper seedling, the real-time growth status model diagram of the pepper seedling in the target greenhouse is constructed in combination with 3D software. The 3D software can integrate these feature information to construct a realistic 3D model that can reflect the actual growth status of the pepper seedling. This model can display the morphology, color, texture and other characteristics of the pepper seedling from multiple angles.

[0069] Through the above series of operations, a model diagram that comprehensively and accurately reflects the real-time growth status of pepper seedlings in the target greenhouse can be constructed. This model diagram integrates the contour, color, texture and other characteristics of the pepper seedlings, and describes the growth status of the pepper seedlings in detail from multiple dimensions. It can provide an intuitive and accurate basis for subsequent judgment of the growth stage, health status and difference from the standard growth status of the pepper seedlings, and help to accurately formulate greenhouse control strategies for the growth needs of pepper seedlings, thereby improving the growth quality and yield of pepper seedlings.

[0070] Preferably, the growth characteristic images and growth water requirement range data corresponding to the pepper seedlings at different growth stages are obtained through the big data network, and a dictionary search tree is constructed according to the growth characteristic images and growth water requirement range data corresponding to the pepper seedlings at different growth stages, such as Figure 2 As shown, specifically:

[0071] S202: Acquire growth characteristic images and growth water requirement range data corresponding to pepper seedlings at different growth stages through a big data network, and construct growth characteristic model diagrams corresponding to pepper seedlings at different growth stages according to the growth characteristic images;

[0072] S204: constructing a dictionary tree, and dividing a number of dictionary branches in the dictionary tree according to different growth stages of peppers;

[0073] S206: Obtaining the growth water requirement range data corresponding to the pepper seedlings at different growth stages, and converting the growth water requirement range data corresponding to the pepper seedlings at different growth stages into specific character string data;

[0074] S208: Mapping each string data to the root node of the corresponding dictionary branch, completing the coding operation on each dictionary branch;

[0075] S210: storing the growth characteristic model graphs corresponding to the pepper seedlings at different growth stages in the leaf nodes of the corresponding dictionary fork branches to obtain a dictionary search tree.

[0076] It should be noted that, first, the growth characteristic images and growth water requirement range data of pepper seedlings at different growth stages are obtained through the big data network. The growth characteristic model diagram is constructed using the growth characteristic image. This step helps to express the appearance characteristics of pepper seedlings at different growth stages in a structured model form. For example, the height of the pepper seedlings, the number and shape of leaves, the morphology of flowers and fruits, etc. are different at different stages such as the seedling stage, the flowering stage, and the fruiting stage. The growth characteristic model diagram can accurately capture these differences. Constructing a dictionary tree is a data structure organization method. By cutting out several dictionary branches in the dictionary tree, a framework structure is provided for the subsequent storage of relevant information of pepper seedlings at different stages. This structure is similar to a classification directory, which can classify and store the information of pepper seedlings at different growth stages for quick search and matching. The growth water requirement range data is converted into specific string data in order to adapt to the data storage and retrieval method of the dictionary tree. Then, these string data are mapped to the root nodes of the corresponding dictionary branches for coding operations. This operation enables each dictionary branch to have an identifier related to the growth water requirement range, which is convenient for quickly locating information related to specific water requirements during search. Finally, the growth characteristic model graphs corresponding to the pepper seedlings at different growth stages are stored in the leaf nodes of the corresponding dictionary fork branches, thus completing the construction of the dictionary search tree. In this way, the dictionary search tree integrates the growth characteristic model graphs and the growth water requirement range data of the pepper seedlings at different growth stages, forming a complete and easy-to-query data structure.

[0077] By constructing such a dictionary search tree, the growth characteristic model diagram and growth water requirement range data of pepper seedlings at different growth stages can be efficiently stored and managed. In practical applications, when it is necessary to query the relevant information of pepper seedlings at a certain growth stage, the corresponding dictionary fork can be quickly and accurately located to obtain the growth characteristic model diagram and growth water requirement range data of the pepper seedlings. This helps to accurately determine the current growth stage of the pepper seedlings, and then formulate reasonable greenhouse control strategies such as irrigation based on their growth water requirement range data, improve the growth quality and yield of pepper seedlings, and realize intelligent pepper cultivation management.

