Soil irrigation demand prediction method based on data analysis

Abstract

The invention discloses a soil irrigation demand prediction method based on data analysis, and relates to the technical field of agricultural irrigation, and the method comprises the following steps: S1, collecting the historical data of soil, and obtaining the irrigation prediction amount driven by a time sequence; s2, according to a plurality of characteristic parameters of the soil at the current moment, generating a multi-factor coupled irrigation prediction quantity; s3, fusion is carried out, and the irrigation prediction amount of the soil at the next moment is determined. According to the method, data of different dimensions are collected to form a complete original data set; two analysis methods are adopted for historical data, and irrigation prediction amount irrigation based on time sequence driving and multi-factor coupling is determined; the finally generated irrigation prediction amount can avoid the problems of hardening, nutrient loss and the like caused by over-wet soil, the crop yield and quality can be improved, and the development requirements of refined agriculture are met.

Description

Technical Field

[0001] This invention relates to the field of agricultural irrigation technology, and more specifically to a method for predicting soil irrigation demand based on data analysis. Background Technology

[0002] In agricultural irrigation forecasting, accurate irrigation volume prediction is the core technological support for achieving water-saving irrigation and increasing crop yields. Currently, existing irrigation forecasting technologies generally suffer from the following shortcomings: First, data collection dimensions are limited, with most schemes focusing only on a single type of data and lacking systematic integration of multi-source data, leading to forecasting models that are detached from actual planting scenarios; second, soil data lacks specificity, making it difficult to accurately reflect the actual water supply capacity of the soil; third, irrigation volume prediction often relies on a single driving logic (such as time-series prediction based solely on historical data), resulting in forecasts that either fail to capture historical irrigation patterns or cannot adapt to real-time environmental changes. These problems collectively make it difficult for existing technologies to balance irrigation efficiency and water-saving needs, failing to meet the practical application requirements of precision agricultural irrigation. Summary of the Invention

[0003] To address the above problems, this invention proposes a soil irrigation demand prediction method based on data analysis.

[0004] The technical solution of this invention is: a method for predicting soil irrigation demand based on data analysis, characterized by comprising the following steps:

[0005] S1. Collect historical soil data to obtain time-driven irrigation forecasts;

[0006] S2. Based on several characteristic parameters of the soil at the current moment, generate a multi-factor coupled irrigation forecast.

[0007] S3. The irrigation forecasts driven by time series and those coupled with multiple factors are fused to determine the irrigation forecast for the soil at the next time step.

[0008] Furthermore, in S1, historical soil data is collected and input into a neural network model to obtain time-driven irrigation predictions; the historical data includes irrigation amounts and precipitation amounts during historical periods.

[0009] In S2, historical soil data, in addition to the above, can also include stratified moisture content, temperature, and electrical conductivity over the past 72 hours; and can also include temperature, precipitation, sunshine, and humidity over the past 24 hours and the next 48 hours.

[0010] Furthermore, S2 includes the following sub-steps:

[0011] S21. Obtain future precipitation replenishment;

[0012] S22. Determine the predicted water requirement for crops;

[0013] S23. Construct a soil moisture characteristic curve for the soil at the current moment;

[0014] S24. Determine soil moisture loss based on the shape parameters of the soil moisture characteristic curve;

[0015] S25. The irrigation forecast is obtained by coupling multiple factors, including future precipitation replenishment, crop water demand, and soil moisture loss.

[0016] The beneficial effects of the above-mentioned further solutions are as follows: In this invention, irrigation demand is broken down into dimensions such as external replenishment (rainfall), crop demand, and soil condition, so that soil moisture originates from the actual water-holding capacity in different directions. The soil moisture characteristic curve uses soil water content Q (expressed as a volume percentage) as the horizontal axis and soil water suction S (expressed as atmospheric pressure) as the vertical axis.

[0017] Furthermore, S22 includes the following sub-steps:

[0018] S221. Weighted fusion of soil relative humidity and wind speed at the current moment yields the soil moisture availability correction coefficient.

