Soil water flux monitoring method, device and equipment and storage medium

Abstract

The invention provides a soil water flux monitoring method, device and equipment and a storage medium, and the soil water flux monitoring method comprises the following steps: arranging an active heat source and monitoring equipment, and obtaining water content data and temperature data of to-be-detected soil; constructing a soil hydrodynamic constraint based on a Richardson equation and the water content data of the to-be-detected soil, and training a pre-constructed water flux monitoring model; based on a convection-conduction-dispersion equation and the temperature data of the to-be-measured soil, constructing a soil thermodynamic constraint, and training the water flux monitoring model; and constructing a soil hydrothermal coupling constraint, and training the water flux monitoring model to obtain a target water flux monitoring model. According to the invention, the problem that the existing monitoring method is difficult to realize high-precision monitoring of the soil water flux at low data cost under different situations can be solved.

Description

Technical Field

[0001] The present invention relates to the technical field of soil water flux monitoring, and particularly relates to a method, device, equipment and storage medium for monitoring soil water flux. Background Art

[0002] Soil water flux plays a key role in many fields such as agricultural production, ecological protection and geological disaster warning. However, there is currently no general direct monitoring equipment for soil water flux. Existing indirect measurement methods have relatively strict requirements for the environment or equipment working conditions, and the physical and chemical properties of soils in China have strong spatio-temporal variability, which poses great challenges to the monitoring of soil water flux. Therefore, there is an urgent need for a high-precision and portable method for measuring soil water flux to meet the actual needs.

[0003] Soil moisture content and soil temperature are commonly used soil monitoring indicators. At present, soil moisture content can be combined with the Richards equation to estimate soil water flux. However, in high-flux or variable-saturation scenarios, the change in moisture content is tiny, resulting in the moisture content data being difficult to contain effective water flux information, thus reducing the accuracy of water flux monitoring. Soil temperature can be combined with the convection-dispersion equation of soil to calculate soil water flux. Among them, the thermal pulse active heat source technology has great application potential, but it has strict control over the calorific value. Therefore, in the face of heterogeneous soils and heat non-closure scenarios, its accuracy will be significantly reduced. There are also methods using hydrothermal coupling to estimate soil water flux, but such methods currently require the hydrothermal characteristics of the soil as a priori, which limits their further popularization and application.

[0004] For the problem that existing monitoring methods are difficult to achieve high-precision monitoring of soil water flux at a low data cost in different scenarios, no effective solution has been proposed yet. Summary of the Invention

[0005] The present invention provides a method, device, equipment and storage medium for monitoring soil water flux to solve the defect that existing monitoring methods are difficult to achieve high-precision monitoring of soil water flux at a low data cost in different scenarios.

[0006] In the first aspect, the present invention provides a method for monitoring soil water flux, including: Deploy an active heat source and monitoring equipment, and obtain the moisture content data and temperature data of the soil to be measured; Based on the Richards equation and the moisture content data of the soil to be measured, construct a soil hydrodynamics constraint, and train a pre-constructed water flux monitoring model; Based on the convection-dispersion equation and the temperature data of the soil to be measured, construct a soil thermodynamics constraint, and train the water flux monitoring model; Construct a soil water-heat coupling constraint, train the water flux monitoring model, and obtain a target water flux monitoring model; Obtain the spatio-temporal coordinates of the measurement position of the soil to be measured, and determine the soil water flux at the measurement position of the soil to be measured through the target water flux monitoring model.

[0007] According to a soil water flux monitoring method provided by the present invention, an active heat source and monitoring equipment are arranged, and water content data and temperature data of the soil to be measured are obtained, including: Determine the monitoring area of the soil to be measured, and arrange an active heat source and monitoring equipment in the monitoring area of the soil to be measured; Based on the monitoring area and the monitoring positions of the monitoring equipment, determine a set of measurement points and a set of collocation points for training; the set of measurement points contains several measurement point coordinates, and the set of collocation points contains several collocation point coordinates; Obtain the water content data and temperature data at the positions where the measurement point coordinates of the soil to be measured are located.

[0008] Before obtaining the water content data and temperature data at the positions where the measurement point coordinates of the soil to be measured are located according to a soil water flux monitoring method provided by the present invention, it includes: performing normalization processing on the measurement point coordinates and collocation point coordinates of the soil to be measured.