[0078] Preferably, the real-time growth state model diagram is imported into the dictionary search tree for search and matching, and the growth water requirement range data of the pepper seedlings in the target greenhouse at the current growth stage is obtained, specifically:

[0079] Obtain the real-time growth status model diagram of the pepper seedlings in the target greenhouse, and obtain the growth characteristic model diagram on each dictionary branch in the dictionary search tree;

[0080] Performing similarity analysis on the real-time growth state model graph and the growth characteristic model graphs on each dictionary fork branch to obtain the similarity between the real-time growth state model graph and the growth characteristic model graphs on each dictionary fork branch;

[0081] Sorting the similarities between the real-time growth state model graph and the growth feature model graphs on each dictionary branch to obtain the maximum similarity;

[0082] The growth characteristic model diagram corresponding to the maximum similarity is marked, the character string data of the dictionary fork of the marked growth characteristic model diagram is obtained, and the obtained character string data is interpreted to obtain the growth water requirement range data of the pepper seedlings in the target greenhouse at the current growth stage.

[0083] It should be noted that the real-time growth state model diagram of the pepper seedlings in the target greenhouse is obtained. This model diagram contains the contour, color, texture and other feature information of the pepper seedlings, which is a comprehensive presentation of the actual growth state of the pepper seedlings in the current greenhouse. At the same time, the growth feature model diagrams on each dictionary fork branch in the dictionary search tree are obtained. These growth feature model diagrams are pre-constructed according to the pepper seedlings at different growth stages, representing the characteristics that pepper seedlings should have at each typical growth stage. The real-time growth state model diagram and the growth feature model diagrams on each dictionary fork branch are analyzed for similarity. This process involves a variety of algorithms to quantify the similarity between the two. For example, the similarity of the two model diagrams in terms of shape structure (contour features), color distribution (color features) and texture features is compared. By calculating the difference or matching degree of these features, the similarity between the real-time growth state model diagram and the growth feature model diagrams on each dictionary fork branch is obtained. The obtained similarities are sorted by size to find the maximum similarity. The growth feature model diagram on the dictionary fork corresponding to this maximum similarity is closest to the actual growth state of the pepper seedlings in the current greenhouse. This sorting operation can select the model that best matches the current state of the pepper seedlings from a large number of predefined growth stage models. Mark the growth feature model graph corresponding to the maximum similarity, and then obtain the string data of the dictionary branch where the growth feature model graph is located. This string data is converted from the growth water requirement range data previously. By interpreting it, the growth water requirement range data of the pepper seedlings in the target greenhouse at the current growth stage can be obtained. This data is the key basis for subsequent greenhouse control operations such as soil moisture management.

[0084] By importing the real-time growth status model graph into the dictionary search tree for search and matching operations, the current growth stage of the pepper seedlings in the target greenhouse can be accurately determined, and then the growth water requirement range data corresponding to the growth stage can be obtained. This process realizes the accurate mapping from the actual growth state of the pepper seedlings to their growth needs, providing a basis for the precise irrigation and scientific management of the pepper seedlings in the greenhouse. It can avoid the unreasonable water supply caused by inaccurate judgment of the growth stage of the pepper seedlings, help improve the growth quality of the pepper seedlings, reduce the occurrence of diseases and pests, and ultimately improve the yield and quality of peppers.

[0085] Preferably, real-time moisture data of the soil in the target greenhouse is collected at several preset time nodes, and the soil moisture suitability deviation coefficient of the pepper seedlings in the target greenhouse is obtained based on the collected real-time moisture data and the growth moisture requirement range data combined with the Bezier curve algorithm analysis, specifically:

[0086] The real-time moisture data of the soil in the target greenhouse is collected at several preset time nodes, the Bezier curve algorithm is introduced, and the real-time moisture data collected at each preset time node is used as a discrete point;

[0087] Connect each discrete point in sequence according to each acquisition time node to construct the Bezier curve of soil moisture dynamics in the target greenhouse;

[0088] Determine an upper constraint boundary and a lower constraint boundary according to the growth water requirement range data of the pepper seedlings in the target greenhouse at the current growth stage, and divide a constraint area in the water dynamic Bezier curve according to the upper constraint boundary and the lower constraint boundary;

[0089] Calculating the total curve length of the moisture dynamic Bezier curve outside the constraint area, and calculating the total curve length of the moisture dynamic Bezier curve within the constraint area;

[0090] The total curve length of the moisture dynamic Bezier curve outside the constraint area is processed by ratio with the total curve length of the moisture dynamic Bezier curve within the constraint area to obtain the soil moisture suitability deviation coefficient of the pepper seedlings in the target greenhouse.

