Intelligent management system and method for water resources in agricultural irrigation area

Abstract

The invention discloses an agricultural irrigation area water resource intelligent management system and method, and relates to the field of intelligent management, and the method comprises the steps: building a correlation model of a crop growth stage and a basic irrigation threshold value through real-time collection of original weather station data, future 24-hour weather forecast, crop types, planting dates and other multi-dimensional data; and in combination with soil humidity sensor data and a channel pump station state, the basic irrigation starting threshold is dynamically adjusted by using an FAO formula or a deep learning algorithm to form a dynamic irrigation decision threshold. And then multi-source data is input into an irrigation decision module, an irrigation starting and stopping instruction is generated by judging the relation between the irrigation state and the soil humidity threshold value, and closed-loop management is achieved. Thus, the limitation of traditional irrigation static threshold management is broken through, and through meteorological data dynamic adaptation and multi-source data fusion decision making, on the premise of guaranteeing crop growth water demand, the water resource allocation efficiency is optimized, and the intelligent management level of an agricultural irrigation area is improved.

Description

Technical Field

[0001] The present application relates to the field of intelligent management, and more specifically, to an intelligent management system and method for water resources in agricultural irrigation areas. Background Art

[0002] The increasingly complex and volatile modern agricultural production environment places higher demands on intelligent and efficient water resource management in agricultural irrigation areas. Rational allocation of limited water resources and precision irrigation are crucial not only for crop growth, development, and yield quality, but also for the sustainable use of water resources. Traditional agricultural irrigation methods rely heavily on empirical experience and fixed irrigation plans, lacking dynamic responsiveness to soil moisture conditions, crop growth stages, and meteorological changes. This leads to water waste or insufficient irrigation, making it difficult to meet the demands of modern agriculture for efficient and scientific water resource management.

[0003] However, existing agricultural irrigation water resource management systems primarily rely on soil moisture sensors to collect real-time data and use fixed threshold parameters to control irrigation starts and stops. However, these methods, often based on static thresholds, struggle to adapt to complex and changing meteorological conditions and crop growth requirements. For example, failure to consider future changes in rainfall and evapotranspiration in weather forecasts can lead to delays and errors in irrigation decisions, impacting the rational allocation and efficient use of water resources. Furthermore, existing technologies generally ignore the differences in water requirements at different crop growth stages, making it difficult to achieve precise and intelligent adjustments to irrigation strategies.

[0004] Therefore, there is an urgent need for an optimized intelligent management system and method for water resources in agricultural irrigation areas. Summary of the Invention

[0005] In order to solve the above technical problems, this application is proposed.

[0006] According to one aspect of the present application, a method for intelligent management of water resources in agricultural irrigation areas is provided, which includes: obtaining original meteorological station data, weather forecast data for the next 24 hours, and current crop types and planting dates; based on the current crop types and planting dates, determining the growth stage of the current crop to obtain the current growth stage of the crop; based on the current growth stage of the crop, querying the current stage basic irrigation start threshold and the current stage basic irrigation stop threshold from a predefined basic soil moisture upper and lower limit threshold table; based on the original meteorological station data and the weather forecast data for the next 24 hours, dynamically adjusting the current stage basic irrigation start threshold to obtain the current stage dynamic irrigation start threshold; obtaining average soil moisture data and channel and pump station status data; and inputting the average soil moisture data, channel and pump station status data, the current stage dynamic irrigation start threshold and the current stage basic irrigation stop threshold into an irrigation decision module to obtain an irrigation decision management result.

[0007] According to another aspect of the present application, an intelligent management system for water resources in an agricultural irrigation area is provided, which includes: a data acquisition module for acquiring original meteorological station data, weather forecast data for the next 24 hours, and current crop types and planting dates; a crop growth stage judgment module for judging the growth stage of the current crop based on the current crop type and planting date to obtain the current growth stage of the crop; an irrigation threshold query module for querying the current stage basic irrigation start threshold and the current stage basic irrigation stop threshold from a predefined basic soil moisture upper and lower limit threshold table based on the current growth stage of the crop; an irrigation start threshold dynamic adjustment module for dynamically adjusting the current stage basic irrigation start threshold based on the original meteorological station data and the weather forecast data for the next 24 hours to obtain the current stage dynamic irrigation start threshold; an equipment and soil status data acquisition module for acquiring average soil moisture data and channel and pump station status data; an irrigation decision module for inputting the average soil moisture data, channel and pump station status data, the current stage dynamic irrigation start threshold and the current stage basic irrigation stop threshold into the irrigation decision module to obtain an irrigation decision management result.

[0008] Compared with the existing technology, the present application provides an intelligent management system and method for water resources in agricultural irrigation areas. It collects multi-dimensional data such as original meteorological station data, weather forecasts for the next 24 hours, crop types and planting dates in real time to build a correlation model between crop growth stages and basic irrigation thresholds. It also combines soil moisture sensor data and channel pump station status, and uses FAO formulas or deep learning algorithms to dynamically adjust the basic irrigation start threshold to form a dynamic irrigation decision threshold. Multi-source data is then input into the irrigation decision module, and by judging the relationship between the irrigation status and the soil moisture threshold, irrigation start and stop instructions are generated to achieve closed-loop management. In this way, the limitations of traditional static irrigation threshold management are broken through. Through dynamic adaptation of meteorological data and fusion decision-making of multi-source data, the efficiency of water resource allocation is optimized while ensuring the water demand for crop growth, and the level of intelligent management of agricultural irrigation areas is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] The above and other purposes, features, and advantages of the present application will become more apparent through a more detailed description of the embodiments of the present application in conjunction with the accompanying drawings. The accompanying drawings are intended to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the drawings, the same reference numerals generally represent the same components or steps.

[0010] Figure 1 This is a flow chart of an intelligent management method for water resources in agricultural irrigation areas according to an embodiment of the present application.

[0011] Figure 2 This is a data flow diagram of the intelligent management method of water resources in agricultural irrigation areas according to an embodiment of the present application.

[0012] Figure 3 This is a flowchart of sub-step S4 of the method for intelligent management of water resources in agricultural irrigation areas according to an embodiment of the present application.

[0013] Figure 4 This is a flowchart of sub-step S4 of the method for intelligent management of water resources in agricultural irrigation areas according to another embodiment of the present application.

[0014] Figure 5 This is a flowchart of sub-step S5 of the method for intelligent management of water resources in agricultural irrigation areas according to an embodiment of the present application.

[0015] Figure 6 This is a flowchart of sub-step S6 of the method for intelligent management of water resources in agricultural irrigation areas according to an embodiment of the present application.

[0016] Figure 7 This is a block diagram of an intelligent management system for water resources in agricultural irrigation areas according to an embodiment of the present application. DETAILED DESCRIPTION

[0017] The following describes embodiments of the present disclosure in more detail with reference to the accompanying drawings. While the drawings illustrate certain embodiments of the present disclosure, it should be understood that the present disclosure can be implemented in various forms and should not be construed as limited to the embodiments described herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are for illustrative purposes only and are not intended to limit the scope of protection of the present disclosure.

[0018] In response to the problems in the above-mentioned background technology, this application proposes an intelligent management method for water resources in agricultural irrigation areas. Figure 1 This is a flow chart of an intelligent management method for water resources in agricultural irrigation areas according to an embodiment of the present application. Figure 2 This is a data flow diagram of the method for intelligent management of water resources in agricultural irrigation areas according to an embodiment of the present application. Figure 1 and Figure 2As shown, the intelligent management method of water resources in agricultural irrigation areas includes the following steps: S1, obtaining original meteorological station data, weather forecast data for the next 24 hours, and current crop types and planting dates; S2, judging the growth stage of the current crop based on the current crop type and planting date to obtain the current growth stage of the crop; S3, querying the current stage basic irrigation start threshold and the current stage basic irrigation stop threshold from a predefined basic soil moisture upper and lower limit threshold table based on the current growth stage of the crop; S4, dynamically adjusting the current stage basic irrigation start threshold based on the original meteorological station data and the weather forecast data for the next 24 hours to obtain the current stage dynamic irrigation start threshold; S5, obtaining average soil moisture data and channel and pump station status data; S6, inputting the average soil moisture data, channel and pump station status data, the current stage dynamic irrigation start threshold and the current stage basic irrigation stop threshold into an irrigation decision module to obtain an irrigation decision management result.