[0009] According to a soil water flux monitoring method provided by the present invention, based on the Richards equation and the water content data of the soil to be measured, construct a soil hydrodynamics constraint, and train a pre-constructed water flux monitoring model, including: Predict the water content, matrix potential, and hydraulic conductivity at the positions where the measurement point coordinates and collocation point coordinates of the soil to be measured are located through the water flux monitoring model, and obtain a water content prediction value; Based on the set of measurement points and collocation points of the soil to be measured and the water content prediction value, construct a soil hydrodynamics constraint, and determine the loss of the soil hydrodynamics constraint of the water flux monitoring model; With the goal of minimizing the loss of the soil hydrodynamics constraint of the water flux monitoring model, train the water flux monitoring model and update the parameters.

[0010] According to a soil water flux monitoring method provided by the present invention, based on the convection-dispersion equation and the temperature data of the soil to be measured, construct a soil thermodynamics constraint, and train the water flux monitoring model, including: Obtain the temperature, thermal dispersion coefficient, thermal conductivity, and volume heat capacity at the positions where the measurement point coordinates and collocation point coordinates of the soil to be measured are located through the water flux monitoring model, and obtain a temperature prediction value; Based on the measurement point set and collocation point set of the soil to be measured and the predicted temperature values, construct soil thermodynamics constraints and determine the soil thermodynamics constraint loss of the water flux monitoring model; With the goal of minimizing the soil thermodynamics constraint loss of the water flux monitoring model, train the water flux monitoring model and update its parameters.

[0011] According to a soil water flux monitoring method provided by the present invention, constructing soil water-heat coupling constraints, training the water flux monitoring model, and obtaining a target water flux monitoring model, including: Determine the water-heat coupling constraint loss based on the soil hydrodynamics constraint loss and soil thermodynamics constraint loss of the water flux monitoring model; With the goal of minimizing the water-heat coupling constraint loss of the water flux monitoring model, train the water flux monitoring model and update its parameters to obtain a target water flux monitoring model.

[0012] According to a soil water flux monitoring method provided by the present invention, determining the soil water flux at the position to be measured of the soil to be measured through the target water flux monitoring model, including: Input the spatio-temporal coordinates of the position to be measured of the soil to be measured into the target water flux monitoring model to obtain the water content and soil temperature of the soil to be measured at the position to be measured; Based on the water content and soil temperature of the soil to be measured at the position to be measured, determine the soil water flux at the position to be measured of the soil to be measured.

[0013] In a second aspect, the present invention also provides a soil water flux monitoring device, including: A data acquisition module, configured to deploy an active heat source and monitoring equipment, and acquire the water content data and temperature data of the soil to be measured; A hydrodynamic training module, configured to construct soil hydrodynamics constraints based on the Richards equation and the water content data of the soil to be measured, and train a pre-constructed water flux monitoring model; A thermodynamic training module, configured to construct soil thermodynamics constraints based on the convection-dispersion equation and the temperature data of the soil to be measured, and train the water flux monitoring model; A coupling training module, configured to construct soil water-heat coupling constraints, train the water flux monitoring model, and obtain a target water flux monitoring model; A monitoring module, configured to acquire the spatio-temporal coordinates of the position to be measured of the soil to be measured, and determine the soil water flux at the position to be measured of the soil to be measured through the target water flux monitoring model.

[0014] In a third aspect, the present invention further provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, the soil water flux monitoring method described in the first aspect above is implemented.

[0015] In a fourth aspect, the present invention further provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the soil water flux monitoring method described in the first aspect above is implemented.

[0016] In a fifth aspect, the present invention further provides a computer program product, including a computer program. When the computer program is executed by a processor, the soil water flux monitoring method described in the first aspect above is implemented.