[0019] S222. Multiply the soil moisture availability correction coefficient by the crop coefficient and reference evapotranspiration at the current time to obtain the predicted water requirement for the crop.

[0020] The beneficial effects of the above-mentioned further solutions are: In this invention, relative humidity (RH) and wind speed (u) are the core meteorological factors of crop transpiration stress, which directly determine the urgency of soil moisture on crops.

[0021] The lower the relative humidity (RH), the drier the air, the stronger the crop leaves' transpiration demand, and the faster the soil moisture is consumed.

[0022] The higher the wind speed, the faster the crop transpiration rate and the faster the evaporation of soil surface water. However, when the wind speed is too high, soil moisture will quickly be lost to a depth that the crop cannot absorb. At this time, the water availability needs to be appropriately reduced (to avoid ineffective irrigation).

[0023] Weighted fusion can provide ,in, , and These represent the first adaptation weight, the second adaptation weight, and the third adaptation weight, respectively. Dimensionless Can be taken , Can be taken , This indicates that an exponentiation operation is being performed.

[0024] Furthermore, in S24, soil moisture loss The expression is:

[0025] ;

[0026] ;

[0027] in, This represents the average density of all mineral particles in the soil. This indicates the soil's own volume density. This indicates the intake suction power value. The shape parameters representing the soil moisture characteristic curve. Indicates soil matrix potential, Represents the proportionality coefficient. This indicates the current soil moisture loss.

[0028] The beneficial effects of the above-mentioned further solutions are as follows: In this invention, the intake suction value is the critical suction parameter characterizing the entry of air into the pores during the dehydration process of saturated soil. Soil matrix potential is the potential energy generated by the adsorption force and capillary force of the soil matrix, reflecting the state of water free energy in unsaturated soil; its value is 0 in saturated soil and less than 0 in unsaturated soil.

[0029] When it is approximately equal to 1, the current matrix potential is close to the air intake value, which means that the large pores of the soil begin to be filled with water, and the soil moisture content rises rapidly at this time.

[0030] When the value is greater than 1, the current matrix potential is greater than the air intake value. The macropores are already filled with water, and the soil begins to fill the medium and small pores. The rate of increase in water content slows down, making it difficult for crop roots to absorb water.

[0031] When the value is less than 1, the current matrix potential is less than the air intake value, the macropores are not yet filled with water, and the soil is in the dry to slightly moist stage.

[0032] It reflects the density of the mineral particles (such as sand and clay) in the soil itself, and is unrelated to the soil's porosity and water content. Soil bulk density reflects the compactness of the soil; the lower the bulk density, the more pores and the looser the soil; conversely, the higher the bulk density, the more compact the soil (fewer pores).

[0033] Furthermore, in S25, the crop water requirement is added to the soil moisture loss, and then the future precipitation replenishment and the current soil effective water content are subtracted in turn. The result is then multiplied by 0 to obtain the irrigation forecast quantity coupled with multiple factors.

[0034] The beneficial effect of the above-mentioned further scheme is that, in this invention, future precipitation replenishment is obtained based on the precipitation amount predicted by short-term weather forecasts.

[0035] The current effective water content of the soil depends on the water absorption capacity of crop root hairs and the soil suction capacity.

[0036] Crop water requirement plus soil moisture loss equals total water consumption (i.e., water consumed by crop growth, soil evaporation, etc.); future rainfall replenishment plus current effective soil moisture content equals existing / future water reserves (i.e., usable water in the soil now plus future rainfall). A positive result indicates that existing reserves are insufficient to meet consumption, and this positive number represents the amount of water needed for additional irrigation. A negative result indicates that existing reserves exceed total consumption, meaning the soil already has enough water and no additional irrigation is needed.

[0037] Furthermore, in S3, the predicted irrigation amount for the soil at the next time step. The expression is:

[0038] ;

[0039] in, This represents time-series driven irrigation forecasts. This represents the irrigation forecast quantity resulting from the coupling of multiple factors. This indicates taking the minimum value.