[0019] In the above-mentioned intelligent water resource management method for agricultural irrigation areas, step S1 obtains raw weather station data, weather forecast data for the next 24 hours, and the current crop type and planting date. The raw weather station data includes net radiation, daily average temperature, daily average wind speed at a height of 2 meters above the ground, and actual air vapor pressure. Specifically, the net radiation is directly measured by a radiation sensor, the daily average temperature is directly measured by a temperature sensor, the daily average wind speed at a height of 2 meters above the ground is directly measured by a wind speed sensor, and the actual air vapor pressure is calculated from the relative humidity and air temperature. It should be understood that in order to provide accurate and dynamic basic information for water resource management in agricultural irrigation areas, thereby achieving precision irrigation, this application collects real-time raw weather station data and combines it with weather forecast data for the next 24 hours, particularly rainfall forecasts. This system can reasonably determine the future replenishment of soil moisture and avoid over-irrigation or under-irrigation due to uncertain weather conditions. In addition, by obtaining the current crop type and planting date and clarifying the crop's growth stage, irrigation thresholds and strategies can be accurately adjusted according to the water demand characteristics of crops at different growth stages, thereby providing scientific and comprehensive basic data support for irrigation decisions.

[0020] In particular, in one possible embodiment, step S1 is implemented as follows: First, raw weather station data is collected in real time from weather stations deployed in agricultural irrigation areas. This data primarily covers key meteorological parameters such as net radiation, daily average temperature, daily average wind speed at a height of 2 meters above the ground, and actual air vapor pressure. Net radiation is directly measured by a radiation sensor, daily average temperature is measured by a temperature sensor, wind speed is monitored by a wind speed sensor, and actual air vapor pressure is calculated based on relative humidity and temperature. The weather station equipment continuously and stably collects and transmits this data, ensuring real-time and accuracy, providing a foundation for subsequent calculations of indicators such as crop evapotranspiration.

[0021] Secondly, 24-hour weather forecast data is typically obtained through data interfaces with meteorological authorities or third-party weather service platforms. This automatically pulls forecast information for the next 24 hours, focusing on rainfall forecasts and other relevant meteorological parameters. This forecast data is formatted and verified to ensure compatibility with intelligent management systems for agricultural irrigation areas, improving its availability and reliability.

[0022] In addition, current crop type and planting date information is entered by farmers or management systems, typically through agronomic management platforms or agricultural IoT terminals. Based on the crop type and planting date, the system automatically determines the crop's current growth stage, serving as a key basis for adjusting irrigation strategies. Crop information and meteorological data are simultaneously uploaded and stored in a database, providing multi-dimensional data support for dynamic decision-making.

[0023] The entire data collection process is centrally managed by the data acquisition module, ensuring real-time, synchronized updates of raw weather station data and weather forecast data. It also integrates crop information to effectively integrate multi-source data. Using network communication technologies such as wireless transmission and IoT protocols, the collected raw data is transmitted to a central processing unit or cloud platform, ensuring data integrity and timeliness. This provides a reliable data foundation for dynamic threshold adjustment and irrigation decision-making in the intelligent irrigation system.

[0024] In the above-mentioned intelligent management method for water resources in agricultural irrigation areas, the step S2, based on the current crop type and planting date, determines the growth stage of the current crop to obtain the current growth stage of the crop. It should be understood that in order to accurately identify the specific development stage of the crop in its growth cycle, thereby providing a scientific basis for precision irrigation and water resource management, this application calculates the current stage of the crop by understanding the type and planting time of the crop, such as the germination period, vegetative growth period, flowering period or maturity period, and there are significant differences in the demand for water at different stages. After executing this judgment, the water requirement characteristics and irrigation thresholds of the crop at the current stage can be clarified, thereby supporting targeted adjustment of the irrigation strategy, realizing the stage-by-stage differentiation of irrigation management, ensuring that irrigation measures are more in line with crop growth needs, avoiding water resource waste or crop growth damage caused by improper water management, and improving the water resource utilization efficiency and crop yield quality of agricultural irrigation areas.

[0025] In particular, in one possible embodiment, the implementation process of step S2 is as follows: First, the system obtains the current crop type information and accurate planting date from the agronomy management platform or the agricultural Internet of Things terminal. Then, based on the pre-established crop growth cycle database or model, the database records in detail the different growth stages of various crops from sowing to harvesting and their corresponding time intervals. Next, the system matches the actual growth days with the time intervals in the crop growth cycle database to determine the specific growth stage the crop is currently in. For example, for corn crops, the system will determine whether it is in the germination stage, seedling stage, jointing stage, filling stage, maturity stage, etc. based on the growth days. In addition, taking into account regional climate and environmental factors, the system may combine local meteorological data to make growth corrections to improve the accuracy of the judgment. The judgment result is automatically updated and stored in the cloud platform database for subsequent irrigation threshold configuration and dynamic adjustment module calls.

[0026] In the above-mentioned intelligent management method for water resources in agricultural irrigation areas, step S3 queries the basic irrigation start threshold and the basic irrigation stop threshold of the current stage from the predefined basic soil moisture upper and lower limit threshold table based on the current growth stage of the crop. Specifically, the present application realizes precise control of irrigation timing and irrigation intensity by utilizing predefined basic soil moisture upper and lower limit thresholds for different growth stages. Crops at different growth stages have different requirements for soil moisture. Reasonable setting of start threshold and stop threshold can ensure that irrigation not only meets the water needs of crops, but also avoids water waste or soil salinization caused by excessive irrigation. By accurately obtaining the basic irrigation start and stop thresholds corresponding to the current growth stage from the predefined threshold table, a standardized and scientific basis is provided for irrigation control, ensuring the systematic and targeted nature of irrigation management. In this way, the appropriate soil moisture range can be dynamically identified according to the different growth stages of the crops, guiding the irrigation system to generate reasonable start and stop commands, and realizing refined management of irrigation operations. This not only ensures that crops receive appropriate water supply during the critical growth period, promotes healthy growth and increases yield, but also improves water resource utilization efficiency, reduces unnecessary water consumption and resource waste, and thus promotes the sustainable use and intelligent management of water resources in agricultural irrigation areas.

[0027] In particular, in one possible embodiment, step S3 is implemented as follows: First, combining crop growth characteristics and agronomic research, a threshold table covering various growth stages, such as the germination stage and the maturity stage, is developed for different crop types, such as corn and wheat. The soil moisture values corresponding to the irrigation start and stop stages are clearly defined and stored in the system database. In specific implementation, the current growth stage is first determined based on the crop type and planting date. For example, corn is determined to be in the grain filling stage. Based on this, the threshold table for the corresponding crop is then located and the growth stage is searched. For example, the grain filling stage of corn is searched in the predefined basic soil moisture upper and lower limit threshold table, as shown in Table 1. The basic irrigation start threshold is found to be 60% (volume moisture content) and the stop threshold is found to be 80% (volume moisture content). Finally, the query results are returned for subsequent dynamic adjustments and irrigation decision-making.

[0028] Table 1 Predefined basic soil moisture upper and lower threshold values (excerpt):

[0029] In the above-mentioned intelligent management method of water resources in agricultural irrigation areas, the step S4 dynamically adjusts the basic irrigation start threshold of the current stage based on the original meteorological station data and the weather forecast data for the next 24 hours to obtain the dynamic irrigation start threshold of the current stage. Specifically, due to relying solely on static thresholds, it is difficult to adapt to the dynamic changes in meteorological conditions and crop growth stages, which can easily lead to waste of water resources or insufficient irrigation. Therefore, the present application quantifies the impact of the environment on soil moisture by integrating real-time meteorological data with short-term rainfall forecasts, so that the irrigation threshold dynamically adapts to the actual water demand of crops, and obtains the dynamic irrigation start threshold of the current stage, thereby optimizing the efficiency of water resource allocation and avoiding irrigation lags or errors caused by static thresholds, thereby realizing closed-loop management of multi-source data fusion, while ensuring the water demand for crop growth, and improving the efficiency of water resource utilization.