[0091] It should be noted that the real-time moisture data of the soil in the target greenhouse is collected at several preset time nodes, which ensures that the dynamic information of soil moisture at different times can be obtained. The Bezier curve algorithm is introduced, and these collected real-time moisture data are used as discrete points to prepare for the construction of a curve that can reflect the trend of soil moisture over time. The Bezier curve of the soil moisture dynamics in the target greenhouse is constructed by connecting each discrete point in sequence according to each acquisition time node. The Bezier curve has the characteristics of smoothness and flexibility, and can fit these discrete moisture data points well, thereby intuitively showing the change process of soil moisture over time, whether it is a gradual increase, decrease or fluctuation.

[0092] The upper constraint boundary and the lower constraint boundary are determined according to the growth water requirement range data of the pepper seedlings in the target greenhouse at the current growth stage. This growth water requirement range is obtained by judging the growth stage of the pepper seedlings before, and is the soil moisture range required for the healthy growth of the pepper seedlings. According to these two boundaries, a constraint area is divided in the water dynamic Bezier curve, which represents the soil moisture in the range suitable for the growth of pepper seedlings. The total curve length of the water dynamic Bezier curve outside the constraint area and the total curve length in the constraint area are calculated. The length of the curve outside the constraint area reflects the degree of deviation of the soil moisture from the suitable range, while the length in the constraint area indicates the duration or proportion of the soil moisture in the suitable range. The total curve length outside the constraint area is processed by ratio with the total curve length in the constraint area to obtain the soil moisture suitability deviation coefficient of the pepper seedlings in the target greenhouse.

[0093] The soil moisture suitability deviation coefficient obtained through the above operation can quantitatively indicate the degree to which the soil moisture in the target greenhouse deviates from the suitable range of the pepper seedlings at the current growth stage. This coefficient provides an intuitive and effective indicator for evaluating soil moisture conditions and can be used to determine whether the soil moisture in the greenhouse needs to be regulated.

[0094] Preferably, the moisture condition of the soil in the target greenhouse is analyzed according to the soil moisture suitability deviation coefficient, and the target greenhouse is regulated and processed according to the analysis result, specifically:

[0095] The soil moisture suitability deviation coefficient of the pepper seedlings in the target greenhouse is compared with the preset coefficient threshold;

[0096] If the soil moisture suitability deviation coefficient of the pepper seedlings in the target greenhouse is not greater than the preset coefficient threshold, it means that the moisture conditions of the soil in the target greenhouse meet the growth requirements of the pepper seedlings in the current growth stage, and the target greenhouse will not be regulated.

[0097] It should be noted that a preset coefficient threshold is set, which is a measurement standard determined in advance based on the growth requirements of pepper seedlings, past experience or experimental data. The calculated soil moisture suitability deviation coefficient is compared with this preset coefficient threshold. This comparison process is a key step in determining whether the soil moisture condition needs to be adjusted. If the soil moisture suitability deviation coefficient is not greater than the preset coefficient threshold, this means that the fluctuation of soil moisture or the degree of deviation from the suitable range is within an acceptable range. In practical terms, the moisture conditions of the soil in the target greenhouse at this time basically meet the growth requirements of the pepper seedlings at the current growth stage. In this case, the target greenhouse is not regulated, which reflects a precise management strategy, avoids unnecessary intervention, reduces the waste of manpower and material resources, and also maintains a relatively stable growth environment in the greenhouse.

[0098] By comparing the soil moisture suitability deviation coefficient with the preset coefficient threshold, it is possible to accurately determine whether the soil moisture conditions in the target greenhouse meet the needs of the pepper seedlings at the current growth stage, thereby deciding whether to regulate the greenhouse. This method realizes intelligent decision-making based on data, reduces unnecessary regulation operations, helps maintain the stability of the soil moisture environment in the greenhouse, provides suitable conditions for the healthy growth of pepper seedlings, and improves the growth quality and final yield of pepper seedlings.

