A soil humidity threshold regulation method and system for efficient cultivation of red plum apricot

Abstract

The application discloses a soil humidity threshold regulation method for efficient cultivation of red plum apricot, and relates to the technical field of soil humidity threshold regulation. The method comprises the following steps: acquiring soil type, root distribution depth and climate conditions, combining a preset red plum apricot growth stage database, determining soil water retention capacity and initial water use efficiency, classifying growth stages by using a k-means clustering algorithm, and acquiring initial humidity thresholds matched with soil and climate. If the climate condition deviation exceeds a threshold value, the humidity threshold value is adjusted by using a linear regression algorithm to ensure adaptation to dynamic environmental changes. Based on the adjusted humidity threshold value and root distribution, a decision tree algorithm is used to evaluate waterlogging risk, a dynamic drainage scheme is generated in combination with a drainage strategy database, and finally the irrigation amount is adjusted in real time by using a proportional-integral-derivative algorithm. The application can dynamically optimize irrigation and drainage strategies, realize precise water management, improve the growth efficiency of red plum apricot, and reduce waterlogging risk.

Description

Technical Field

[0001] This invention relates to the field of soil moisture threshold regulation technology, and in particular to a method and system for regulating soil moisture threshold for the efficient cultivation of red plum apricot. Background Technology

[0002] As a fruit tree with high economic value, the efficient cultivation of the red plum apricot is crucial for agricultural production. Soil moisture is a key factor affecting the growth and development of the red plum apricot, directly related to the root system's ability to absorb nutrients and the health of the tree.

[0003] Existing soil moisture control methods rely on single environmental parameters, neglecting the interaction of factors such as soil type, climate conditions, tree age structure, and root depth, resulting in low control precision. Soil type and root depth are two key technical attributes. Differences in soil type directly affect water retention and transport efficiency, while root depth determines the actual utilization rate of water within the root zone. Soil type influences root growth patterns and distribution depth. Existing methods struggle to adjust humidity thresholds based on the dynamic changes of these factors, potentially leading to stunted growth or root damage in red plum and apricot trees during rainy or dry seasons due to improper water management.

[0004] How to dynamically adjust the soil moisture threshold based on the interaction between soil type and root distribution depth to adapt to the needs of different growth stages of red plum apricot has become a key issue for efficient cultivation.

[0005] It should be noted that the information disclosed in the background section above is only used to enhance the understanding of the background of the present invention, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0006] In view of this, the present invention provides a method and system for controlling soil moisture threshold for efficient cultivation of red plum apricot, in order to solve the problem that existing technologies are difficult to accurately control soil moisture threshold, and to realize the dynamic adjustment of soil moisture threshold according to the interaction of soil type and root distribution depth, so as to adapt to the needs of different growth stages of red plum apricot.

[0007] This invention provides a method for controlling soil moisture threshold for high-efficiency cultivation of red plum apricot, comprising:

[0008] By acquiring soil type, root distribution depth and climate conditions, and combining them with a pre-set database of red plum apricot growth stages, soil water retention capacity and initial water use efficiency can be obtained.

[0009] Based on the soil water retention capacity and initial water use efficiency, the k-means clustering algorithm was used to obtain the classification results of the growth stages of the red plum apricot.

[0010] Based on the growth stage classification results and the preset humidity threshold database, the initial humidity threshold corresponding to the soil type and local climate type is matched to obtain the initial humidity control range.

[0011] Compare real-time rainfall or evaporation data with the expected climate values ​​corresponding to the initial humidity control range. When the deviation exceeds the preset deviation range, analyze the climate conditions using a linear regression algorithm to obtain the adjusted humidity control range.

[0012] Based on the adjusted humidity control range and root distribution depth, the water accumulation risk level of the red plum apricot is obtained. Based on the water accumulation risk level, a dynamic drainage plan is generated in conjunction with the drainage strategy database.

[0013] Based on the dynamic drainage scheme and the adjusted humidity control range, the irrigation amount is dynamically adjusted using a proportional-integral-differential algorithm to obtain precise irrigation parameters for each tree.

[0014] In some optional embodiments, the step of obtaining the growth stage classification results of the red plum apricot using a k-means clustering algorithm based on the soil water retention capacity and initial water use efficiency further includes:

[0015] Soil moisture content and root depth information are collected by sensors, and combined with the soil water retention capacity and initial water use efficiency, a soil environment dataset for red plum and apricot trees is obtained.

[0016] Based on the soil environment dataset of the red plum and apricot trees, the initial water use efficiency and root distribution characteristics of each red plum and apricot tree at different growth stages were calculated to obtain a feature vector set.

[0017] The feature vector set was clustered using the k-means clustering algorithm, and through further iterative optimization, the classification results of the growth stages of the red plum apricot were obtained.

[0018] In some optional embodiments, the step of obtaining soil type, root distribution depth, and climate conditions, combined with a preset database of red plum apricot growth stages, to obtain soil water retention capacity and initial water use efficiency, further includes:

[0019] Acquire data on soil type, root distribution depth, and climate conditions collected by sensors;

[0020] Based on a pre-set database of red plum apricot growth stages, soil characteristic parameters matching soil type and growth stage are queried to obtain soil water retention capacity.

[0021] Based on the soil water retention capacity and root distribution depth, the initial water use efficiency is obtained using the random forest algorithm.

[0022] In some optional embodiments, the step of matching the initial humidity threshold corresponding to the soil type and local climate type based on the growth stage classification results and a preset humidity threshold database to obtain the initial humidity control range further includes:

[0023] Based on the growth stage classification results and the preset humidity threshold database, the preliminary humidity threshold for each tree at the current growth stage is obtained.

[0024] By matching the initial humidity threshold with the soil type and local climate type, the initial humidity control range for each tree is obtained.

[0025] In some optional embodiments, the step of comparing real-time rainfall or evaporation data with the expected climate value corresponding to the initial humidity control range, and obtaining the adjusted humidity control range by analyzing climate conditions through a linear regression algorithm when the deviation exceeds a preset deviation range, further includes:

[0026] Compare the real-time rainfall and evaporation with the expected climate values ​​corresponding to the initial humidity control range. If the deviation exceeds a preset deviation threshold, obtain the rainfall and evaporation from the real-time monitoring data.