[0017] Compared with the prior art, the present invention has the following beneficial effects: The soil water flux monitoring method provided by the present invention can achieve high-precision monitoring of soil water flux only by using soil moisture content and temperature data in combination with an active heat source, and the data cost is lower. Moreover, this method has higher monitoring accuracy in variable saturation and high-flux scenarios. Compared with traditional active heat source technologies such as heat pulse, this method does not require strict control of the heat source heating amount, and makes up for the deficiency of heat pulse technology in considering soil heterogeneity with a hydrothermal coupling technology; compared with existing hydrothermal coupling technologies, this method does not require the hydrothermal characteristics of the soil as prior information and has better generalization performance. Through this method, the problem that existing monitoring methods are difficult to achieve high-precision monitoring of soil water flux at a low data cost in different scenarios can be solved. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0019] Figure 1 is a flowchart of the soil water flux monitoring method provided by the present invention; Figure 2 is a schematic diagram of the water flux acquisition process in an embodiment of the present invention; Figure 3 is a generalization diagram of the training process of the water flux monitoring model in an embodiment of the present invention; Figure 4 is a comparison bar chart of the water flux monitoring errors of the method of the present invention and the method using only moisture content data in different flow state scenarios in an embodiment of the present invention; Figure 5It is a comparison chart of using this method and only using water content data to estimate the water flux field and its error in sandy soil and loamy sandy soil in the embodiments of the present invention; Figure 6 It is the effect diagram of the model inverting soil hydrothermal parameters in the embodiments of the present invention; Figure 7 It is the structural block diagram of a soil water flux monitoring device provided by the present invention; Figure 8 It is the structural schematic diagram of an electronic device provided by the present invention. Specific embodiments

[0020] To make the objectives, technical solutions and advantages of the present invention clearer, the technical solutions in the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0021] The present invention provides a method for monitoring soil water flux, Figure 1 It is the flow chart of the soil water flux monitoring method provided by the present invention, as Figure 1 shown, and the method includes the following steps: Step S101, arrange active heat sources and monitoring devices, and obtain the water content data and temperature data of the soil to be measured; Step S102, based on the Richards equation and the water content data of the soil to be measured, construct a soil hydrodynamics constraint, and train the pre-constructed water flux monitoring model; Step S103, based on the convection-diffusion equation and the temperature data of the soil to be measured, construct a soil thermodynamics constraint, and train the water flux monitoring model; Step S104, construct a soil hydrothermal coupling constraint, train the water flux monitoring model, and obtain the target water flux monitoring model; Step S105, obtain the spatio-temporal coordinates of the position to be measured of the soil to be measured, and determine the soil water flux at the position to be measured of the soil to be measured through the target water flux monitoring model.

[0022] In this method, first, an active heat source and soil moisture content and temperature monitoring devices are deployed, and the true moisture content data and temperature data of the corresponding positions of the soil to be measured are obtained through the monitoring devices. Then, based on the Richards equation and the moisture content data of the soil to be measured, a soil hydrodynamics constraint is constructed, and the pre-constructed water flux monitoring model is trained to initially solve the hydraulic state of the soil profile. Among them, the basic architecture of the water flux monitoring model can adopt Physics-Informed Neural Networks (PINN). Then, the convective-dispersive equation and the soil temperature data are used to construct a soil thermodynamics constraint to continue training the water flux monitoring model, and the soil thermodynamics parameters are initialized. Then, a soil water-heat coupling constraint is constructed, and the water flux monitoring model is trained to update all soil water-heat parameters and states. Finally, the spatio-temporal coordinates of the position to be monitored are input into the trained target water flux monitoring model to obtain the soil water flux. In the above process, only the soil moisture content and temperature data are used, combined with an active heat source, to achieve high-precision monitoring of the soil water flux, and the data cost is lower. Moreover, this method has higher monitoring accuracy in variable saturation and high-throughput scenarios. Compared with traditional active heat source technologies such as thermal pulses, this method does not require strict control of the heat input of the heat source, and makes up for the lack of consideration of soil heterogeneity by thermal pulse technology with water-heat coupling technology; compared with existing water-heat coupling technologies, this method does not require the water-heat characteristics of the soil as prior information and has better generalization performance. Through this method, the problem that existing monitoring methods are difficult to achieve high-precision monitoring of soil water flux at low data costs in different scenarios can be solved.

[0023] Figure 2 is a schematic diagram of the water flux acquisition process in an embodiment of the present invention. As Figure 2 shown, in some of these embodiments, step S101, deploying an active heat source and monitoring devices, and obtaining the moisture content data and temperature data of the soil to be measured, includes: determining the monitoring area of the soil to be measured, and deploying an active heat source and monitoring devices in the monitoring area of the soil to be measured; based on the monitoring area and the monitoring positions of the monitoring devices, determining a set of measurement points and a set of collocation points for training; the set of measurement points contains several measurement point coordinates, and the set of collocation points contains several collocation point coordinates; obtaining the moisture content data and temperature data of the positions where the measurement point coordinates of the soil to be measured are located.