[0099] Preferably, the moisture condition of the soil in the target greenhouse is analyzed according to the soil moisture suitability deviation coefficient, and the target greenhouse is regulated and processed according to the analysis result, specifically:

[0100] If the soil moisture suitability deviation coefficient of the pepper seedlings in the target greenhouse is greater than the preset coefficient threshold, it means that the moisture condition of the soil in the target greenhouse does not meet the growth requirements of the pepper seedlings in the current growth stage, then the average value of the real-time moisture data collected at several preset time nodes is calculated to obtain the average moisture data;

[0101] The average value of the moisture data is compared with a preset threshold value. If the average value of the moisture data is greater than the preset threshold value, the soil in the target greenhouse is determined to be in a drought state; otherwise, the soil in the target greenhouse is determined to be in an over-humid state.

[0102] Preferably, the moisture condition of the soil in the target greenhouse is analyzed according to the soil moisture suitability deviation coefficient, and the target greenhouse is regulated according to the analysis result, further comprising the following steps:

[0103] If the soil in the target greenhouse is too wet, the real-time rainfall in the preset range of the target greenhouse is obtained. If the real-time rainfall in the preset range of the target greenhouse is greater than the preset rainfall, the roof of the target greenhouse is controlled to be closed to achieve the rain-proof function;

[0104] If the real-time rainfall within the preset range of the target greenhouse is not greater than the preset rainfall, the real-time light intensity within the preset range of the target greenhouse is obtained; if the real-time light intensity within the preset range of the target greenhouse is greater than the preset intensity value, the roof of the target greenhouse is controlled to open to realize the natural sunlight dehumidification function.

[0105] Preferably, the moisture condition of the soil in the target greenhouse is analyzed according to the soil moisture suitability deviation coefficient, and the target greenhouse is regulated according to the analysis result, further comprising the following steps:

[0106] If the soil in the target greenhouse is in a drought condition, the real-time rainfall in the preset range of the target greenhouse is obtained. If the real-time rainfall in the preset range of the target greenhouse is greater than the preset rainfall, the roof of the target greenhouse is controlled to open to realize the natural rainwater irrigation function;

[0107] If the real-time rainfall within the preset range of the target greenhouse is not greater than the preset rainfall, the real-time light intensity within the preset range of the target greenhouse is obtained; if the real-time light intensity within the preset range of the target greenhouse is greater than the preset intensity value, the roof of the target greenhouse is controlled to be closed to enhance the moisturizing function of the target greenhouse.

[0108] It should be noted that when the soil moisture suitability deviation coefficient is greater than the preset coefficient threshold, it indicates that the soil moisture conditions do not meet the current growth requirements of the pepper seedlings. At this time, the average value of the real-time moisture data collected at the preset time node is calculated, and the soil is judged to be dry or too wet by comparing it with the preset threshold. This step converts the abstract concept of the deviation coefficient into a judgment of the actual soil moisture condition (drought or over-wet), providing a basis for subsequent precise regulation.

[0109] If the soil is too wet, first use the sensor to obtain the real-time rainfall in the preset range of the target greenhouse. When the rainfall is greater than the preset rainfall, close the roof to block the rain and prevent more moisture from entering the greenhouse and aggravating the over-humidity. This is an active defense measure based on external rainfall conditions to prevent further increase in soil moisture. If the rainfall is not greater than the preset rainfall, obtain the real-time light intensity. When the light intensity is greater than the preset intensity value, open the roof to use natural sunlight for dehumidification. This is an effective way to use natural sunlight and air circulation to reduce soil moisture, and to adjust the humidity by reasonably controlling the state of the greenhouse roof.

[0110] When the soil is in a drought condition, also pay attention to the real-time rainfall. If the rainfall is greater than the preset rainfall, open the roof to allow natural rainwater to irrigate the soil. Using natural precipitation to replenish soil moisture is an energy-saving and environmentally friendly irrigation method. When the rainfall is not greater than the preset rainfall, check the real-time light intensity. If the light intensity is greater than the preset intensity value, close the roof to enhance the moisturizing function and reduce soil moisture evaporation to maintain the existing soil moisture content and avoid worsening of the drought.

[0111] It should also be noted that if the soil is too wet, the ventilation system, drainage system, etc. in the greenhouse can be controlled to open; if the soil is dry, the irrigation system, etc. can be controlled to open.