[0027] The effects of rainfall and evaporation on environmental humidity were analyzed using a linear regression algorithm, and the data analysis results were obtained.

[0028] Based on the data analysis results, the initial humidity range is adjusted to generate the adjusted humidity control range.

[0029] In some optional embodiments, the step of obtaining the waterlogging risk level of the red plum apricot based on the adjusted humidity control range and root distribution depth, and generating a dynamic drainage plan based on the waterlogging risk level and in conjunction with a drainage strategy database, further includes:

[0030] If the soil moisture exceeds the preset moisture risk threshold, a decision tree algorithm is used to preliminarily classify the water accumulation risk of each tree based on the root distribution depth to obtain an initial risk level.

[0031] Clustering algorithms are used to group tree locations to obtain location grouping results. If the location grouping results show that the trees are located in low-lying areas, the risk prediction weights are adjusted according to the initial risk level to obtain a weighted risk level.

[0032] Based on the weighted risk level and soil moisture data, a logistic regression algorithm is used to perform a secondary verification of the water accumulation risk to obtain the final risk level.

[0033] Query the drainage parameters corresponding to soil type and root distribution in the preset drainage strategy database, analyze their correlation with the final risk level, and obtain optimized drainage parameters;

[0034] Based on the optimized drainage parameters and the root distribution of each tree, the allocation weight of the dynamic drainage scheme is calculated to obtain the weight distribution result. If the weight distribution result meets the preset weight distribution threshold, a dynamic drainage scheme for each tree is generated.

[0035] In some optional embodiments, the step of dynamically adjusting the irrigation amount using a proportional-integral-differential algorithm based on the dynamic drainage scheme and the adjusted humidity control range to obtain precise irrigation parameters for each tree further includes:

[0036] The soil moisture value is collected in real time by sensors and compared with the preset moisture irrigation threshold.

[0037] If the soil moisture value is lower than the preset moisture threshold, the irrigation adjustment coefficient is calculated by the proportional-integral-differential algorithm to obtain the irrigation adjustment value.

[0038] Based on the irrigation adjustment value and combined with the dynamic drainage scheme, the irrigation water allocation for each tree is calculated to obtain the precise irrigation parameters for each tree.

[0039] This invention provides a soil moisture threshold control system for high-efficiency cultivation of red plum apricot, comprising:

[0040] The data acquisition module is used to obtain soil type, root distribution depth and climate conditions, and combined with the preset red plum apricot growth stage database, to obtain soil water retention capacity and initial water use efficiency.

[0041] The growth classification module is used to obtain the growth stage classification results of the red plum apricot based on the soil water retention capacity and initial water use efficiency using the k-means clustering algorithm.

[0042] The threshold matching module is used to match the initial humidity threshold corresponding to the soil type and local climate type based on the growth stage classification results and the preset humidity threshold database, so as to obtain the initial humidity control range.

[0043] The dynamic adjustment module is used to compare real-time rainfall or evaporation data with the expected climate value corresponding to the initial humidity control range. When the deviation exceeds the preset deviation range, the climate conditions are analyzed through a linear regression algorithm to obtain the adjusted humidity control range.

[0044] The risk assessment module is used to obtain the water accumulation risk level of the red plum apricot based on the adjusted humidity control range and root distribution depth, and to generate a dynamic drainage plan based on the water accumulation risk level and the drainage strategy database.

[0045] The precision control module is used to dynamically adjust the irrigation amount based on the dynamic drainage scheme and the adjusted humidity control range, using a differential algorithm to obtain precise irrigation parameters for each tree.

[0046] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit the invention.

[0047] The soil moisture threshold control method and system for high-efficiency cultivation of red plum apricot of the present invention has the following beneficial effects:

[0048] This method obtains soil type, root depth, and climatic conditions, and combines this with a pre-defined database of red plum apricot growth stages to determine soil water retention capacity and initial water use efficiency. A k-means clustering algorithm is used to classify growth stages, obtaining initial humidity thresholds that match soil and climate. If climatic conditions deviate beyond the threshold, a linear regression algorithm is used to adjust the humidity threshold to ensure adaptation to dynamic environmental changes. Based on the adjusted humidity threshold and root distribution, a decision tree algorithm is used to assess waterlogging risk, and a dynamic drainage plan is generated using a drainage strategy database. Finally, a proportional-integral-differential algorithm is used to adjust irrigation volume in real time. This invention can dynamically optimize irrigation and drainage strategies, achieve precise water management, improve the growth efficiency of red plum apricots, and reduce the risk of waterlogging. Attached Figure Description

[0049] Other features, objects, and advantages of the invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings.

[0050] Figure 1 This is a flowchart of a soil moisture threshold control method for high-efficiency cultivation of red plum apricot according to an embodiment of the present invention;

[0051] Figure 2 This is a schematic diagram of a soil moisture threshold control system for high-efficiency cultivation of red plum apricot according to an embodiment of the present invention. Detailed Implementation

[0052] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, they are provided so that the invention will be more comprehensive and complete, and will fully convey the concept of the exemplary embodiments to those skilled in the art. The described features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.

[0053] Furthermore, the accompanying drawings are merely illustrative of the invention and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and therefore repeated descriptions of them will be omitted. Some block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities. These functional entities can be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor devices and / or microcontroller devices.

[0054] The flowchart shown in the attached diagram is merely an illustrative example and does not necessarily include all steps. For example, some steps may be broken down, while others may be combined or partially combined. Therefore, the actual execution order may change depending on the specific circumstances.

[0055] like Figure 1 As shown in the figure, this invention provides a method for controlling soil moisture threshold for efficient cultivation of red plum apricot. This method integrates multi-source data and multi-algorithm collaboration to dynamically optimize irrigation and drainage strategies, achieve precise water management, improve the growth efficiency of red plum apricot, and reduce the risk of waterlogging.