[0030] In particular, in one possible embodiment, Figure 3 FIG4 is a flow chart of sub-step S4 of the method for intelligent management of water resources in agricultural irrigation areas according to an embodiment of the present application. Figure 3 As shown, the step S4 includes: S41, calculating the reference crop evapotranspiration based on the original weather station data; S42, extracting the future predicted rainfall from the weather forecast data for the next 24 hours; S43, calculating the dynamic adjustment amount based on the reference crop evapotranspiration and the future predicted rainfall; S44, adding the dynamic adjustment amount to the basic irrigation start threshold of the previous stage to obtain the dynamic irrigation start threshold of the current stage.

[0031] Specifically, the step S41 calculates the reference crop evapotranspiration based on the original weather station data. It should be understood that since the reference crop evapotranspiration is a key indicator for measuring the total amount of crop water evaporation and transpiration, it directly reflects the theoretical water requirement of the crop under specific meteorological conditions. Therefore, the present application can scientifically calculate the potential water loss of the crop on the same day or within a certain period of time by utilizing the real-time and accurate environmental data such as net radiation, air temperature, wind speed and air humidity provided by the original weather station. The obtained reference crop evapotranspiration provides a scientific basis for irrigation management, quantifies the water demand of crops, avoids blind irrigation or insufficient water, and can reasonably take into account the actual evaporation and transpiration of crops in the subsequent dynamic adjustment of irrigation thresholds, ensuring that irrigation decisions are in line with the physiological water requirements of crops and improving the efficiency of water resource utilization.

[0032] In particular, in one possible embodiment, the step S41 of calculating the reference crop evapotranspiration based on the original weather station data includes: inputting the original weather station data into the FAO Penman-Monteith formula to obtain the reference crop evapotranspiration, the formula being: ;in, is the slope of the saturated vapor pressure curve related to air temperature, is the net radiation, is the soil heat flux density, is the wet / dry constant, is the daily average temperature, is the daily average wind speed at a height of 2 meters above the ground, is the saturated water vapor pressure in the air, is the actual water vapor pressure of air, is the reference crop evapotranspiration.

[0033] Specifically, step S42 extracts the future predicted rainfall from the weather forecast data for the next 24 hours. Specifically, because traditional irrigation management does not consider the impact of future weather changes on soil moisture, it can easily lead to delayed or inaccurate irrigation decisions. Therefore, this application extracts the future predicted rainfall to predict the replenishment of soil moisture by natural precipitation in the future period, and uses it as a key parameter for dynamically adjusting the irrigation start threshold. This parameter complements the reference crop evapotranspiration, allowing irrigation decisions to proactively adapt to meteorological changes, avoiding over-irrigation due to unforeseen rainfall or under-irrigation due to ignoring the possibility of precipitation, thereby achieving dynamic optimization of irrigation thresholds based on weather forecasts.

[0034] In particular, in one possible embodiment, step S42 is implemented as follows: First, an automated data connection channel is established with the official data interface of a meteorological service agency or a third-party meteorological platform to periodically receive a data packet containing a refined hourly or time-segmented weather forecast for the next 24 hours. This data packet is typically encapsulated in a standardized JSON or XML format and contains key fields such as geographic coordinates, timestamp, meteorological element code, and corresponding numerical value. After receiving the raw forecast data, the system first performs format parsing and standardization. Using pre-set meteorological element mapping rules, the code fields are converted into internally unified parameter names, for example, mapping "rainfall_amount" to "forecastPrecipitation."

[0035] The system then filters forecast data for the corresponding areas based on the geographic location of the agricultural irrigation districts and verifies the accuracy of the data's timeframe, ensuring that the extracted rainfall data strictly corresponds to the next 24-hour window. To ensure data reliability, the system incorporates a multi-dimensional validation mechanism: it uses time series continuity testing to identify data gaps or jumps, utilizes historical meteorological statistical characteristic values for rationality verification, and flags or corrects outliers that violate physical laws.

[0036] For missing short-term data segments, the system uses a spatiotemporal interpolation algorithm to fill in the gaps. Specifically, this algorithm horizontally utilizes a weighted average of contemporaneous data from neighboring meteorological stations, and vertically integrates historical rainfall patterns for the same period to perform time series compensation. After data cleaning and correction, the system extracts features specific to agricultural irrigation needs. For hourly rainfall forecasts, the system calculates the cumulative total rainfall over the next 24 hours. If the forecast data is divided into time periods, such as three-hour intervals, the rainfall for each period is accumulated and summed.

[0037] The system also extracts rainfall intensity distribution characteristics, such as the maximum hourly rainfall and its occurrence time, to inform subsequent dynamic irrigation strategy adjustments. Ultimately, the processed predicted rainfall data is stored in a standardized numerical format in a designated data table within the system database and pushed to the irrigation decision-making module in real time via a message queue, providing real-time meteorological information for dynamic adjustments to irrigation thresholds.

[0038] Specifically, step S43 calculates a dynamic adjustment value based on the reference crop evapotranspiration and the future predicted rainfall. It should be understood that, because static thresholds in traditional irrigation management cannot balance the dynamic balance between crop evapotranspiration water demand and future precipitation replenishment, calculating a dynamic adjustment value based on the reference crop evapotranspiration and future predicted rainfall quantifies the combined impact of meteorological factors on soil moisture, converting actual crop evapotranspiration loss and expected natural precipitation into calculable adjustment parameters. This allows for dynamic correction of the basic irrigation initiation threshold, addressing the irrigation decision-making lag caused by the lack of multi-source data fusion in traditional methods and enabling real-time adaptation of irrigation thresholds to meteorological changes.

[0039] In particular, in a possible embodiment, the step S43 of calculating the dynamic adjustment amount based on the reference crop evapotranspiration and the future predicted rainfall includes: calculating the dynamic adjustment amount based on the reference crop evapotranspiration and the future predicted rainfall using the following formula: ;in, is the reference crop evapotranspiration, Predicting future rainfall, is the evapotranspiration conversion function, is the rainfall conversion function, and are the evapotranspiration weight and rainfall weight, It is a dynamic adjustment amount.

[0040] It should be understood that the reference crop evapotranspiration and future rainfall forecasts The units are typically physical units, such as millimeters (mm), but they ultimately affect a soil moisture threshold in percentage units. Therefore, it's illogical to directly add or subtract millimeter values from a percentage threshold. A conversion function maps these input values, which have clear physical meaning, into an adjustment factor representing the intensity of the impact, allowing it to be multiplied by the weight and included in the calculation of the final adjustment.

[0041] It should be understood that the reference crop evapotranspiration The larger the value, the faster the soil will lose water in the future, and the more likely it is that irrigation will be started earlier. Therefore, the adjustment factor output by the evapotranspiration conversion function should also be larger. In one embodiment, the evapotranspiration conversion function uses a piecewise linear function. Specifically, the evapotranspiration conversion function is expressed as: ;in, represents the low impact threshold, represents a high impact threshold, represents the reference crop evapotranspiration, represents the linear region response coefficient, Represents the acceleration zone response coefficient.

[0042] Specifically, several key evapotranspiration nodes are preset, such as low impact threshold (e.g. 2 mm / day) and high impact threshold (e.g. 6 mm / day). Below the low impact threshold, its impact on soil moisture is negligible. The output can be set to 0 to prevent the system from overreacting to minor weather fluctuations. Between these two thresholds, The output can be The linear response coefficient indicates the size of the adjustment factor generated when the evapotranspiration increases by 1 unit within the normal evapotranspiration range. It represents the basic response sensitivity. For example, it can be set to 0.8. When the "high impact threshold" is exceeded, it indicates that the weather is extremely hot and dry, and the risk of water loss has increased dramatically. At this time, the growth slope of the function can be designed to be larger to produce a stronger early warning signal, prompting a significant increase in the irrigation threshold. That is, the acceleration zone response coefficient should be greater than the linear zone response coefficient. The acceleration zone response coefficient represents the adjustment factor generated by each unit increase in evapotranspiration under high evapotranspiration intensity. For example, it can be set to 1.5. Here, the low impact threshold, high impact threshold, linear zone response coefficient, and acceleration zone response coefficient are only examples and are not specifically limited. The specific values can be adjusted according to local climate characteristics and crop types.