[0112] Through the above series of operations, the soil moisture status is judged according to the soil moisture suitability deviation coefficient, and then the target greenhouse is precisely regulated according to different conditions (drought or over-humidity) and external environmental factors (rainfall, light intensity). This regulation method realizes the intelligent management of greenhouse soil moisture, can effectively deal with the situation of over-humidity or drought of soil, and makes full use of natural conditions (rainfall, sunlight) to regulate soil moisture, reduce human intervention and improve the stability of the growth environment of pepper seedlings, thus helping to improve the growth quality and yield of pepper seedlings.

[0113] The pepper cultivation rain shelter greenhouse control method may further comprise the following steps:

[0114] Obtain pest feature images corresponding to pepper seedlings infected with various types of pests through a big data network, perform feature transformation processing on the pest feature images, and obtain a wavelet coefficient matrix of pepper seedlings infected with various types of pests;

[0115] Constructing a knowledge graph, importing the wavelet coefficient matrix of pepper seedlings infected with various types of pests into the knowledge graph, and regularly updating the knowledge graph;

[0116] Acquire the actual growth rate data of the pepper seedlings in each sub-area of ​​the target greenhouse within a preset time period; compare the actual growth rate data of the pepper seedlings in each sub-area within the preset time period with the preset growth rate;

[0117] Obtaining a sub-region corresponding to an actual growth rate data not greater than a preset growth rate, and marking it as an abnormal growth sub-region; obtaining growth image information of pepper seedlings in the abnormal growth sub-region, performing matrix transformation processing on the growth image information, and obtaining a wavelet coefficient matrix of the pepper seedlings in the abnormal growth sub-region;

[0118] Calculating the structural similarity index between the wavelet coefficient matrix of the pepper seedlings in the abnormal growth sub-region and the wavelet coefficient matrix of the pepper seedlings infected with various types of pests to obtain a plurality of structural similarity indexes;

[0119] Each structural similarity index is compared with a preset similarity index value; if there is a case where at least one structural similarity index is greater than the preset similarity index value, the abnormal growth sub-region is marked as an insect pest infection region, and the insect pest type of the insect pest infection region is matched in the knowledge graph based on the structural similarity index being greater than the preset similarity index value.

[0120] It should be noted that the wavelet coefficient matrix can effectively represent the characteristic information in the pest feature image, which can capture the details and texture information of the image at different scales and directions. The knowledge graph is a semantic network that can organize the wavelet coefficient matrices of different types of pests in a structured way to facilitate subsequent query and matching. The knowledge graph is updated regularly to ensure that it contains the latest pest feature information to adapt to the ever-changing pest situation.

[0121] When the actual growth rate of a sub-region is not greater than the preset growth rate, the sub-region is marked as an abnormal growth sub-region. This is based on the expectation of the normal growth rate of pepper seedlings. If the growth rate is lower than expected, there are likely to be factors that affect growth, such as pests, diseases or other environmental problems. The structural similarity index can quantitatively represent the degree of structural similarity between the two matrices. After obtaining several structural similarity indices, they are compared with the preset similarity index values. If there is at least one structural similarity index greater than the preset similarity index value, the abnormal growth sub-region is marked as a pest-infected region, and the pest type is matched in the knowledge graph based on this situation greater than the preset similarity index value. This process accurately determines the pest type by comparing the similarity between the image features of the abnormal growth area and the known pest features. By analyzing the images of pepper seedlings in abnormal growth areas, using the wavelet coefficient matrix and structural similarity index to match the pest features in the knowledge graph, it is possible to accurately determine whether the area is infected by pests and the type of infected pests. This enables intelligent monitoring and precise judgment of pepper seedling pests, helps to take targeted prevention and control measures in a timely manner, reduces the impact of pests on the growth of pepper seedlings, and improves the yield and quality of peppers.

[0122] The pepper cultivation rain shelter greenhouse control method may further comprise the following steps:

[0123] Retrieving the reproduction rate of the pest-infected area under various environmental characteristic conditions in the pest-infected area from a big data network according to the pest type in the pest-infected area;

[0124] Taking the environmental characteristic conditions as conditional nodes and the corresponding reproduction rates as variable nodes, generating a network topology graph according to the conditional nodes and the variable nodes, constructing a graph-embedded network model, embedding the network topology graph into the graph-embedded network model, and obtaining a pest reproduction rate prediction model;

[0125] Acquire real-time environmental characteristic conditions of the pest-infected area at a number of time points, import the real-time environmental characteristic conditions acquired at each time point into the pest reproduction rate prediction model, and obtain the predicted reproduction rate of the pests in the pest-infected area at each time point;