[0056] S11. Data Acquisition: Acquire soil type, root distribution depth, and climatic conditions. Combined with a pre-set database of red plum apricot growth stages, obtain soil water retention capacity and initial water use efficiency. Specifically, firstly, soil moisture content data is collected by soil moisture sensor layer by layer according to the soil profile. Combined with soil texture analysis algorithm, soil type and its corresponding water retention capacity are determined. Simultaneously, root distribution depth data is acquired through root scanner, and the distribution characteristics of roots at different depths are calculated based on the root density decay model. Climate condition data such as rainfall and evaporation are obtained from meteorological station API. Combined with stage water requirement data in the growth stage database, the initial water use efficiency is calculated using the water balance equation, providing a precise and dynamic data foundation for subsequent growth stage classification and humidity control.

[0057] S12. Growth Classification: Based on the soil water retention capacity and initial water use efficiency, the k-means clustering algorithm is used to obtain the growth stage classification results of the red plum apricot. Specifically, firstly, the soil water retention capacity and initial water use efficiency of each tree are constructed as a two-dimensional feature vector to form a sample dataset; by setting the number of clusters k=3, corresponding to the seedling stage, growth stage, and maturity stage, random centroids are initialized, and then the Euclidean distance is iteratively calculated, continuously optimizing the centroid position until convergence, dividing the trees into different clusters. The classification effectiveness is verified by calculating the ratio of the inter-cluster sum of squares to the intra-cluster sum of squares, thereby generating accurate growth stage labels for each tree, providing a basis for subsequent personalized humidity control.

[0058] S13. Threshold Matching: Based on the growth stage classification results and a preset humidity threshold database, the system matches the initial humidity thresholds corresponding to the soil type and local climate type to obtain the initial humidity control range. Specifically, the system first uses the growth stage classification label as the query keyword, combined with real-time acquired soil type and climate type data, to retrieve the initial humidity thresholds verified by historical data from the database. Then, a weighted average algorithm is used to dynamically adjust the initial thresholds by incorporating current environmental factors. The calculation formula comprehensively considers the influence of meteorological parameters such as rainfall and evaporation. Finally, fine-tuning is performed based on individual tree differences to generate an initial humidity control range that combines environmental adaptability with individual specificity, providing a core parameter benchmark for subsequent precision irrigation and risk control.

[0059] S14. Dynamic Adjustment: Compare real-time rainfall or evaporation data with the expected climate values ​​corresponding to the initial humidity control range. When the deviation exceeds a preset deviation range, analyze the climate conditions using a linear regression algorithm to obtain an adjusted humidity control range. Specifically, the system monitors rainfall and evaporation data in real time and compares them with the expected climate values ​​corresponding to the initial humidity control range. When the deviation between the monitored data and the expected values ​​exceeds a preset threshold, the system automatically triggers a linear regression analysis process: Based on historical climate datasets, a regression model is constructed with rainfall and evaporation as independent variables and environmental humidity as the dependent variable. By calculating the regression coefficients, a more accurate humidity prediction value under the current climate conditions is obtained, and the initial humidity control range is dynamically corrected accordingly. This generates an adjusted humidity control range that better reflects actual environmental changes, ensuring that the water management strategy can respond to climate fluctuations in real time and maintain the humidity stability of the red plum apricot growing environment.

[0060] S15: Risk Assessment: Based on the adjusted humidity control range and root distribution depth, the waterlogging risk level of the red plum apricot is obtained. Based on the waterlogging risk level, a dynamic drainage plan is generated in conjunction with the drainage strategy database. Specifically, firstly, based on real-time soil moisture data and root distribution characteristics, a preliminary risk classification is performed using preset humidity threshold rules. If the soil moisture content exceeds the set threshold and the root distribution is shallow, a high-risk judgment is triggered. Subsequently, a clustering algorithm is used to group and analyze the tree locations, and the risk weights of trees located in low-lying areas are adjusted. Then, a logistic regression algorithm is used for secondary verification to generate the final waterlogging risk level. Based on this risk level, the system matches drainage parameters corresponding to soil type and root distribution depth from the preset drainage strategy database. A personalized drainage plan is generated for each tree using a dynamic weight allocation algorithm, and corrected and optimized in conjunction with real-time environmental data. Finally, an automatically executable dynamic drainage control command is output to achieve precise and efficient waterlogging risk prevention and control.

[0061] S16. Precise Control: Based on the dynamic drainage scheme and the adjusted humidity control range, a differential algorithm is used to dynamically adjust the irrigation volume to obtain precise irrigation parameters for each tree. Specifically, the system uses a proportional-integral-differential algorithm to dynamically adjust the irrigation volume in real time based on the dynamic drainage scheme and the adjusted humidity control range, obtaining precise irrigation parameters for each tree. High-precision soil moisture sensors deployed in the root zone continuously collect data. When the current soil moisture is detected to be lower than the target threshold, the algorithm calculates the humidity deviation value and outputs a control quantity based on historical deviation accumulation and trends. The control quantity is converted into a specific irrigation volume and executed by the precision drip irrigation equipment. During irrigation, the output is adjusted in real time based on sensor feedback. Simultaneously, combined with the water drainage time estimated by the soil permeability model in the drainage scheme, the irrigation interval is intelligently arranged to form a closed-loop control. This ensures that soil moisture is stably maintained within the target range, achieving precise irrigation management based on the individual needs of each tree.

[0062] Through the above steps, this embodiment can dynamically adjust the soil moisture threshold according to the interaction between soil type and root distribution depth to adapt to the needs of different growth stages of red plum apricot.

[0063] In some embodiments, based on the above embodiments, the process of obtaining soil water retention capacity and initial water use efficiency by acquiring soil type, root distribution depth, and climate conditions, combined with a preset red plum apricot growth stage database, includes the following steps:

[0064] Data on soil type, root distribution depth, and climate conditions collected by sensors are acquired and stored as a structured dataset.

[0065] Based on a pre-set database of red plum apricot growth stages, soil characteristic parameters matching soil type and growth stage are queried to obtain soil water retention capacity.