[0040] The beneficial effects of the above-mentioned further scheme are as follows: In this invention, based on the deviation between the predicted values ​​of the two paths, the mean of path consistency and the minimum of irrigation conservatism are adaptively balanced. The more consistent the paths are, the more the mean is trusted; the greater the divergence between the paths, the more it is biased towards the conservative minimum. When the deviation between the predicted values ​​of the two paths is extremely small, that is... When the predicted values ​​approach zero, the time-series driven and multi-factor coupled predictions are consistent, indicating that the information from the two paths corroborates each other, and the mean is the most reasonable result. When the predicted values ​​from the two paths deviate significantly, the core constraint of the irrigation scenario is to avoid over-irrigation (over-irrigation leads to soil compaction and nutrient loss). The multi-factor coupled path is calculated based on the actual physical needs of the soil, weather, and crops, which is closer to the actual water shortage situation. Therefore, it biases towards the minimum value (usually the multi-factor coupled prediction) to ensure the rationality of the irrigation amount. When the predicted values ​​from the two paths deviate moderately, the result is a weighted combination of the mean and the minimum value, which retains the information from both paths without deviating excessively from the conservative requirements.

[0041] The beneficial effects of this invention are as follows: This invention collects data from different dimensions to form a complete original dataset; it employs two analysis methods on historical data to determine irrigation forecasts based on time-series driving and multi-factor coupling; this invention also adopts dual-path fusion prediction, which learns the time-series variation pattern of irrigation volume through historical data and quantifies physical needs such as crop water requirements and soil loss based on real-time multi-source data. When the deviation between the two paths is small, the average value is taken to improve accuracy; when the deviation is large, a conservative value is used to avoid over-irrigation. This solves the shortcomings of existing single-driving logic that cannot take into account both historical patterns and real-time changes, achieving a balance between the accuracy and safety of irrigation prediction; the final generated irrigation forecast can avoid problems such as soil compaction and nutrient loss caused by excessive moisture, which is conducive to improving crop yield and quality and meets the development needs of precision agriculture. Attached Figure Description

[0042] Figure 1 This is a flowchart of a data analysis-based method for predicting soil irrigation demand. Detailed Implementation

[0043] The embodiments of the present invention will be further described below with reference to the accompanying drawings.

[0044] like Figure 1 As shown, this invention provides a method for predicting soil irrigation demand based on data analysis, comprising the following steps:

[0045] S1. Collect historical soil data to obtain time-driven irrigation forecasts;

[0046] S2. Based on several characteristic parameters of the soil at the current moment, generate a multi-factor coupled irrigation forecast.

[0047] S3. The irrigation forecasts driven by time series and those coupled with multiple factors are fused to determine the irrigation forecast for the soil at the next time step.

[0048] Furthermore, in S1, historical soil data is collected and input into a neural network model to obtain time-driven irrigation predictions; the historical data includes irrigation amounts and precipitation amounts during historical periods.

[0049] In S2, historical soil data, in addition to the above, can also include stratified moisture content, temperature, and electrical conductivity over the past 72 hours; and can also include temperature, precipitation, sunshine, and humidity over the past 24 hours and the next 48 hours.

[0050] In this embodiment of the invention, S2 includes the following sub-steps:

[0051] S21. Obtain future precipitation replenishment;

[0052] S22. Determine the predicted water requirement for crops;

[0053] S23. Construct a soil moisture characteristic curve for the soil at the current moment;

[0054] S24. Determine soil moisture loss based on the shape parameters of the soil moisture characteristic curve;

[0055] S25. The irrigation forecast is obtained by coupling multiple factors, including future precipitation replenishment, crop water demand, and soil moisture loss.

[0056] In this invention, irrigation demand is broken down into dimensions such as external replenishment (rainfall), crop demand, and soil condition, so that soil moisture originates from the actual water-holding capacity in different directions. The soil moisture characteristic curve uses soil water content Q (expressed as a volume percentage) as the horizontal axis and soil water suction S (expressed as atmospheric pressure) as the vertical axis.

[0057] In this embodiment of the invention, S22 includes the following sub-steps:

[0058] S221. Weighted fusion of soil relative humidity and wind speed at the current moment yields the soil moisture availability correction coefficient.