[0126] Performing weighted average processing on the predicted reproduction rate of pests in the pest-infected area at each time point to obtain the predicted reproduction rate of pests in the pest-infected area within a preset time period;

[0127] If the predicted reproduction rate of the pests in the pest-infected area within the preset time period is greater than the preset rate value, then the reproduction-inhibiting environmental conditions that have a reproduction-inhibiting effect on the pests in the pest-infected area are retrieved from the big data network according to the pest type in the pest-infected area;

[0128] The corresponding environmental control system in the target greenhouse is controlled and processed according to the reproduction inhibition environmental condition, so that the environment in the insect pest infection area reaches the reproduction inhibition environmental condition.

[0129] It should be noted that, first, the reproduction rate under different environmental characteristic conditions is retrieved in the big data network according to the type of pests in the pest-infected area. The environmental characteristic conditions here cover a variety of factors that may affect the reproduction of pests, such as temperature, humidity, and light. The environmental characteristic conditions are used as conditional nodes, and the corresponding reproduction rates are used as variable nodes to construct a network topology graph, and then a graph-embedded network model is generated to obtain a pest reproduction rate prediction model. This model construction method can effectively capture the complex relationship between environmental factors and pest reproduction rates. By embedding the network topology graph into the graph-embedded network model, the model can accurately predict the pest reproduction rate according to the input environmental characteristic conditions. The real-time environmental characteristic conditions of the pest-infected area are obtained at multiple time nodes, and they are imported into the pest reproduction rate prediction model to obtain the predicted reproduction rate of the pest at each time node. Since the environmental conditions at different times may be different, predictions are made separately at each time node. Then these predicted reproduction rates are weighted averaged to obtain the predicted reproduction rate within the preset time period. The weighted average takes into account the importance or representativeness of different time nodes, so that the prediction results can better reflect the pest reproduction trend within the entire preset time period. If the predicted reproduction rate of the pest within the preset time period is greater than the preset rate value, it indicates that the pest has a tendency to reproduce rapidly and measures need to be taken to suppress it. At this time, according to the type of pest, the reproduction inhibition environmental conditions that have a reproduction inhibition effect on the pest are retrieved in the big data network. For example, the reproduction of some pests will be inhibited under low temperature, low humidity or specific light conditions. Finally, according to the retrieved reproduction inhibition environmental conditions, the environmental control system of the target greenhouse is controlled so that the environment in the pest-infected area reaches the reproduction inhibition environmental conditions, thereby effectively controlling the reproduction of pests and reducing the harm of pests to pepper cultivation. According to the prediction results, when the reproduction rate of pests may be too high, the reproduction inhibition environmental conditions are obtained and implemented in time, and the reproduction of pests is suppressed by regulating the greenhouse environment. The forward-looking management of pest reproduction in the process of pepper cultivation is realized, which helps to reduce the damage of pests to pepper seedlings, improve the yield and quality of peppers, and reduce the use of chemical pesticides, so as to achieve more environmentally friendly and efficient pepper cultivation management.

[0130] The present invention also discloses a rain shelter greenhouse control system based on intelligent perception, such as Figure 3 As shown, the rain shelter cultivation greenhouse control system includes a memory 10 and a processor 20, and the memory 10 stores a rain shelter cultivation greenhouse control method program. When the rain shelter cultivation greenhouse control method program is executed by the processor 20, any one of the rain shelter cultivation greenhouse control method steps is implemented.

[0131] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A rain shelter greenhouse control method based on intelligent perception, characterized in that: The following steps are involved: Acquire a real-time growth status image of the pepper seedlings in the target greenhouse, and construct a real-time growth status model diagram of the pepper seedlings in the target greenhouse according to the real-time growth status image; The growth characteristic images and growth water requirement range data corresponding to the pepper seedlings at different growth stages are obtained through the big data network, and a dictionary search tree is constructed according to the growth characteristic images and growth water requirement range data corresponding to the pepper seedlings at different growth stages; Importing the real-time growth state model diagram into the dictionary search tree for search and matching, and obtaining the growth water requirement range data of the pepper seedlings in the target greenhouse at the current growth stage; At a number of preset time points, real-time moisture data of the soil in the target greenhouse is collected, and the soil moisture suitability deviation coefficient of the pepper seedlings in the target greenhouse is obtained by combining the collected real-time moisture data with the growth moisture requirement range data and the Bezier curve algorithm analysis; The moisture condition of the soil in the target greenhouse is analyzed according to the soil moisture suitability deviation coefficient, and the target greenhouse is regulated and processed according to the analysis result.