[0066] Based on the soil water retention capacity and root distribution depth, the initial water use efficiency is obtained using the random forest algorithm.

[0067] The system collects data in real time through a network of multiple sensors deployed in the field. The soil moisture sensor uses a TDR-315L model, collecting data in layers from 0-100 cm depth, with each 10 cm layer representing a measurement. Assuming the measured moisture content at a depth of 0-30 cm is 15% for sandy soil and 25% for clay soil, soil particle size distribution data is obtained using a soil texture analyzer. According to the USDA soil classification standard, when the sand content greater than 0.05 mm exceeds 70% and the clay content less than 0.002 mm is less than 20%, the soil type is determined to be sandy loam. Its water retention capacity, calculated as 20% after determining the porosity to be 0.4 using a stereomicroscope, is also considered. A CI-600 model root scanner is used for scanning. Assuming the roots of the red plum apricot seedlings are mainly distributed at a depth of 0-40 cm, the root density formula is used...

[0068]

[0069] in In depth Root density at ground level, root density at ground surface =0.5 g / cm³, attenuation coefficient k=0.03, the root density at 40 cm is calculated to be 0.15 g / cm³. Climate data is acquired in real time through the meteorological station API interface. Assuming an average daily rainfall of 5 mm and evapotranspiration of 3 mm, the net water input is 2 mm / day. All data, after preprocessing, are stored in a structured format in a cloud database.

[0070] The system calls a pre-set database of red plum apricot growth stages for data matching. This database contains characteristic parameters of various soil types at different growth stages, with the seedling stage requiring 0.8 mm / day of water. Through data query and matching algorithms, when the system identifies sandy loam as the soil type and the tree as a seedling, the database returns a field capacity of 20% for the soil under these conditions. Subsequently, the water balance equation is used... The calculation is performed, where FC is 20% of field capacity. With a soil moisture content of 15%, ET as evapotranspiration of 3 mm / day, and D as drainage (assumed to be 0), the calculated water supply is 20% × 15% - 3 = 0 mm / day, which is different from the water requirement of 0.8 mm / day.

[0071] After obtaining data on soil water retention capacity and root distribution depth, the system uses a random forest algorithm to calculate the initial water use efficiency. The algorithm uses soil water retention capacity of 20%, root density of 0.15 g / cm³, and net water input of 2 mm / day as feature inputs. Through a trained prediction model, the initial water use efficiency is calculated to be (0.8 / 2) × 100% = 40%. This result indicates that sandy loam soil has a weak water retention capacity, requiring irrigation frequency optimization to 0.5 mm / day to match the root water absorption needs, providing a scientific basis for subsequent precision irrigation decisions. The entire process is fully automated, ensuring the accuracy and timeliness of data processing.

[0072] In some embodiments, based on the above embodiments, the process of obtaining the growth stage classification results of red plum apricot using the k-means clustering algorithm according to the soil water retention capacity and initial water use efficiency includes the following steps:

[0073] Soil moisture content and root depth information are collected by sensors, and combined with the soil water retention capacity and initial water use efficiency, a soil environment dataset for red plum and apricot trees is obtained.

[0074] Based on the soil environment dataset of the red plum and apricot trees, the initial water use efficiency and root distribution characteristics of each red plum and apricot tree at different growth stages were calculated to obtain a feature vector set.

[0075] The feature vector set was clustered using the k-means clustering algorithm, and through further iterative optimization, the classification results of the growth stages of the red plum apricot were obtained.

[0076] The system collected soil characteristic data from 100 red plum and apricot trees by deploying a soil sensor network and root detection equipment in the field. The soil type of each tree was accurately determined and quantified into water retention capacity parameters, with clay having a water retention capacity of 0.45, loam 0.30, and sandy soil 0.15. The initial water use efficiency of the root distribution was 0.60, 0.45, and 0.30, respectively.

[0077] Based on the collected dataset, the system constructs a two-dimensional feature vector for each tree. The first dimension represents the soil's water retention capacity, and the second dimension represents the initial water use efficiency. For example, a tree growing in clay soil might have a feature vector of [0.45, 0.60], while a tree in sandy soil might have a feature vector of [0.15, 0.30]. These feature vectors comprehensively describe the water use characteristics of each tree and serve as input data for the k-means clustering algorithm.

[0078] In the clustering analysis phase, the algorithm sets k=3 to correspond to the three main growth stages of the red plum apricot. Initially, three centroids are randomly selected, located at [0.40, 0.55], [0.25, 0.40], and [0.10, 0.25]. The Euclidean distance from each tree's feature vector to each centroid is calculated. For example, the distance from a tree at [0.45, 0.60] to the centroid at [0.40, 0.55] is... The trees are then assigned to the nearest cluster. After five iterations of optimization, the algorithm converges and yields three stable clusters: Cluster 1 contains 40 trees with centroids of [0.43, 0.58], corresponding to the seedling stage; Cluster 2 contains 35 trees with centroids of [0.28, 0.43], corresponding to the growth stage; and Cluster 3 contains 25 trees with centroids of [0.12, 0.28], corresponding to the maturity stage.

[0079] Cluster analysis revealed that trees in clay soils were mostly distributed in seedling clusters, benefiting from their high water retention capacity, which provided favorable conditions for early growth. Trees in sandy and loam soils, on the other hand, were mainly concentrated in the growing and mature stages, with relatively low water use efficiency limiting their early development. Each tree was ultimately assigned a corresponding growth stage label, which was then associated with its unique identifier and stored in a database. These classification results were directly applied to optimize irrigation system scheduling; for example, the irrigation frequency for seedling trees could be appropriately increased to 0.2 m³ per day, thus achieving precise water management.

[0080] In some embodiments, based on the above embodiments, the process of matching the initial humidity threshold corresponding to the soil type and local climate type with the growth stage classification results and a preset humidity threshold database to obtain the initial humidity control range includes the following steps:

[0081] Based on the growth stage classification results and the preset humidity threshold database, the preliminary humidity threshold for each tree at the current growth stage is obtained.

[0082] By matching the preliminary humidity threshold with the growth stage classification results, the initial humidity control range for each tree is obtained.