[0059] S222. Multiply the soil moisture availability correction coefficient by the crop coefficient and reference evapotranspiration at the current time to obtain the predicted water requirement for the crop.

[0060] In this invention, relative humidity (RH) and wind speed (u) are the core meteorological factors of crop transpiration stress, which directly determine the urgency of soil moisture on crops.

[0061] The lower the relative humidity (RH), the drier the air, the stronger the crop leaves' transpiration demand, and the faster the soil moisture is consumed.

[0062] The higher the wind speed, the faster the crop transpiration rate and the faster the evaporation of soil surface water. However, when the wind speed is too high, soil moisture will quickly be lost to a depth that the crop cannot absorb. At this time, the water availability needs to be appropriately reduced (to avoid ineffective irrigation).

[0063] Weighted fusion can provide ,in, , and These represent the first adaptation weight, the second adaptation weight, and the third adaptation weight, respectively. Dimensionless Can be taken , Can be taken , This indicates that an exponentiation operation is being performed.

[0064] In this embodiment of the invention, in S24, soil moisture loss... The expression is:

[0065] ;

[0066] ;

[0067] in, This represents the average density of all mineral particles in the soil. This indicates the soil's own volume density. This indicates the intake suction power value. The shape parameters representing the soil moisture characteristic curve. Indicates soil matrix potential, Represents the proportionality coefficient. This indicates the current soil moisture loss.

[0068] In this invention, the intake suction value is the critical suction parameter characterizing the entry of air into the pores during the dehydration process of saturated soil. Soil matrix potential is the potential energy generated by the adsorption and capillary forces of the soil matrix, reflecting the state of water free energy in unsaturated soil; its value is 0 in saturated soil and less than 0 in unsaturated soil.

[0069] When it is approximately equal to 1, the current matrix potential is close to the air intake value, which means that the large pores of the soil begin to be filled with water, and the soil moisture content rises rapidly at this time.

[0070] When the value is greater than 1, the current matrix potential is greater than the air intake value. The macropores are already filled with water, and the soil begins to fill the medium and small pores. The rate of increase in water content slows down, making it difficult for crop roots to absorb water.

[0071] When the value is less than 1, the current matrix potential is less than the air intake value, the macropores are not yet filled with water, and the soil is in the dry to slightly moist stage.

[0072] It reflects the density of the mineral particles (such as sand and clay) in the soil itself, and is unrelated to the soil's porosity and water content. Soil bulk density reflects the compactness of the soil; the lower the bulk density, the more pores and the looser the soil; conversely, the higher the bulk density, the more compact the soil (fewer pores).

[0073] In this embodiment of the invention, in S25, the crop water requirement and soil moisture loss are added together, and then the future precipitation replenishment and the current soil effective water content are subtracted in turn. The result is then multiplied by 0 to obtain the irrigation prediction quantity coupled with multiple factors.

[0074] In this invention, future precipitation replenishment is obtained based on precipitation data from short-term weather forecasts.

[0075] The current effective water content of the soil depends on the water absorption capacity of crop root hairs and the soil suction capacity.

[0076] Crop water requirement plus soil moisture loss equals total water consumption (i.e., water consumed by crop growth, soil evaporation, etc.); future rainfall replenishment plus current effective soil moisture content equals existing / future water reserves (i.e., usable water in the soil now plus future rainfall). A positive result indicates that existing reserves are insufficient to meet consumption, and this positive number represents the amount of water needed for additional irrigation. A negative result indicates that existing reserves exceed total consumption, meaning the soil already has enough water and no additional irrigation is needed.

[0077] In this embodiment of the invention, S3 represents the predicted irrigation amount for the soil at the next time step. The expression is:

[0078] ;

[0079] in, This represents time-series driven irrigation forecasts. This represents the irrigation forecast quantity resulting from the coupling of multiple factors. This indicates taking the minimum value.