[0024] Further, before obtaining the moisture content data and temperature data of the positions where the measurement point coordinates of the soil to be measured are located, it includes: normalizing the measurement point coordinates and collocation point coordinates of the soil to be measured.

[0025] Exemplarily, after determining the soil water flux monitoring area If the upward direction is defined as the positive direction of the spatial coordinate, the spatial coordinate is expressed as , and the time coordinate is expressed as Then, arrange the active heat source at , and arrange the soil moisture content and temperature monitoring equipment at . The two are in a straight line longitudinally. In this embodiment, the active heat source is the starting position line of the 0m space, and the soil moisture content and temperature monitoring equipment are arranged at -0.05, -0.45, and -0.85m respectively.

[0026] According to the monitoring area and the soil moisture content and temperature monitoring positions, the measurement point coordinates and collocation point coordinates for training can be obtained:

[0027] Among them, is the measurement point coordinate, is the original measurement point space coordinate, is the original measurement point time coordinate; is the set of points composed of all measurement point coordinates; is the collocation point coordinate, is the original collocation point space coordinate, is the original collocation point time coordinate, is the Latin hypercube sampling function. The obtained measurement point and collocation point coordinates need to be normalized:

[0028]

[0029] Among them, is the normalized measurement point space coordinate; is the normalized measurement point time coordinate; is the normalized measurement point coordinate; is the normalized collocation point space coordinate; is the normalized measurement point time coordinate; is the normalized collocation point coordinate; is the lower boundary position (m) of the flux monitoring area; is the upper boundary position (m) of the flux monitoring area; is the initial time (day) of the flux monitoring area; is the termination time (day) of the flux monitoring area. At the same time, according to the measurement point set match the corresponding soil volumetric water content , soil temperature data to produce the dataset required for soil water flux monitoring.

[0030] Figure 3It is a generalization diagram of the training process of the water flux monitoring model in the embodiments of the present invention. As shown in Figure 3 In some of these embodiments, in step S102, based on the Richards equation and the water content data of the soil to be measured, a soil hydrodynamics constraint is constructed, and the pre-constructed water flux monitoring model is trained, including: predicting the water content, matrix potential, and hydraulic conductivity at the measurement point coordinates and collocation point coordinates of the soil to be measured through the water flux monitoring model to obtain the predicted water content value; constructing a soil hydrodynamics constraint based on the measurement point set and collocation point set of the soil to be measured and the predicted water content value, and determining the soil hydrodynamics constraint loss of the water flux monitoring model; training and updating the parameters of the water flux monitoring model with the goal of minimizing the soil hydrodynamics constraint loss of the water flux monitoring model.

[0031] Exemplarily, first, a soil hydrodynamics neural network is built to estimate the matrix potential , water content , and hydraulic conductivity of the soil profile:

[0032]

[0033]

[0034] Among them, is the predicted value of the soil matrix potential (m); is the neural network for predicting the soil matrix potential (m); is the predicted value of the soil water content (m 3 m -3 ); is the neural network for predicting the soil water content (m 3 m -3 ); is the predicted value of the soil hydraulic conductivity (m·day -1 ); is the mapping function for predicting the soil hydraulic conductivity (m·day -1 ); are the updatable parameters of the neural network; are the updatable parameters of the hydraulic conductivity mapping function.

[0035] Input the measurement point set , collocation point set into the soil hydrodynamics neural network, and construct a soil hydrodynamics constraint based on the measured soil water content data:

[0036]

[0037]

[0038] Among them, and are the hydrodynamic observation loss and the hydrodynamic collocation residual respectively; is the sequence number of the points in the point set; and are the numbers of the measurement points and the collocation points respectively; is the neural network self-differentiation operation; and are the weight coefficients of the hydrodynamic observation loss and the hydrodynamic collocation residual respectively; is the soil hydrodynamics constraint loss.

[0039] Finally, based on the soil hydrodynamics constraint loss , train the water flux monitoring model and update the parameters:

[0040] Among them, , and are all the hydrodynamics parameters to be updated.

[0041] Correspondingly, in step S103, based on the convection-diffusion equation and the temperature data of the soil to be measured, construct the soil thermodynamics constraint and train the water flux monitoring model, including: obtaining the temperature, thermal dispersion coefficient, thermal conductivity and volumetric heat capacity at the positions of the measurement points and the collocation points of the soil to be measured through the water flux monitoring model, and obtaining the temperature prediction value; based on the measurement point set and the collocation point set of the soil to be measured and the temperature prediction value, construct the soil thermodynamics constraint and determine the soil thermodynamics constraint loss of the water flux monitoring model; aiming at minimizing the soil thermodynamics constraint loss of the water flux monitoring model, train and update the parameters of the water flux monitoring model.