[0083] The system uses machine learning models (such as the random forest algorithm) to accurately classify the growth stage of each tree. This model generates classification labels by analyzing tree feature data (including leaf area, tree height, root depth, etc.), ensuring a classification accuracy of over 95%. For example, after a tree is classified as "growing stage", the system immediately extracts an initial humidity threshold from a preset humidity threshold database that precisely matches the soil type and climate conditions.

[0084] Taking a real-world application scenario as an example: When the system detects that the soil type of a tree is clay and the climate is temperate humid, it queries the database to return the initial humidity threshold of 25%-35% for the tree during its growth period. The system then dynamically adjusts this threshold based on real-time environmental data. By connecting to the weather station API, it obtains current rainfall and evaporation data (e.g., 5mm rainfall, 3mm evaporation). A weighted average algorithm is used for precise calculation. The weather station data is standardized to obtain a rainfall impact factor of 0.2 and an evaporation impact factor of 0.1. Then, these are substituted into the formula: Adjusted threshold = Initial threshold × (1 + Rainfall impact factor × 0.3 - Evaporation impact factor × 0.2) = 25% × (1 + 0.2 × 0.3 - 0.1 × 0.2) = 25.5%. The upper limit is calculated using the same method to obtain 35.5%, thus adjusting the humidity threshold range to 25.5%-35.5%.

[0085] To further improve the accuracy of regulation, the system also considers individual differences among trees. By measuring the root water absorption capacity of each tree using an infrared spectrometer, the system will fine-tune the humidity control range by ±2% for trees with a water absorption rate 10% higher than normal. For example, if the water absorption rate of a tree in its growth stage is 10% higher than normal, the system will further adjust its humidity control range from 25.5%-35.5% to 26%-36%.

[0086] The entire process is based on an intelligent humidity control method that integrates multi-source data. This ensures the scientific nature of the decision-making process while fully considering individual differences, providing the optimal growth environment management plan for each red plum and apricot tree.

[0087] In some embodiments, based on the above embodiments, comparing real-time rainfall or evaporation data with the expected climate value corresponding to the initial humidity control range, and when the deviation exceeds a preset deviation range, the process of analyzing climate conditions through a linear regression algorithm to obtain the adjusted humidity control range includes the following steps:

[0088] Compare the real-time rainfall and evaporation data with the expected climate values ​​corresponding to the initial humidity control range. If the deviation exceeds a preset deviation threshold, obtain the rainfall and evaporation data from the real-time monitoring data.

[0089] The impact of rainfall and evaporation data on environmental humidity was analyzed using a linear regression algorithm, and the data analysis results were obtained.

[0090] Based on the data analysis results, the initial humidity range is adjusted to generate the adjusted humidity control range.

[0091] The system collects climate data in real time through a deployed network of environmental sensors. Suppose the system detects a rainfall of 50 mm / day and an evaporation of 30 mm / day, calculating a net water input of 50 - 30 = 20 mm / day. Then, according to a preset humidity conversion model (each 1 mm / day of net water corresponds to a 1% change in humidity), it calculates that the current climate conditions will cause a 20% increase in soil moisture, rising from a baseline humidity of 70% to 90%. This calculated result exceeds the system's preset humidity deviation threshold (80% + 10% = 90%), triggering a linear regression analysis mechanism.

[0092] The system retrieves the latest 30 sets of climate data from the historical database, including rainfall sequences [40, 55, 60, 45, 50] mm / day, evaporation sequences [25, 30, 35, 28, 32] mm / day, and corresponding observed humidity values ​​[75%, 78%, 82%, 76%, 79%]. A prediction model is then established using a linear regression algorithm to calculate...

[0093]

[0094] in, It is the predicted ambient humidity value. It's rainfall data. It is the evaporation rate and the regression coefficient. =50.2, =0.5, =-0.3. Substituting the current measured data X1=50, X2=30 into the model, the predicted environmental humidity value Y=66.2% is calculated.

[0095] Although the predicted value was within the initial control range of 60%-80%, the system still initiated an optimization adjustment procedure because the previous climate-derived value had triggered a deviation alarm. Based on the scientific prediction of the regression model, the system adjusted the humidity control range from the original 60%-80% to 65%-75%, making it closer to the predicted value under current climate conditions. After the adjustment, the system immediately sent a command to the execution device through the IoT control module: when the humidity sensor reported that the current ambient humidity was 85%, the system automatically started the dehumidifier, operating at a power of 500W, reducing the humidity by 5% per hour, and expected to reduce the humidity to around the target range of 70% within 4 hours.

[0096] The entire control process, from data acquisition and analysis to equipment control, is automated by the central processing system. Rigorous data verification and logical judgment at each stage ensure the accuracy and reliability of the control process, ultimately achieving intelligent humidity management without human intervention. This control method, based on historical data modeling and real-time monitoring feedback, guarantees both the scientific nature of decision-making and the timeliness of execution, effectively maintaining the humidity stability of the red plum apricot's growing environment.

[0097] In some embodiments, based on the above embodiments, the process of obtaining the waterlogging risk level of the red plum apricot based on the adjusted humidity control range and root distribution depth, and generating a dynamic drainage plan based on the waterlogging risk level and in conjunction with the drainage strategy database, includes the following steps:

[0098] If the soil moisture exceeds the preset moisture threshold, a decision tree algorithm is used to preliminarily classify the water accumulation risk of each tree based on the root distribution depth to obtain an initial risk level.

[0099] Clustering algorithms are used to group tree locations to obtain location grouping results. If the location grouping results show that the trees are located in low-lying areas, the risk prediction weights are adjusted according to the initial risk level to obtain a weighted risk level.

[0100] Based on the weighted risk level and soil moisture data, a logistic regression algorithm is used to perform a secondary verification of the water accumulation risk to obtain the final risk level.

[0101] Query the drainage parameters corresponding to soil type and root distribution in the preset drainage strategy database, analyze their correlation with the final risk level, and obtain optimized drainage parameters;

[0102] Based on the optimized drainage parameters and the root distribution of each tree, the allocation weight of the dynamic drainage scheme is calculated to obtain the weight distribution result. If the weight distribution result meets the preset weight distribution threshold, a dynamic drainage scheme for each tree is generated.