[0043] Correspondingly, the rainfall conversion function needs to consider the agricultural concept of effective rainfall. Not all predicted rainfall can effectively replenish soil moisture for crop roots. Therefore, the rainfall conversion function can also use a piecewise linear function. Specifically, the rainfall conversion function is expressed as: ;in, represents the future predicted rainfall, Indicates the invalid rainfall threshold, represents the soil saturation infiltration threshold, represents the conversion coefficient of the effective absorption area, Indicates the maximum effective water replenishment factor.

[0044] Specifically, an invalid rainfall threshold (such as 1 mm) can be defined in the above formula. When the value is lower than this, these trace precipitations are likely to evaporate before reaching the ground or on the soil surface, and have no substantial contribution to deep soil moisture. The output should be 0. When the invalid threshold is exceeded and the absorption range is ideal (such as 1-30 mm), The output of The increase in the effective absorption area conversion coefficient is used to indicate the size of the water replenishment benefit adjustment factor generated by each unit increase in rainfall within the effective rainfall range. For example, it can be set to 1.0. However, when When the soil saturation infiltration threshold (such as 30mm) is exceeded, it means that heavy rain may occur, and the excess water may not be fully absorbed by the soil and will form surface runoff. The growth rate should slow significantly, even approaching a saturation ceiling, the maximum effective water replenishment factor, which can be set to 29.0, for example. This ensures that the irrigation initiation threshold is not excessively lowered due to a very high rainfall forecast, thus avoiding the misjudgment that the crop root zone may still be water-deficient after a heavy rain. The ineffective rainfall threshold, soil saturation infiltration threshold, effective absorption zone conversion coefficient, and maximum effective water replenishment factor are only examples and are not specifically defined. Specific values can be adjusted according to local climate characteristics and crop types.

[0045] At the same time, evapotranspiration weight and rainfall weight are used to quantify the relative importance of reference crop evapotranspiration and future predicted rainfall to the dynamic adjustment amount. They can be statically configured based on the prior knowledge of agricultural experts and regional climate characteristics. For example, in a region with perennial drought and little rain and strong evaporation, water evaporation is the dominant factor affecting soil moisture, so will be given a much higher The value of Set to 0.75, Set to 0.1. This makes the dynamic adjustment naturally more sensitive to drought risk. On the contrary, in a monsoon climate with abundant rainfall and obvious seasonality, accurate prediction and utilization of rainfall is the key to water-saving irrigation, so The weight of Same or higher.

[0046] Here, the reference crop evapotranspiration and the future predicted rainfall are converted by the conversion function respectively, and then weighted by the evapotranspiration weight and the rainfall weight. If the dynamic adjustment amount is used as the schedulable interval, then it is obviously composed of the schedulable positive value area, that is, , and the dispatchable negative region, i.e. Thus, within the short time sequence, the schedulable positive value region and the schedulable negative value region will inevitably have a short time sequence significant asymmetric characteristic, thereby causing an asymmetric short time significant error in the dynamic adjustment amount.

[0047] Therefore, in order to improve the robustness of the dynamic adjustment amount relative to short-term timing errors during real-time calculation, a weighted short-term normative correlation compensation mechanism is introduced, that is, the normative balance correlation within the short-term time series is used as a correlation compensation for significant asymmetric oscillations to ensure the short-term coupling correlation between the positive and negative regions, thereby improving the robustness of the short-term timing errors of the dynamic adjustment amount.

[0048] That is, in another possible embodiment of this embodiment, the dynamic adjustment amount is calculated based on the reference crop evapotranspiration and the future predicted rainfall, and it also includes: determining the evapotranspiration weight and the rainfall weight, and dynamically correcting the evapotranspiration weight and the rainfall weight to obtain the corrected evapotranspiration weight and the corrected rainfall weight.

[0049] Specifically, the evapotranspiration weight and the rainfall weight are dynamically corrected to obtain the corrected evapotranspiration weight and the corrected rainfall weight, including: first, determining the function conversion value of the reference crop evapotranspiration and the future predicted rainfall within a predetermined time to obtain the reference crop evapotranspiration conversion vector and the future predicted rainfall conversion vector, that is, obtaining Function distribution in short time series dimension ,as well as Corresponding function distribution It should be understood that reference crop evapotranspiration (e.g., actual crop evaporation loss) and future predicted rainfall (e.g., natural precipitation replenishment) are typically expressed in physical units. Their impact on soil moisture is not a simple linear relationship and may exhibit complex dynamic changes over short time series. To enable the dynamic adjustment value to be used as a schedulable interval and to address the significant short-term asymmetry that may occur between its positive and negative regions, these physical quantities are converted into standardized functional distributions with time series characteristics. In one specific embodiment, the reference crop evapotranspiration and future predicted rainfall obtained for each of the past eight hours are substituted into the evapotranspiration conversion function and the rainfall conversion function to obtain the reference crop evapotranspiration conversion vector and the future predicted rainfall conversion vector. In this embodiment, the reference crop evapotranspiration conversion vector and the future predicted rainfall conversion vector each have eight elements. They clearly depict the dynamic impact of evapotranspiration and rainfall on the water balance of agricultural fields within a specific time window, laying a solid foundation for subsequent refined analysis and ensuring the validity and consistency of the data before entering complex models.

[0050] Then, the canonical correlation matrix of the reference crop evapotranspiration conversion vector and the future predicted rainfall conversion vector is calculated, which is expressed as: ;in, represents the reference crop evapotranspiration conversion vector, represents the future predicted rainfall conversion vector, represents the transpose of a vector, represents matrix multiplication, Indicates the calculation of L2 norm, represents the canonical correlation matrix.

[0051] It should be understood that these two key meteorological factors, evapotranspiration and rainfall, do not exist in isolation; their interaction has a holistic impact on soil moisture conditions. Understanding their overall, linear relationship within a short time series is necessary to establish a fundamental, macroscopic correlation model, representing a canonical equilibrium correlation between evapotranspiration and rainfall. By calculating the canonical correlation matrix, we can quantify the overall statistical dependence of evapotranspiration and rainfall within a predetermined time series, revealing their synergistic or antagonistic trends under typical meteorological conditions. For example, under persistently clear and dry weather, evapotranspiration and rainfall may be negatively correlated, while in rainy seasons, they may be weakly correlated. This provides an important reference benchmark for subsequent, more complex asymmetric analyses and a baseline understanding of the average interaction between the two for irrigation systems.

[0052] And, the gauge correlation of its significant asymmetric oscillation can be expressed as follows: ;in, The first vector representing the reference crop evapotranspiration conversion vector Reference crop evapotranspiration conversion value, represents the first vector of the future predicted rainfall transformation The future predicted rainfall conversion value, Represents the significantly asymmetric oscillatory canonical correlation matrix The value of the position.

[0053] Relying solely on conventional canonical correlations is insufficient to capture the significant asymmetric oscillations that can occur in evapotranspiration and rainfall over short time series. For example, a sudden, heavy rainfall event can have a short-term impact on soil moisture that far exceeds its average impact, and the intensity and duration of this impact may be asymmetric with the linear effect of evapotranspiration. Alternatively, under extreme heat and drought conditions, evapotranspiration can increase dramatically. If this asymmetry is not accounted for, it can lead to asymmetric, short-term, significant errors in the dynamic adjustment, thus affecting the timeliness and accuracy of irrigation decisions. By calculating a canonical correlation matrix of significant asymmetric oscillations, we can further explore the nonlinear and unbalanced dynamic connections between evapotranspiration and rainfall over short time series, especially during rapid changes or extreme events. This allows us to identify local oscillations that may lead to significant deviations in the dynamic adjustment, enabling irrigation systems to respond more sensitively to sudden meteorological changes and achieving a significant representation of time series paths. By calculating the significantly asymmetric oscillation canonical correlation matrix, it can accurately capture those instantaneous and asymmetric interactions between evapotranspiration and rainfall that do not conform to the conventional linear relationship. For example, when short-term heavy rainfall is predicted, the matrix will highlight the rapid replenishment effect of rainfall on soil moisture, providing key abnormal signals for subsequent weight corrections, ensuring that incorrect irrigation decisions are not made due to inertia.