[0042] Exemplarily, first, build a soil thermodynamics neural network to estimate the temperature of the soil profile, the thermal dispersion coefficient the thermal conductivity the volumetric heat capacity

[0043]

[0044]

[0045] Among them, is the predicted value of soil temperature (°C); is the neural network for predicting soil temperature (°C), are the updatable parameters of this network; is the predicted value of soil thermal conductivity ( ); is the mapping function for predicting soil thermal conductivity ( ), are the updatable parameters of this mapping function; is the predicted value of soil volumetric heat capacity ( ); is the volumetric heat capacity of water body ( ); is the volumetric heat capacity of soil particles ( ); is the soil saturated volumetric water content (m 3 m -3 ).

[0046] Then, input the measurement point set , collocation point set into the soil thermodynamics neural network, and construct the soil thermodynamics constraint based on the measured soil temperature data:

[0047]

[0048]

[0049] Among them, and are the thermodynamics observation loss and thermodynamics collocation point residual respectively; and are the weight coefficients of the thermodynamics observation loss and thermodynamics collocation point residual respectively; is the soil thermodynamics constraint loss, is the estimated value of thermal dispersion coefficient. represents the predicted value of soil temperature at sequence number i , represents the soil temperature at sequence number i , represents the updatable parameters of the hydraulic conductivity mapping function, represents the sequence number i at which the soil thermal conductivity is located, represents the sequence number i at which the hydraulic conductivity is located, represents the sequence number i at which the soil matric potential is located, Indicates the sequence number i of the spatial coordinates at

[0050] Based on the soil thermodynamics constraint loss , train the network and update the parameters to initialize the thermodynamics parameters:

[0051] where and are both thermodynamics parameters to be updated.

[0052] Based on the above embodiments, in step S104, construct a soil water-heat coupling constraint, train a water flux monitoring model, and obtain a target water flux monitoring model, including: determining a water-heat coupling constraint loss based on the soil hydrodynamics constraint loss and the soil thermodynamics constraint loss of the water flux monitoring model; training and parameter updating the water flux monitoring model with the goal of minimizing the water-heat coupling constraint loss of the water flux monitoring model to obtain a target water flux monitoring model.

[0053] Exemplarily, the water-heat coupling constraint loss can be composed of the hydrodynamics constraint loss and the thermodynamics constraint loss , and the specific formula is as follows:

[0054] where is the water-heat coupling constraint loss; is the weight coefficient of the hydrodynamics constraint loss; is the weight coefficient of the thermodynamics constraint loss. Update all model parameters according to the water-heat coupling constraint loss :

[0055] where , , , and are all model parameters to be updated.

[0056] During the above training process, the parameter update is divided into three stages. The parameters updated in the first stage only involve the hydrodynamics constraint, and the parameters updated in the second stage only involve the thermodynamics constraint. However, neither water nor heat alone can well approximate the true soil water-heat state distribution. Therefore, in the third stage, these parameters are updated simultaneously to impose the concept of water-heat coupling. Moreover, adding the first stage and the second stage additionally before the simultaneous update can reduce the training difficulty of the water flux monitoring model and prevent overfitting or model collapse.

[0057] In some of these embodiments, in step S105, determining the soil water flux at the position to be measured of the soil to be measured through the target water flux monitoring model includes: inputting the spatio-temporal coordinates of the position to be measured of the soil to be measured into the target water flux monitoring model to obtain the water content and soil temperature of the soil to be measured at the position to be measured; and determining the soil water flux at the position to be measured of the soil to be measured based on the water content and soil temperature of the soil to be measured at the position to be measured.

[0058] Exemplarily, specify any position , input it into the trained model, obtain the soil water and heat state at the corresponding position, and calculate the soil water flux using Darcy's law:

[0059] Among them, is the calculated water flux (m·day -1 ).