[0103] Based on soil moisture data and root distribution characteristics, the system automatically assesses the waterlogging risk level of each tree. It acquires soil moisture content data at a depth of 0-30 cm using sensors (e.g., 0-10 cm: 45%, 10-20 cm: 38%, 20-30 cm: 32%), and combines this with root distribution depth measured by ground-penetrating radar (mainly distributed at 0-25 cm, with a maximum depth of 30 cm). Following preset rules (moisture content >40% indicates high risk), the system classifies the tree as high-risk.

[0104] The system retrieves corresponding drainage parameters from a pre-set drainage strategy database based on the risk level. Taking clay soil with a root depth of 0.5 meters as an example, the system finds a basic drainage rate of 0.2 cubic meters per hour and a pipe diameter of 0.3 meters. Trees in the area are grouped according to water depth, categorized into three levels: high-risk clusters (water depth > 0.7 meters), medium-risk clusters (0.4-0.7 meters), and low-risk clusters (< 0.4 meters), and assigned different drainage rates for each level: 0.25 cubic meters per hour, 0.15 cubic meters per hour, and 0.1 cubic meters per hour, respectively.

[0105] Finally, the system generates a personalized drainage plan for each tree, outputting it to the drainage control system in a standardized JSON format: {“tree_id”:1,“drainage_rate”:0.25, “pipe_diameter”:0.3}. The entire process, from risk assessment to drainage plan generation, is fully automated, ensuring that drainage measures are precisely matched to the actual conditions of each tree.

[0106] In some embodiments, based on the above embodiments, the process of dynamically adjusting the irrigation amount using a proportional-integral-differential algorithm according to the dynamic drainage scheme and the adjusted humidity control range to obtain precise irrigation parameters for each tree includes the following steps:

[0107] The soil moisture value is collected in real time by sensors and compared with the preset moisture threshold.

[0108] If the soil moisture value is lower than the preset moisture threshold, the irrigation adjustment coefficient is calculated by the proportional-integral-differential algorithm to obtain the irrigation adjustment value.

[0109] Based on the irrigation adjustment value and combined with the dynamic drainage scheme, the irrigation water allocation for each tree is calculated to obtain the precise irrigation parameters for each tree.

[0110] The system monitors soil moisture in real time using high-precision soil moisture sensors (model SHT35, measurement range 0-100%RH, accuracy ±1.5%RH). One sensor is buried 5 cm deep at the base of each tree, collecting data every 10 minutes and transmitting the data to a cloud database via a LoRa wireless module.

[0111] Taking one tree as an example, the sensor detects that the current soil moisture is 25%RH, which is lower than the preset threshold of 35%RH, triggering an irrigation request. The cloud system uses a proportional-integral-derivative (PID) control algorithm to dynamically calculate the irrigation amount. The PID algorithm formula is:

[0112]

[0113] in, For the total control variable, the humidity deviation value e(t) = target humidity - current humidity = 35%RH - 25%RH = 10%RH, Kp is the proportional coefficient = 0.5, Ki is the integral coefficient = 0.1, Kd is the derivative coefficient = 0.2, ∫e(t)dt represents the cumulative amount of deviation over time, and de(t) / dt represents the rate of change of deviation.

[0114] The calculated total control quantity u(t) = 5.167, which translates to an irrigation volume of 5.167 L / h. The system irrigates using a precision drip irrigation device (Netafim dripper, adjustable flow rate 0-10 L / h) for 30 minutes, with an actual irrigation volume of 2.583 L.

[0115] After irrigation, the sensor detected that the humidity rose to 33%RH. The system recalculated e(t) = 2%RH and recalculated the PID control quantity: the total control quantity u(t) = 1.04. The irrigation quantity was adjusted to 1.04L / h and continued for 10 minutes. An additional 0.173L of irrigation was added to stabilize the humidity within the target range of 35%RH ± 1%RH.

[0116] Meanwhile, the system calculates the drainage time based on the soil permeability model (K=0.02cm / s) and uses Darcy's formula t=irrigation water volume / (soil permeability coefficient×assumed drainage area×hydraulic slope) to estimate the drainage time. The calculated drainage time is about 2 hours. Based on this, the system suspends irrigation for 2 hours to avoid water accumulation.

[0117] All operational data is stored in the cloud. Combined with historical irrigation records (average daily irrigation volume of 5L / tree), the system continuously optimizes humidity threshold adjustments using a random forest machine learning model (prediction accuracy of 85%), forming a complete intelligent irrigation closed-loop control system. This system achieves precise water and fertilizer management by dynamically adjusting irrigation volume based on real-time soil moisture while also considering drainage needs.

[0118] like Figure 2 As shown in the figure, this embodiment of the invention provides a soil moisture threshold control system for the efficient cultivation of red plum apricot. The system can dynamically optimize irrigation and drainage strategies, achieve precise water management, improve the growth efficiency of red plum apricot, and reduce the risk of waterlogging.

[0119] The data acquisition module M201 is used to acquire soil type, root distribution depth and climate conditions, and combined with the preset red plum apricot growth stage database, to obtain soil water retention capacity and initial water use efficiency.

[0120] The growth classification module M202 is used to obtain the growth stage classification results of the red plum apricot based on the soil water retention capacity and initial water use efficiency using the k-means clustering algorithm.

[0121] The threshold matching module M203 is used to match the initial humidity threshold corresponding to the soil type and local climate type based on the growth stage classification results and the preset humidity threshold database, so as to obtain the initial humidity control range.

[0122] The dynamic adjustment module M204 is used to compare real-time rainfall or evaporation data with the expected climate value corresponding to the initial humidity control range. When the deviation exceeds the preset deviation range, the climate conditions are analyzed through a linear regression algorithm to obtain the adjusted humidity control range.

[0123] The risk assessment module M205 is used to obtain the water accumulation risk level of the red plum apricot based on the adjusted humidity control range and root distribution depth, and to generate a dynamic drainage plan based on the water accumulation risk level and the drainage strategy database.