[0054] Thus, since the canonical correlation matrix and the significantly asymmetric oscillatory canonical correlation matrix The canonical correlation matrix and the significantly asymmetric oscillation canonical correlation matrix have the same dynamic correlation sensitivity, and the canonical correlation matrix and the significantly asymmetric oscillation canonical correlation matrix are subjected to F-norm canonical correlation correlation modulation to obtain a canonical correlation correlation modulation matrix: ;in, represents the canonical correlation matrix, represents the significantly asymmetric oscillatory canonical correlation matrix, represents the transpose of a vector, represents matrix multiplication, represents element-wise multiplication, represents the canonical correlation modulation matrix.

[0055] To achieve more robust irrigation decisions, it is necessary to consider both the overall linear correlation between evapotranspiration and rainfall (representing typical meteorological patterns) and their significant asymmetric oscillations within short time series (representing extreme or unexpected meteorological events). Simply superimposing or averaging these two types of information cannot effectively integrate their unique contributions. Through the correlation modulation of the F-norm-like canonical correlation, these two correlations can be cleverly integrated, allowing the canonical equilibrium correlation to serve as a correlation compensation for the significant asymmetric oscillations, thereby forming a more adaptable integrated correlation model. Here, the canonical correlation correlation modulation matrix not only reflects the conventional correlation between evapotranspiration and rainfall but, more importantly, allows for dynamic and weighted adjustments to this correlation by asymmetric oscillations within the short time series, ensuring that irrigation decisions can simultaneously account for long-term trends and short-term fluctuations and that short-term coupled correlations between positive and negative regions are maintained. The resulting canonical correlation correlation modulation matrix acts as a dynamic regulator, flexibly adjusting the weights of their influence on irrigation decisions based on the actual interaction patterns of evapotranspiration and rainfall within the short time series.

[0056] The short-term distribution of the function is further mapped, and the weight is corrected based on the norm. That is, based on the normative correlation modulation matrix and the reference crop evapotranspiration conversion vector, the evapotranspiration weight is corrected to obtain a corrected evapotranspiration weight, and based on the normative correlation modulation matrix and the future predicted rainfall conversion vector, the rainfall weight is corrected to obtain a corrected rainfall weight, which is expressed as: ;in, and are the evapotranspiration weight and rainfall weight, Fan correlation modulation matrix, represents the reference crop evapotranspiration conversion vector, represents the future predicted rainfall conversion vector, represents matrix multiplication, Indicates the calculation of L2 norm, represents the first preset parameter, represents the second preset parameter, for example, Of course, this is just an example and can be adjusted according to the irrigation area environment, crop type and historical data. represents the weight of the corrected evapotranspiration, Represents the corrected rainfall weight.

[0057] In other words, the original evapotranspiration and rainfall weights are static and cannot reflect the dynamic impact of meteorological conditions on crop water demand and soil moisture replenishment in real time. In actual agricultural production, the relative importance of evapotranspiration and rainfall on soil water balance is constantly changing. For example, in the dry season, evapotranspiration has a much greater impact on soil moisture than rainfall, while the opposite is true in the rainy season. By introducing a dynamic correction mechanism, these weights can be made adaptive, more accurately reflecting the actual contribution of evapotranspiration and rainfall to irrigation decisions under current meteorological conditions and optimizing the weighting coefficients. Specifically, based on the short-term significant asymmetric balance between schedulable positive and schedulable negative regions, the canonical correlation compensation for significant asymmetric oscillations under time series path significance is performed by accurately capturing the canonical significant correlation in the short-term dimension. This optimizes the weighting coefficients to ensure the short-term coupled correlation between positive and negative regions, thereby improving the robustness of the short-term time series errors of the dynamic adjustment variable. Ultimately, these dynamically corrected weights can more accurately guide the calculation of the dynamic adjustment variable, ensuring that the adjustment of the irrigation initiation threshold is more aligned with actual meteorological conditions and crop water demand. For example, when persistent high temperatures and high wind speeds lead to high evapotranspiration and no rainfall is predicted, the evapotranspiration weight may be significantly increased, prompting the irrigation initiation threshold to be advanced to prevent crop damage due to water shortages. Conversely, when effective rainfall is predicted, the rainfall weight may be increased, thereby lowering the irrigation initiation threshold, avoiding unnecessary irrigation and conserving water resources. This adaptive weight adjustment mechanism significantly enhances the intelligence level and decision-making accuracy of water resource management systems in agricultural irrigation areas, maximizes water resource utilization efficiency, and ensures healthy crop growth.

[0058] Specifically, step S44 adds the dynamic adjustment amount to the previous stage's basic irrigation initiation threshold to obtain the current stage's dynamic irrigation initiation threshold. Specifically, by linearly superimposing the previous stage's basic irrigation initiation threshold and the dynamic adjustment amount, a dynamic irrigation initiation threshold for the current stage is generated that combines the water requirements of the crop growth stage with real-time meteorological adaptability. This dynamic adjustment can be dynamically adjusted based on the combined influence of current meteorological factors. For example, when heavy rainfall is predicted within the next 24 hours, the dynamic adjustment amount is negative, and after superposition, the initiation threshold is lowered to reduce unnecessary irrigation. When high temperatures and strong evapotranspiration occur, the adjustment amount is positive, raising the initiation threshold to replenish water in advance.

[0059] In particular, in another possible embodiment, Figure 4 FIG4 is a flowchart of sub-step S4 of the method for intelligent management of water resources in agricultural irrigation areas according to another embodiment of the present application. Figure 4As shown, the step S4 includes: S41, structuring the original weather station data and the weather forecast data for the next 24 hours and arranging them into an external condition coding vector; S42, performing external condition element association coding based on a fully connected layer on the external condition coding vector to obtain an implicit association coding vector between external condition elements; S43, performing decoding regression on the implicit association coding vector between external condition elements to obtain a dynamic adjustment amount; S44, adding the dynamic adjustment amount to the basic irrigation start threshold of the previous stage to obtain the dynamic irrigation start threshold of the current stage.

[0060] Specifically, in step S41, the original weather station data and the weather forecast data for the next 24 hours are structured and arranged into an external condition coding vector. It should be understood that since the original weather station data and the weather forecast data for the next 24 hours are multi-source heterogeneous data, their data formats, units and dimensions are different, and they cannot be effectively identified and processed when directly input into the model, making it difficult to mine the implicit associations between the data. Therefore, they need to be structured to form a coding vector in a unified format. Specifically, by structuring the multi-source meteorological data and arranging them into an external condition coding vector. After structuring and coding, the original weather station data and the weather forecast data for the next 24 hours are converted into a standardized vector form, and the dimensional consistency and feature correlation of the data are enhanced, which facilitates the subsequent mining of the nonlinear relationship between meteorological elements.

[0061] In particular, in a possible embodiment, the implementation process of step S41 is as follows: First, multi-dimensional data cleaning is performed on the parameters such as net radiation, daily average temperature, daily average wind speed at a height of 2 meters above the ground, actual water vapor pressure of the air in the original meteorological station data, and the predicted rainfall in the weather forecast data for the next 24 hours. Outliers are filtered out by setting thresholds, and missing data are supplemented by time series interpolation. At the same time, data that obviously violates regional meteorological laws is eliminated in combination with the climate characteristic library of the area where the agricultural irrigation area is located (for example, extreme abnormal data of continuous 30°C without precipitation during the rainy season in the south). Secondly, data standardization is performed, and Z-Score standardization is used for continuous data, where the Z-Score standardization formula is: ,in is the mean, is the standard deviation, For continuous data, it is mapped to a distribution with a mean of 0 and a variance of 1; for non-negative data such as rainfall, Min-Max normalization is used, where the Min-Max normalization formula is ,in, is non-negative data, and Representing the maximum and minimum values in non-negative data, ensuring that all data lies in the interval [0, 1] and eliminating dimensionality differences. Next, a structured data model is constructed, defining key features such as net radiation, daily average temperature, wind speed, actual air vapor pressure, and predicted rainfall as the base dimensions of the vector. Derived dimensions are supplemented based on the correlation between meteorological elements. For data with time series characteristics, they are expanded into continuous dimensions in chronological order, and short-term trend features are extracted using a sliding window technique. Finally, the dimensions are arranged in the logical order of original meteorological parameters, derived meteorological indicators, predicted rainfall, and time series features to form a fixed-length external condition encoding vector. For example, the vector format is defined as [net radiation, daily average temperature, wind speed, actual air vapor pressure, rainfall in the next hour, rainfall in the next two hours, ..., rainfall in the next 24 hours, accumulated rainfall in the six hours, accumulated rainfall in the twelve hours] and stored in floating-point tensor format, ensuring direct input into the fully connected layer for subsequent correlation encoding. This provides standardized data input for the deep learning model to extract implicit correlations between meteorological elements.