2. The control method of a rain shelter cultivation greenhouse based on intelligent perception according to claim 1 is characterized in that: A real-time growth state model diagram of pepper seedlings in the target greenhouse is constructed according to the real-time growth state image, specifically: A Canny edge detection algorithm is introduced to perform feature extraction on the real-time growth status image based on the Canny edge detection algorithm to obtain contour feature information of each part of the pepper seedling; Convert the real-time growth status image from the RGB color space to the HSV color space, and extract the color feature information of each part of the pepper seedling in the HSV space; A gray-level co-occurrence matrix extraction algorithm is introduced, and the gray-level co-occurrence frequencies of pixel pairs in each preset direction in the real-time growth status image are extracted based on the gray-level co-occurrence matrix extraction algorithm, and the texture feature information of each part of the pepper seedling is determined according to the obtained gray-level co-occurrence frequencies; According to the contour feature information, color feature information and texture feature information of each part of the pepper seedlings and combined with three-dimensional software, a real-time growth status model diagram of the pepper seedlings in the target greenhouse is constructed.

3. The control method of a rain shelter cultivation greenhouse based on intelligent perception according to claim 1 is characterized in that: The growth characteristic images and growth water requirement range data corresponding to pepper seedlings at different growth stages are obtained through the big data network, and a dictionary search tree is constructed according to the growth characteristic images and growth water requirement range data corresponding to pepper seedlings at different growth stages, specifically: Acquire growth characteristic images and growth water requirement range data corresponding to pepper seedlings at different growth stages through a big data network, and construct growth characteristic model diagrams corresponding to pepper seedlings at different growth stages according to the growth characteristic images; Constructing a dictionary tree, and dividing a number of dictionary branches in the dictionary tree according to different growth stages of peppers; Obtain the growth water requirement range data corresponding to the pepper seedlings at different growth stages, and convert the growth water requirement range data corresponding to the pepper seedlings at different growth stages into specific string data; Map each string data to the root node of the corresponding dictionary branch, and complete the coding operation on each dictionary branch; The growth characteristic model graphs corresponding to the pepper seedlings at different growth stages are stored in the leaf nodes of the corresponding dictionary fork branches to obtain a dictionary search tree.

4. The method for controlling a rain shelter cultivation greenhouse based on intelligent perception according to claim 1, characterized in that: The real-time growth state model diagram is imported into the dictionary search tree for search and matching, and the growth water requirement range data of the pepper seedlings in the target greenhouse at the current growth stage is obtained, specifically: Obtain the real-time growth status model diagram of the pepper seedlings in the target greenhouse, and obtain the growth characteristic model diagram on each dictionary branch in the dictionary search tree; Performing similarity analysis on the real-time growth state model graph and the growth characteristic model graphs on each dictionary fork branch to obtain the similarity between the real-time growth state model graph and the growth characteristic model graphs on each dictionary fork branch; Sorting the similarities between the real-time growth state model graph and the growth feature model graphs on each dictionary branch to obtain the maximum similarity; The growth characteristic model diagram corresponding to the maximum similarity is marked, the character string data of the dictionary fork of the marked growth characteristic model diagram is obtained, and the obtained character string data is interpreted to obtain the growth water requirement range data of the pepper seedlings in the target greenhouse at the current growth stage.

5. The control method of a rain shelter cultivation greenhouse based on intelligent perception according to claim 1 is characterized in that: At several preset time points, real-time soil moisture data in the target greenhouse is collected, and the soil moisture suitability deviation coefficient of the pepper seedlings in the target greenhouse is obtained based on the collected real-time moisture data and the growth moisture requirement range data combined with the Bezier curve algorithm analysis, which is specifically: The real-time moisture data of the soil in the target greenhouse is collected at several preset time nodes, the Bezier curve algorithm is introduced, and the real-time moisture data collected at each preset time node is used as a discrete point; The discrete points are connected in sequence according to the acquisition time nodes to construct the Bezier curve of soil moisture dynamics in the target greenhouse; Determine an upper constraint boundary and a lower constraint boundary according to the growth water requirement range data of the pepper seedlings in the target greenhouse at the current growth stage, and divide a constraint area in the water dynamic Bezier curve according to the upper constraint boundary and the lower constraint boundary; Calculating the total curve length of the moisture dynamic Bezier curve outside the constraint area, and calculating the total curve length of the moisture dynamic Bezier curve within the constraint area; The total curve length of the moisture dynamic Bezier curve outside the constraint area is processed by ratio with the total curve length of the moisture dynamic Bezier curve within the constraint area to obtain the soil moisture suitability deviation coefficient of the pepper seedlings in the target greenhouse.