[0124] The precision control module M206 is used to dynamically adjust the irrigation amount according to the dynamic drainage scheme and the adjusted humidity control range, using a differential algorithm to obtain precise irrigation parameters for each tree.

[0125] The data acquisition module M201 collects real-time data on soil type, root depth, and climate conditions (such as rainfall and evaporation) through a network of sensors deployed in the field (e.g., the TDR-315L soil moisture sensor and the CI-600 root scanner) and by accessing a weather station API. After acquiring the data, it uses a pre-defined database of red plum apricot growth stages for joint analysis. For example, it analyzes the particle size distribution data collected by the sensors using the USDA soil classification standard to determine the soil type (e.g., sandy loam) and calculate its water retention capacity (e.g., field capacity of 20%). Simultaneously, it combines a root distribution model and the water balance equation to calculate the initial water use efficiency (e.g., 40% utilization rate under conditions of a seedling water requirement of 0.8 mm / day and a net water input of 2 mm / day). The module's outputs—water retention capacity and initial water use efficiency—serve as the quantitative basis for all subsequent analyses and decisions.

[0126] The growth classification module M202 receives the processing results from M201 and intelligently identifies the growth stage of each tree based on its soil water retention capacity and initial water use efficiency. These two feature data points for each tree are constructed into a two-dimensional feature vector (e.g., [0.45, 0.60]). Subsequently, the k-means clustering algorithm is used, with k=3 (corresponding to seedling stage, growth stage, and maturity stage). The cluster centers are iteratively optimized by calculating Euclidean distance, ultimately dividing all trees into different growth stage clusters (e.g., cluster 1 centroid [0.43, 0.60] is identified as "seedling stage"). The output of this module provides a crucial basis for subsequent differentiated management and precise control.

[0127] The threshold matching module M203 first accesses a preset humidity threshold database based on the growth stage classification results (e.g., "growing stage") provided by M202, combined with the soil type and climate type provided by M201. This database stores optimal humidity parameters under different conditions, validated by historical data (e.g., the initial humidity threshold for clay, temperate climate, and growing stage is 25%-35%). Next, the module incorporates real-time environmental data (e.g., current rainfall of 5mm and evaporation of 3mm), and dynamically corrects the initial threshold using a weighted average algorithm (adjusted threshold = initial threshold × (1 + rainfall impact factor × 0.3 - evaporation impact factor × 0.2)). Finally, it generates a preliminary humidity control range (e.g., 25.5%-35.5%) for each tree.

[0128] The dynamic adjustment module M204 performs a secondary calibration on the initial range output by M203. This module continuously compares real-time rainfall or evaporation with the climate prediction values ​​used by module M203. When the deviation exceeds a preset range (e.g., ±10%), the calibration process is triggered. The module uses a linear regression algorithm to build a prediction model based on historical climate datasets (e.g., rainfall, evaporation, and actual humidity data from the past 30 days) and substitutes real-time data to calculate a predicted humidity value that better matches the current climate conditions. Finally, it outputs an adjusted humidity control range (e.g., from 60%-80% to 65%-75%), making it more responsive to real-world environmental changes.

[0129] The risk assessment module M205 employs a decision tree algorithm (such as CART) to establish classification rules based on soil moisture and root depth at each layer, and waterlogging risk levels (high, medium, low) as labels (e.g., "0-10 cm water content > 40%" is considered high risk). Based on the assessed risk level, the module further queries the drainage strategy database for matching drainage parameters (e.g., for clay soil, the drainage rate is 0.2 m³ / h) and uses K-means clustering to group trees within the area according to their waterlogging severity. Finally, it generates a dynamic drainage plan for each tree, specifying the required drainage amount and priority.

[0130] The M206 precision control module monitors soil moisture in real time using sensors and dynamically calculates the required irrigation amount using a proportional-integral-derivative (PID) algorithm to achieve precise water replenishment (e.g., calculating an irrigation rate of 6 L / h). Simultaneously, it strictly references drainage plans, using soil permeability models (such as Darcy's formula) to calculate drainage time and performs necessary drainage operations or suspends irrigation after irrigation to prevent waterlogging. This module ultimately outputs precise irrigation parameters for each tree (such as irrigation volume, irrigation duration, and interval), driving the irrigation and drainage equipment to form a complete closed loop.

[0131] The soil moisture threshold control system for high-efficiency cultivation of red plum apricot provided in this embodiment determines the soil water retention characteristics and water use efficiency through data fusion, accurately classifies the growth stages based on machine learning, dynamically optimizes the moisture threshold by combining environmental factors, and uses decision tree to assess waterlogging risk and generate drainage schemes. Finally, the PID algorithm realizes the coordinated control of irrigation amount and drainage strategy, thereby dynamically optimizing irrigation and drainage strategies, achieving precise water management, improving the growth efficiency of red plum apricot and reducing the risk of waterlogging.

[0132] The above description, in conjunction with specific preferred embodiments, provides a further detailed explanation of the present invention. It should not be construed that the specific implementation of the present invention is limited to these descriptions. For those skilled in the art, various simple deductions or substitutions can be made without departing from the concept of the present invention, and all such modifications and substitutions should be considered within the scope of protection of the present invention.

Claims

1. A method for controlling soil moisture threshold for high-efficiency cultivation of red plum apricot, characterized in that, include: By acquiring soil type, root distribution depth and climate conditions, and combining them with a pre-set database of red plum apricot growth stages, soil water retention capacity and initial water use efficiency can be obtained. Based on the soil water retention capacity and initial water use efficiency, the k-means clustering algorithm was used to obtain the classification results of the growth stages of the red plum apricot. Based on the growth stage classification results and the preset humidity threshold database, the initial humidity threshold corresponding to the soil type and local climate type is matched to obtain the initial humidity control range. Compare the real-time rainfall or evaporation with the expected climate values ​​corresponding to the initial humidity control range. When the deviation exceeds the preset deviation range, analyze the climate conditions through a linear regression algorithm to obtain the adjusted humidity control range. Based on the adjusted humidity control range and root distribution depth, the water accumulation risk level of the red plum apricot is obtained. Based on the water accumulation risk level, a dynamic drainage plan is generated in conjunction with the drainage strategy database. Based on the dynamic drainage scheme and the adjusted humidity control range, the irrigation amount is dynamically adjusted using a proportional-integral-differential algorithm to obtain precise irrigation parameters for each tree.