[0062] Specifically, the step S42 performs external condition element association coding based on the fully connected layer on the external condition coding vector to obtain an implicit association coding vector between external condition elements. It should be understood that due to the complex nonlinear associations between the meteorological elements in the external condition coding vector, it is impossible to directly reflect the interaction mechanism between the elements. Therefore, it is necessary to map the weight matrix and activation function of the fully connected layer to explore the implicit association rules between the meteorological elements. Specifically, the neural network structure of the fully connected layer is used to perform feature transformation and abstraction on the implicit association coding vector between the external condition elements, and map the external condition coding vector to an implicit association coding vector between the external condition elements that can represent the element association in a high-dimensional space. Through layer-by-layer calculations of multiple layers of fully connected layers, deep features are extracted to provide a more physically meaningful association feature representation for subsequent decoding regression, thereby realizing intelligent calculation of dynamic adjustment quantities.

[0063] In particular, in one possible embodiment, the implementation process of step S42 is as follows: First, a multi-layer fully connected neural network structure is constructed, with the number of neurons in the input layer consistent with the dimension of the external condition encoding vector. For example, if it contains 10 meteorological features, the input layer is 10-dimensional. The hidden layer is set to 2-3 layers, such as 50-30-20 neurons. The output layer dimension is set according to the requirements of the implicit correlation features, such as generating a 10-dimensional hidden vector, and the ReLU activation function is used to enhance the nonlinear expression capability. Secondly, historical meteorological data and corresponding dynamic adjustment labels are used, and the actual adjustment value calculated by the FAO formula is used as the supervision signal for model training. The weight matrix and bias term of the fully connected layer are optimized using the backpropagation algorithm. L2 regularization and dropout are used, such as a retention rate of 0.8, to prevent overfitting. Then, the standardized external condition encoding vector is input into the trained fully connected network. After matrix operations and activation function mapping are performed in each hidden layer, the combined features between meteorological elements are gradually extracted. For example, the first layer extracts the linear combination of temperature and wind speed, and the second layer extracts the nonlinear combination of temperature, wind speed, and humidity. Finally, the implicit correlation coding vector between external condition elements is obtained from the output of the last hidden layer. This vector contains the correlation feature representation of meteorological elements in high-dimensional space. For example, each dimension represents a combination pattern of meteorological elements, which can be directly used in subsequent decoding regression to calculate the dynamic adjustment amount.

[0064] Specifically, step S43 performs decoding regression on the implicit correlation coding vector between the external condition elements to obtain a dynamic adjustment value. Specifically, because the implicit correlation coding vector between the external condition elements obtained through fully connected layer encoding is an abstract feature representation in a high-dimensional space and cannot be directly used for irrigation threshold adjustment, it must be mapped into a specific numerical dynamic adjustment value through decoding regression. Accordingly, through the decoding regression model, the meteorological factor coupling relationship represented in the implicit correlation coding vector between the external condition elements is converted into a quantifiable dynamic adjustment value. The dynamic adjustment value generated by decoding regression can accurately reflect the degree of impact of the synergistic effect of multiple meteorological factors on crop water demand. For example, when the implicit correlation coding vector between the external condition elements represents the characteristics of "strong radiation + high temperature + low wind speed", the positive adjustment value obtained by regression can scientifically reflect the increase in irrigation demand caused by increased evapotranspiration. When the characteristic of "high predicted rainfall" is included, the negative adjustment value can reasonably suppress unnecessary irrigation. Ultimately, the irrigation threshold is precisely adjusted in response to dynamic changes in meteorological conditions, reducing irrigation lag or over-irrigation.

[0065] In particular, in one possible embodiment, step S43 is implemented as follows: First, a decoding regression model architecture is constructed, using a single-layer or multi-layer fully connected network as the decoder. For example, the input layer dimension is consistent with the implicit correlation encoding vector between external condition factors, the output layer is a one-dimensional adjustment value, the activation function is a linear function or a Reluctant Unit (ReLU) function, and the loss function uses the mean squared error to measure the deviation between the predicted adjustment value and the actual adjustment value. Second, a training dataset is constructed using the implicit correlation encoding vector between external condition factors corresponding to historical meteorological data and the dynamic adjustment value calculated using the FAO formula. For example, 1000 sets of samples under different meteorological conditions over the past three years are collected. The model parameters are optimized using batch gradient descent, with a learning rate of 0.01 and a batch size of 32. Training is iterated for 200 rounds until the loss function converges. Then, the implicit correlation encoding vector between external condition factors obtained by encoding the fully connected layer is input into the trained decoding regression model. A forward propagation calculation is performed, such as multiplying the input vector by the weight matrix and adding the bias, to obtain a predicted dynamic adjustment value. This value can be positive or negative, where a positive adjustment value indicates an increase in the irrigation start threshold, while a negative adjustment value indicates a decrease in the threshold. Finally, the dynamic adjustment amount obtained by regression is physically verified. Combined with the water demand characteristics of the crop growth stage and the water resource conditions of the irrigation area, the upper and lower limits of the adjustment amount are set, such as the ±20% basic threshold, to prevent abnormal adjustments caused by extreme weather forecasts and ensure that the adjustment results meet the actual needs of agricultural irrigation.

[0066] Specifically, in step S44, the dynamic irrigation start threshold value of the previous stage is added to the dynamic adjustment value to obtain the dynamic irrigation start threshold value of the current stage. In particular, in one possible embodiment, the implementation process of step S44 is as follows: first, the basic irrigation start threshold value corresponding to the current crop growth stage is queried from a predefined basic soil moisture upper and lower limit threshold table. This threshold value has been pre-set according to the different growth stages of the crop varieties, for example, the basic threshold value of the corn seedling stage is 60% soil relative humidity; second, a dynamic adjustment value is obtained by calculating using the FAO formula or a deep learning algorithm. This adjustment value comprehensively considers factors such as reference crop evapotranspiration and predicted rainfall, for example, a +5% adjustment value is calculated; then, the basic threshold value and the adjustment value are checked for dimensional consistency, both of which are soil relative humidity percentages. If there are unit differences, they are standardized; finally, a linear superposition operation is performed to obtain the dynamic irrigation start threshold value of the current stage, for example, 60% + 5% = 65%, and the result is checked for rationality. If it does not exceed 80% of the soil field water holding capacity, if it exceeds, the threshold boundary value is used to ensure that the dynamic threshold value conforms to the actual agricultural irrigation conditions.

[0067] In the aforementioned intelligent water resource management method for agricultural irrigation areas, step S5 acquires average soil moisture data as well as channel and pump station status data. Specifically, by acquiring average soil moisture data, the current soil moisture content is quantified, providing a direct basis for determining whether irrigation should be started or stopped. Furthermore, acquiring channel and pump station status data clarifies the operating status of irrigation equipment, enabling irrigation decision-making to generate executable start and stop instructions based on the equipment's actual operating conditions, thereby achieving closed-loop irrigation management through multi-source data fusion. This ensures that the irrigation decision-making module generates accurate irrigation instructions based on actual soil moisture conditions and equipment operating status.