6. The method for controlling a rain shelter greenhouse based on intelligent perception according to claim 5, characterized in that: The moisture condition of the soil in the target greenhouse is analyzed according to the soil moisture suitability deviation coefficient, and the target greenhouse is regulated and processed according to the analysis result, specifically: The soil moisture suitability deviation coefficient of the pepper seedlings in the target greenhouse is compared with the preset coefficient threshold; If the soil moisture suitability deviation coefficient of the pepper seedlings in the target greenhouse is not greater than the preset coefficient threshold, it means that the moisture conditions of the soil in the target greenhouse meet the growth requirements of the pepper seedlings in the current growth stage, and the target greenhouse will not be regulated.

7. The control method of a rain shelter cultivation greenhouse based on intelligent perception according to claim 5 is characterized in that: The moisture condition of the soil in the target greenhouse is analyzed according to the soil moisture suitability deviation coefficient, and the target greenhouse is regulated and processed according to the analysis result, specifically: If the soil moisture suitability deviation coefficient of the pepper seedlings in the target greenhouse is greater than the preset coefficient threshold, it means that the moisture condition of the soil in the target greenhouse does not meet the growth requirements of the pepper seedlings in the current growth stage, then the average value of the real-time moisture data collected at several preset time nodes is calculated to obtain the average moisture data; The average value of the moisture data is compared with a preset threshold value. If the average value of the moisture data is greater than the preset threshold value, the soil in the target greenhouse is determined to be in a drought state; otherwise, the soil in the target greenhouse is determined to be in an over-humid state.

8. The method for controlling a rain shelter cultivation greenhouse based on intelligent perception according to claim 7, characterized in that: The moisture condition of the soil in the target greenhouse is analyzed according to the soil moisture suitability deviation coefficient, and the target greenhouse is regulated according to the analysis result, further comprising the following steps: If the soil in the target greenhouse is too wet, the real-time rainfall in the preset range of the target greenhouse is obtained. If the real-time rainfall in the preset range of the target greenhouse is greater than the preset rainfall, the roof of the target greenhouse is controlled to be closed to achieve the rain-proof function; If the real-time rainfall within the preset range of the target greenhouse is not greater than the preset rainfall, the real-time light intensity within the preset range of the target greenhouse is obtained; if the real-time light intensity within the preset range of the target greenhouse is greater than the preset intensity value, the roof of the target greenhouse is controlled to open to realize the natural sunlight dehumidification function.

9. The method for controlling a rain shelter cultivation greenhouse based on intelligent perception according to claim 7, characterized in that: The moisture condition of the soil in the target greenhouse is analyzed according to the soil moisture suitability deviation coefficient, and the target greenhouse is regulated according to the analysis result, further comprising the following steps: If the soil in the target greenhouse is in a drought condition, the real-time rainfall in the preset range of the target greenhouse is obtained. If the real-time rainfall in the preset range of the target greenhouse is greater than the preset rainfall, the roof of the target greenhouse is controlled to open to realize the natural rainwater irrigation function; If the real-time rainfall within the preset range of the target greenhouse is not greater than the preset rainfall, the real-time light intensity within the preset range of the target greenhouse is obtained; if the real-time light intensity within the preset range of the target greenhouse is greater than the preset intensity value, the roof of the target greenhouse is controlled to be closed to enhance the moisturizing function of the target greenhouse.

10. A rain shelter greenhouse control system based on intelligent perception, characterized in that: The rain shelter cultivation greenhouse control system includes a memory and a processor, wherein the memory stores a rain shelter cultivation greenhouse control method program, and when the rain shelter cultivation greenhouse control method program is executed by the processor, the rain shelter cultivation greenhouse control method steps as described in any one of claims 1 to 9 are implemented.

Images