[0080] In this invention, based on the deviation between the predicted values ​​of the two paths, an adaptive balance is struck between the mean of path consistency and the minimum of irrigation conservatism. The more consistent the paths are, the more trust is placed in the mean; the greater the divergence between the paths, the more the focus shifts towards the conservative minimum. When the deviation between the predicted values ​​of the two paths is extremely small, i.e. When the predicted values ​​approach zero, the time-series driven and multi-factor coupled predictions are consistent, indicating that the information from the two paths corroborates each other, and the mean is the most reasonable result. When the predicted values ​​from the two paths deviate significantly, the core constraint of the irrigation scenario is to avoid over-irrigation (over-irrigation leads to soil compaction and nutrient loss). The multi-factor coupled path is calculated based on the actual physical needs of the soil, weather, and crops, which is closer to the actual water shortage situation. Therefore, it biases towards the minimum value (usually the multi-factor coupled prediction) to ensure the rationality of the irrigation amount. When the predicted values ​​from the two paths deviate moderately, the result is a weighted combination of the mean and the minimum value, which retains the information from both paths without deviating excessively from the conservative requirements.

[0081] Those skilled in the art will recognize that the embodiments described herein are intended to help the reader understand the principles of the invention, and should be understood that the scope of protection of the invention is not limited to such specific statements and embodiments. Those skilled in the art can make various other specific modifications and combinations based on the technical teachings disclosed in this invention without departing from the spirit of the invention, and these modifications and combinations are still within the scope of protection of this invention.

Claims

1. A method for predicting soil irrigation needs based on data analysis, characterized by, The method comprises the following steps: S1, collecting historical data of soil to obtain time series driven irrigation prediction quantity; S2, generating multi-factor coupled irrigation prediction quantity according to a plurality of characteristic parameters of the soil at the current time; S3, fusing the time series driven irrigation prediction quantity and the multi-factor coupled irrigation prediction quantity to determine the irrigation prediction quantity of the soil at the next time.

2. The data analysis based soil irrigation requirement prediction method of claim 1, wherein, In the S1, the historical data of the soil is input into a neural network model to obtain the time series driven irrigation prediction quantity; wherein the historical data comprises irrigation quantity and precipitation in a historical period.

3. The data analysis based soil irrigation requirement prediction method of claim 1, wherein, The S2 comprises the following sub-steps: S21, obtaining future precipitation replenishment quantity; S22, determining crop predicted water requirement; S23, constructing soil moisture characteristic curve for the soil at the current time; S24, determining soil moisture loss according to shape parameters of the soil moisture characteristic curve; S25, obtaining the multi-factor coupled irrigation prediction quantity by adding the future precipitation replenishment quantity, the crop water requirement and the soil moisture loss.

4. The data analysis based soil irrigation requirement prediction method of claim 3, wherein, The S22 comprises the following sub-steps: S221, performing weighted fusion on the soil relative humidity and wind speed at the current time to obtain soil moisture availability correction coefficient; S222, multiplying the soil moisture availability correction coefficient, the crop coefficient of the crop at the current time and the reference evapotranspiration to obtain the crop predicted water requirement.

5. The data analysis based soil irrigation requirement prediction method of claim 3, wherein, In the S24, the soil moisture loss The expression is: ； ； wherein, denotes the mean value of the density of all mineral particles in the soil, denotes the self-volume density of the soil, denotes the air-entry suction value, denotes a shape parameter of the soil moisture characteristic curve, denotes the soil matric potential, denotes a proportionality coefficient, denotes the soil moisture depletion at the current time instant.

6. The data analysis based soil irrigation requirement prediction method of claim 3, wherein, In the S25, after adding the crop water requirement and the soil moisture loss, the future precipitation replenishment quantity and the effective soil moisture content at the current time are subtracted in turn, and the maximum value between the operation result and 0 is obtained to obtain the multi-factor coupled irrigation prediction quantity.

7. The data analysis based soil irrigation requirement prediction method of claim 1, wherein, In the S3, the irrigation prediction of the next time of the soil The expression is: ； wherein, represents a timing-driven irrigation prediction, represents a multi-factor coupled irrigation prediction, represents a minimum.

Images