[0068] In particular, in one possible embodiment, Figure 5 FIG. 5 is a flow chart of sub-step S5 of the method for intelligent management of water resources in agricultural irrigation areas according to an embodiment of the present application. Figure 5 As shown, the step S5 includes: S51, obtaining original soil moisture sensor readings collected by soil moisture sensors deployed at multiple locations in the field; S52, calculating the average value of the original soil moisture sensor readings to obtain the average soil moisture data.

[0069] Specifically, step S51 obtains raw soil moisture sensor readings collected by soil moisture sensors deployed at multiple locations in the field. It should be understood that because data collected by a single soil moisture sensor cannot cover the differences in field microenvironments, it can easily lead to insufficient representation of soil moisture data. Based on this, the present application collects raw soil moisture sensor readings using sensors at multiple locations to obtain more comprehensive field soil moisture distribution information.

[0070] In particular, in one possible embodiment, the implementation process of step S51 is as follows: First, according to the differences in field crop planting layout and soil type, 3-5 soil moisture sensors, such as TDR or FDR principle sensors, are evenly distributed in the plot. The sensor burial depth needs to match the main root layer of the crop, such as 20-40 cm in the corn field, and avoid being close to special areas such as ridges and ditches. Secondly, the raw humidity readings collected by each sensor are transmitted to the edge computing gateway in real time through the LoRa or NB-IoT wireless communication module. The transmission frequency is set to once every 10 minutes to ensure data timeliness. Then, the raw soil moisture sensor readings are preliminarily preprocessed in the edge computing gateway, and obvious abnormal values are filtered out by setting reasonable thresholds, such as invalid data exceeding 100%, and persistent abnormal readings caused by sensor failures are marked. Finally, the preprocessed raw soil moisture sensor readings are synchronously stored in the local database and the cloud server according to the timestamp, providing a standardized raw data set for the subsequent calculation of the average soil moisture.

[0071] Specifically, step S52 calculates the average of the raw soil moisture sensor readings to obtain the average soil moisture data. Soil moisture at different locations in a field varies spatially, and a single sensor reading cannot represent the overall soil moisture conditions. Therefore, it is necessary to calculate the average of the raw soil moisture sensor readings at multiple locations to eliminate the randomness and limitations of single-point data and form a moisture index that is representative of the field.

[0072] In particular, in one possible embodiment, the implementation process of step S52 is as follows: First, the raw soil moisture sensor readings at the same timestamp of the soil moisture sensors deployed at multiple locations in the field are retrieved from the edge computing gateway or cloud database. The raw data must include the volumetric water content percentage and acquisition time of each point. Secondly, the raw soil moisture sensor readings are filtered out of abnormal values, using the 3σ principle, that is, values exceeding the mean ±3 times the standard deviation are considered abnormal, or a physically reasonable threshold is set, such as 0-100%, to filter out invalid data generated by sensor failure or environmental interference, such as readings of -5% or 110%. Then, the arithmetic average of all valid raw readings is calculated: ,in, is the average soil moisture, is the number of effective sensors, For the The system then calculates the effective readings of each sensor, retaining two decimal places to ensure accuracy. Finally, the calculation results are validated for rationality, such as by comparing them with field water capacity to ensure the average value does not exceed 80%. Once verified, the average soil moisture data is stored in a time series in a real-time database and synchronized with the irrigation decision module for subsequent threshold comparisons.

[0073] In the aforementioned intelligent water resource management method for agricultural irrigation areas, step S6 inputs the average soil moisture data, channel and pump station status data, the current dynamic irrigation start threshold, and the current basic irrigation stop threshold into the irrigation decision module to obtain an irrigation decision management result. Specifically, by combining multiple data sources, including average soil moisture data, channel and pump station status data, dynamic irrigation start threshold, and basic irrigation stop threshold, the limitations of static threshold management can be overcome through correlation analysis and dynamic adaptation between the data. For example, average soil moisture data is calculated from readings from multiple soil moisture sensors in the field and reflects soil moisture content in real time. This data does not directly adjust other parameters, but is compared with the dynamic irrigation start threshold and basic irrigation stop threshold to serve as the core basis for determining whether to start or stop irrigation. For example, irrigation is triggered when the average moisture level falls below the start threshold, and terminated when it exceeds the stop threshold. Channel and pump station status data, including channel flow, water level, and pump station start and stop and operating status, is used to confirm the actual operating conditions of the current irrigation equipment. Instead of directly adjusting parameters, it is used to determine whether the irrigation system is in an operating state, such as "irrigating" or "not irrigating," to determine whether to generate a start or stop instruction.

[0074] In particular, in one possible embodiment, Figure 6 FIG. 6 is a flow chart of sub-step S6 of the method for intelligent management of water resources in agricultural irrigation areas according to an embodiment of the present application. Figure 6 As shown, the step S6 includes: S61, judging the current irrigation status based on the channel and pump station status data; S62, if the current irrigation status is that irrigation is not currently in progress, judging whether the average soil moisture data is lower than the dynamic irrigation start threshold of the current stage, and if so, generating an irrigation start instruction; S63, if the current irrigation status is that irrigation is currently in progress, judging whether the average soil moisture data is higher than the basic irrigation stop threshold of the current stage, and if so, generating an irrigation stop instruction.

[0075] Specifically, step S61 determines the current irrigation status based on the channel and pump station status data. Specifically, a two-tiered judgment logic is first established. The first tier determines whether the pump station is powered on based on the pump station's start / stop status relay signal and a current threshold. In one embodiment, the current threshold is 10% of the rated current. The second tier determines whether the water flow is effective by combining the measured channel flow and a minimum irrigation flow threshold. In one embodiment, the minimum irrigation flow threshold is 0.3 m³ / s. Finally, the judgment result is written to the SCADA system's real-time database. A state change trigger mechanism is also implemented. For example, when the state transitions from "not irrigating" to "irrigating," the start time and peak flow rate are automatically recorded. A persistent abnormal state triggers an audible and visual alarm and issues a troubleshooting ticket. In one embodiment, if the current irrigation status is not irrigating, the average soil moisture data is determined to be below the dynamic irrigation start threshold for the current stage. If so, an irrigation start instruction is generated. Conversely, if the current irrigation status is irrigating, the average soil moisture data is determined to be above the basic irrigation stop threshold for the current stage. If so, an irrigation stop instruction is generated.

[0076] In summary, the intelligent management method of water resources in agricultural irrigation areas based on the embodiment of the present application is explained. It constructs a correlation model between the crop growth stage and the basic irrigation threshold by collecting multi-dimensional data such as original meteorological station data, weather forecast for the next 24 hours, crop types and planting dates in real time, and combines soil moisture sensor data and channel pump station status to dynamically adjust the basic irrigation start threshold using the FAO formula or deep learning algorithm to form a dynamic irrigation decision threshold. Multi-source data is then input into the irrigation decision module, and by judging the relationship between the irrigation status and the soil moisture threshold, an irrigation start-stop instruction is generated to achieve closed-loop management. In this way, the limitations of traditional static irrigation threshold management are broken through. Through dynamic adaptation of meteorological data and fusion decision-making of multi-source data, the water resource allocation efficiency is optimized while ensuring the water demand for crop growth, and the level of intelligent management of agricultural irrigation areas is improved.

[0077] Figure 7 FIG is a block diagram of an intelligent management system for water resources in agricultural irrigation areas according to an embodiment of the present application. Figure 7As shown, according to an embodiment of the present application, the intelligent management system 100 for water resources in an agricultural irrigation area includes: a data acquisition module 110 for acquiring original meteorological station data, weather forecast data for the next 24 hours, and the current crop type and planting date; a crop growth stage judgment module 120 for judging the growth stage of the current crop based on the current crop type and planting date to obtain the current growth stage of the crop; an irrigation threshold query module 130 for querying the current stage basic irrigation start threshold and the current stage basic irrigation stop threshold from a predefined basic soil moisture upper and lower limit threshold table based on the current growth stage of the crop; an irrigation start threshold dynamic adjustment module 140 for dynamically adjusting the current stage basic irrigation start threshold based on the original meteorological station data and the weather forecast data for the next 24 hours to obtain the current stage dynamic irrigation start threshold; an equipment and soil status data acquisition module 150 for acquiring average soil moisture data and channel and pump station status data; an irrigation decision module 160 for inputting the average soil moisture data, channel and pump station status data, the current stage dynamic irrigation start threshold and the current stage basic irrigation stop threshold into an irrigation decision module to obtain an irrigation decision management result.