2. The method as described in claim 1, characterized in that, The classification results of the growth stages of the red plum apricot are obtained by using the k-means clustering algorithm based on the soil water retention capacity and initial water use efficiency, including: Soil moisture content and root depth information are collected by sensors, and combined with the soil water retention capacity and initial water use efficiency, a soil environment dataset for red plum and apricot trees is obtained. Based on the soil environment dataset of the red plum and apricot trees, the initial water use efficiency and root distribution characteristics of each red plum and apricot tree at different growth stages were calculated to obtain a feature vector set. The feature vector set was clustered using the k-means clustering algorithm, and through further iterative optimization, the classification results of the growth stages of the red plum apricot were obtained.

3. The method as described in claim 1, characterized in that, The process involves acquiring soil type, root distribution depth, and climatic conditions, combined with a pre-set database of red plum apricot growth stages, to obtain soil water retention capacity and initial water use efficiency, including: Acquire data on soil type, root distribution depth, and climate conditions collected by sensors; Based on a pre-set database of red plum apricot growth stages, soil characteristic parameters matching soil type and growth stage are queried to obtain soil water retention capacity. Based on the soil water retention capacity and root distribution depth, the initial water use efficiency is obtained using the random forest algorithm.

4. The method as described in claim 1, characterized in that, The method involves matching initial humidity thresholds corresponding to the soil type and local climate type based on the growth stage classification results and a preset humidity threshold database to obtain the initial humidity control range, including: Based on the growth stage classification results and the preset humidity threshold database, the preliminary humidity threshold for each tree at the current growth stage is obtained. By matching the initial humidity threshold with the soil type and local climate type, the initial humidity control range for each tree is obtained.

5. The method as described in claim 1, characterized in that, The process of comparing real-time rainfall or evaporation with the expected climate values ​​corresponding to the initial humidity control range, and when the deviation exceeds a preset deviation range, using a linear regression algorithm to analyze climate conditions and obtain an adjusted humidity control range, includes: Compare the real-time rainfall and evaporation with the expected climate values ​​corresponding to the initial humidity control range. If the deviation exceeds a preset deviation threshold, obtain the rainfall and evaporation from the real-time monitoring data. The effects of rainfall and evaporation on environmental humidity were analyzed using a linear regression algorithm, and the data analysis results were obtained. Based on the data analysis results, the initial humidity range is adjusted to generate the adjusted humidity control range.

6. The method as described in claim 1, characterized in that, Based on the adjusted humidity control range and root distribution depth, the waterlogging risk level of the red plum apricot is obtained. Based on the waterlogging risk level and in conjunction with the drainage strategy database, a dynamic drainage plan is generated, including: If the soil moisture exceeds the preset moisture risk threshold, a decision tree algorithm is used to preliminarily classify the water accumulation risk of each tree based on the root distribution depth to obtain an initial risk level. Clustering algorithms are used to group tree locations to obtain location grouping results. If the location grouping results show that the trees are located in low-lying areas, the risk prediction weights are adjusted according to the initial risk level to obtain a weighted risk level. Based on the weighted risk level and soil moisture data, a logistic regression algorithm is used to perform a secondary verification of the water accumulation risk to obtain the final risk level. Query the drainage parameters corresponding to soil type and root distribution in the preset drainage strategy database, analyze their correlation with the final risk level, and obtain optimized drainage parameters; Based on the optimized drainage parameters and the root distribution of each tree, the allocation weight of the dynamic drainage scheme is calculated to obtain the weight distribution result. If the weight distribution result meets the preset weight distribution threshold, a dynamic drainage scheme for each tree is generated.

7. The method as described in claim 1, characterized in that, The process of dynamically adjusting the irrigation amount using a proportional-integral-differential algorithm based on the dynamic drainage scheme and the adjusted humidity control range to obtain precise irrigation parameters for each tree includes: Soil moisture is collected in real time by sensors and compared with preset moisture irrigation thresholds; If the soil moisture is lower than the preset moisture irrigation threshold, the irrigation volume adjustment coefficient is calculated by the proportional-integral-differential algorithm to obtain the irrigation volume adjustment value. Based on the irrigation adjustment value and combined with the dynamic drainage scheme, the irrigation water allocation for each tree is calculated to obtain the precise irrigation parameters for each tree.

8. A soil moisture threshold control system for high-efficiency cultivation of red plum apricot, characterized in that, The system includes: The data acquisition module is used to obtain soil type, root distribution depth and climate conditions, and combined with the preset red plum apricot growth stage database, to obtain soil water retention capacity and initial water use efficiency. The growth classification module is used to obtain the growth stage classification results of the red plum apricot based on the soil water retention capacity and initial water use efficiency using the k-means clustering algorithm. The threshold matching module is used to match the initial humidity threshold corresponding to the soil type and local climate type based on the growth stage classification results and the preset humidity threshold database, so as to obtain the initial humidity control range. The dynamic adjustment module is used to compare real-time rainfall or evaporation data with the expected climate value corresponding to the initial humidity control range. When the deviation exceeds the preset deviation range, the climate conditions are analyzed through a linear regression algorithm to obtain the adjusted humidity control range. The risk assessment module is used to obtain the water accumulation risk level of the red plum apricot based on the adjusted humidity control range and root distribution depth, and to generate a dynamic drainage plan based on the water accumulation risk level and the drainage strategy database. The precision control module is used to dynamically adjust the irrigation amount based on the dynamic drainage scheme and the adjusted humidity control range, using a differential algorithm to obtain precise irrigation parameters for each tree.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the program implements the steps of the method according to any one of claims 1 to 7.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps of the method according to any one of claims 1 to 7.

Images