[0078] As described above, the intelligent management system 100 for water resources in an agricultural irrigation area according to an embodiment of the present application can be implemented in various wireless terminals, such as a server having an equipment asset operation and maintenance management algorithm based on intelligent networking. In one possible implementation, the intelligent management system 100 for water resources in an agricultural irrigation area according to an embodiment of the present application can be integrated into a wireless terminal as a software module and / or a hardware module. For example, the intelligent management system 100 for water resources in an agricultural irrigation area can be a software module in the operating system of the wireless terminal, or can be an application developed for the wireless terminal; of course, the intelligent management system 100 for water resources in an agricultural irrigation area can also be one of the many hardware modules of the wireless terminal.

[0079] Alternatively, in another example, the agricultural irrigation area water resources intelligent management system 100 and the wireless terminal may also be separate devices, and the agricultural irrigation area water resources intelligent management system 100 may be connected to the wireless terminal via a wired and / or wireless network and transmit interactive information in accordance with an agreed data format.

[0080] Here, those skilled in the art will appreciate that the specific operations of each step in the above-mentioned intelligent management system for agricultural irrigation water resources have been described in detail above. Figures 1 to 6 The invention has been introduced in detail in the description of the intelligent management method of water resources in agricultural irrigation areas, and therefore, its repeated description will be omitted.

Claims

1. A method for intelligent management of water resources in agricultural irrigation areas, characterized in that: include: Obtaining original weather station data, weather forecast data for the next 24 hours, and the current crop type and planting date; based on the current crop type and planting date, determining the current crop growth stage to obtain the current crop growth stage; based on the current crop growth stage, querying a current stage basic irrigation start threshold and a current stage basic irrigation stop threshold from a predefined basic soil moisture upper and lower limit threshold table; based on the original weather station data and the weather forecast data for the next 24 hours, dynamically adjusting the current stage basic irrigation start threshold to obtain a current stage dynamic irrigation start threshold; Obtain average soil moisture data and channel and pump station status data; input the average soil moisture data, channel and pump station status data, the current stage dynamic irrigation start threshold and the current stage basic irrigation stop threshold into the irrigation decision module to obtain an irrigation decision management result.

2. The method for intelligent management of water resources in agricultural irrigation areas according to claim 1, characterized in that: Obtaining average soil moisture data includes: obtaining original soil moisture sensor readings collected by soil moisture sensors deployed at multiple locations in the field; and calculating an average value of the original soil moisture sensor readings to obtain the average soil moisture data.

3. The method for intelligent management of water resources in agricultural irrigation areas according to claim 1, characterized in that: Based on the original meteorological station data and the weather forecast data for the next 24 hours, the basic irrigation start threshold of the current stage is dynamically adjusted to obtain the dynamic irrigation start threshold of the current stage, including: calculating a reference crop evapotranspiration based on the original meteorological station data; extracting future predicted rainfall from the weather forecast data for the next 24 hours; calculating a dynamic adjustment amount based on the reference crop evapotranspiration and the future predicted rainfall; and adding the dynamic adjustment amount to the basic irrigation start threshold of the previous stage to obtain the dynamic irrigation start threshold of the current stage.

4. The method for intelligent management of water resources in agricultural irrigation areas according to claim 3, characterized in that: Calculating reference crop evapotranspiration based on the original weather station data includes: inputting the original weather station data into a FAO Penman-Monteith formula to obtain the reference crop evapotranspiration.

5. The method for intelligent management of water resources in agricultural irrigation areas according to claim 3, characterized in that: Calculating a dynamic adjustment amount based on the reference crop evapotranspiration and the future predicted rainfall includes: calculating the dynamic adjustment amount based on the reference crop evapotranspiration and the future predicted rainfall using the following formula, where the formula is: ;in, is the reference crop evapotranspiration, Predicting future rainfall, is the evapotranspiration conversion function, is the rainfall conversion function, and are the evapotranspiration weight and rainfall weight.

6. The method for intelligent management of water resources in agricultural irrigation areas according to claim 5, characterized in that: Calculating the dynamic adjustment amount based on the reference crop evapotranspiration and the future predicted rainfall also includes: determining the evapotranspiration weight and the rainfall weight, and dynamically correcting the evapotranspiration weight and the rainfall weight to obtain the corrected evapotranspiration weight and the corrected rainfall weight.

7. The method for intelligent management of water resources in agricultural irrigation areas according to claim 5, characterized in that: Dynamically correcting the evapotranspiration weight and the rainfall weight to obtain a corrected evapotranspiration weight and a corrected rainfall weight, including: determining a function conversion value of a reference crop evapotranspiration and a future predicted rainfall within a predetermined time to obtain a reference crop evapotranspiration conversion vector and a future predicted rainfall conversion vector; calculating a canonical correlation matrix of the reference crop evapotranspiration conversion vector and the future predicted rainfall conversion vector; calculating a significant asymmetric oscillation canonical correlation matrix of the reference crop evapotranspiration conversion vector and the future predicted rainfall conversion vector; performing an F-norm canonical correlation-modulated correlation on the canonical correlation matrix and the significant asymmetric oscillation canonical correlation matrix to obtain a canonical correlation-modulated matrix; correcting the evapotranspiration weight based on the canonical correlation-modulated matrix and the reference crop evapotranspiration conversion vector to obtain a corrected evapotranspiration weight; and correcting the rainfall weight based on the canonical correlation-modulated matrix and the future predicted rainfall conversion vector to obtain a corrected rainfall weight.

8. The method for intelligent management of water resources in agricultural irrigation areas according to claim 3, characterized in that: Based on the original weather station data and the weather forecast data for the next 24 hours, the basic irrigation start threshold of the current stage is dynamically adjusted to obtain the dynamic irrigation start threshold of the current stage, including: structuring the original weather station data and the weather forecast data for the next 24 hours and arranging them into an external condition coding vector; performing external condition element association coding based on a fully connected layer on the external condition coding vector to obtain an implicit association coding vector between external condition elements; performing decoding regression on the implicit association coding vector between external condition elements to obtain a dynamic adjustment amount; and adding the dynamic adjustment amount to the basic irrigation start threshold of the previous stage to obtain the dynamic irrigation start threshold of the current stage.

9. The method for intelligent management of water resources in agricultural irrigation areas according to claim 1, characterized in that: The average soil moisture data, the channel and pump station status data, the dynamic irrigation start threshold value of the current stage, and the basic irrigation stop threshold value of the current stage are input into the irrigation decision module to obtain an irrigation decision management result, including: judging the current irrigation status based on the channel and pump station status data; if the current irrigation status is that irrigation is not currently in progress, judging whether the average soil moisture data is lower than the dynamic irrigation start threshold value of the current stage, and if so, generating an irrigation start instruction; if the current irrigation status is that irrigation is currently in progress, judging whether the average soil moisture data is higher than the basic irrigation stop threshold value of the current stage, and if so, generating an irrigation stop instruction.

10. An intelligent management system for water resources in agricultural irrigation areas, characterized in that: include: Data acquisition module, used to obtain raw weather station data, weather forecast data for the next 24 hours, and current crop types and planting dates; a crop growth stage determination module for determining the current crop growth stage based on the current crop type and planting date to obtain the current crop growth stage; an irrigation threshold query module for querying the current stage basic irrigation start threshold and the current stage basic irrigation stop threshold from a predefined basic soil moisture upper and lower threshold table based on the current crop growth stage; an irrigation start threshold dynamic adjustment module, configured to dynamically adjust the basic irrigation start threshold for the current stage based on the original meteorological station data and the weather forecast data for the next 24 hours to obtain a dynamic irrigation start threshold for the current stage; and an equipment and soil status data acquisition module, configured to acquire average soil moisture data and channel and pump station status data. The irrigation decision module is used to input the average soil moisture data, channel and pump station status data, the current stage dynamic irrigation start threshold and the current stage basic irrigation stop threshold into the irrigation decision module to obtain an irrigation decision management result.

Images