[0060] To verify the effectiveness of the above method, the following compares the estimation results of the water flux field of this method with those driven only by water content data at a depth of 1 m in sandy soil, sandy loam, loamy sand, and loam under various different flow states such as free drainage, seepage surface, and groundwater within 2.5 days, as Figure 4 shown, Figure 4 is the comparison bar chart of the water flux monitoring errors of this method and those using only water content data in different flow state scenarios in the embodiments of the present invention. And in this experiment, the inversion of the flux field in sandy loam was emphasized, as Figure 5 shown, Figure 5 is the comparison chart of the estimated water flux field and its error of this method and those using only water content data in sandy soil and sandy loam in the embodiments of the present invention. And in this experiment, the water and heat characteristics inverted by this method were also compared with the true values, and the results are as Figure 6 shown, Figure 6 is the effect diagram of the model inversion of soil water and heat parameters in the embodiments of the present invention. The results show that the water flux monitoring method using active thermal tracer proposed by the present invention has higher flux monitoring accuracy, performs well in different flow states and different soils, and has excellent generalization ability.

[0061] The present invention also provides a soil water flux monitoring device. The soil water flux monitoring device provided by the present invention will be described below. The soil water flux monitoring device described below can be correspondingly referred to the soil water flux monitoring method described above. Figure 7 is the structural block diagram of the soil water flux monitoring device provided by the present invention, as Figure 7 shown. The device includes: The data acquisition module 701 is used to deploy active heat sources and monitoring devices, and acquire the water content data and temperature data of the soil to be measured. The hydrodynamic training module 702 is used to construct soil hydrodynamics constraints based on the Richards equation and the water content data of the soil to be measured, and train the pre-constructed water flux monitoring model. The thermodynamics training module 703 is used to construct soil thermodynamics constraints based on the convection-diffusion equation and the temperature data of the soil to be measured, and train the water flux monitoring model. The coupled training module 704 is used to construct soil hydrothermal coupling constraints, train the water flux monitoring model, and obtain the target water flux monitoring model. The monitoring module 705 is used to obtain the spatio-temporal coordinates of the position to be measured of the soil to be measured, and determine the soil water flux of the position to be measured of the soil to be measured through the target water flux monitoring model.

[0062] When this device is in use, first, active heat sources and soil water content and temperature monitoring devices are deployed. The data acquisition module 701 acquires the real water content data and temperature data of the corresponding position of the soil to be measured through the monitoring devices. Then, the hydrodynamic training module 702 constructs soil hydrodynamics constraints based on the Richards equation and the water content data of the soil to be measured, and trains the pre-constructed water flux monitoring model to initially solve the hydraulic state of the soil profile. Among them, the basic architecture of the water flux monitoring model can adopt Physics-Informed Neural Networks (PINN). The thermodynamics training module 703 then uses the convection-diffusion equation and the soil temperature data to construct soil thermodynamics constraints to continue training the water flux monitoring model and initialize the soil thermodynamics parameters. Then, the coupled training module 704 constructs soil hydrothermal coupling constraints and trains the water flux monitoring model to update all soil hydrothermal parameters and states. Finally, the monitoring module 705 inputs the spatio-temporal coordinates of the position to be monitored into the trained target water flux monitoring model to obtain the soil water flux. In the above process, only the soil water content and temperature data are used, combined with an active heat source, to achieve high-precision monitoring of the soil water flux, and the data cost is lower. Moreover, this device has higher monitoring accuracy in variable saturation and high-flux scenarios. Compared with traditional active heat source technologies such as heat pulse, this method does not require strict control of the heat input of the heat source, and makes up for the deficiency of the heat pulse technology in considering soil heterogeneity with the hydrothermal coupling technology; compared with the existing hydrothermal coupling technologies, this device does not require the hydrothermal characteristics of the soil as prior information and has better generalization performance. Through this device, the problem that the existing monitoring methods are difficult to achieve high-precision monitoring of the soil water flux at a low data cost in different scenarios can be solved.

[0063] Figure 8Illustrates a schematic diagram of the physical structure of an electronic device, as Figure 8 shown. The electronic device may include: a processor 801, a communications interface 802, a memory 803, and a communication bus 804. Among them, the processor 801, the communications interface 802, and the memory 803 communicate with each other through the communication bus 804. The processor 801 may call the logical instructions in the memory 803 to execute a method for monitoring soil water flux. The method includes: Deploy an active heat source and monitoring devices, and obtain the water content data and temperature data of the soil to be measured; Based on the Richards equation and the water content data of the soil to be measured, construct a soil hydrodynamics constraint, and train a pre-constructed water flux monitoring model; Based on the convection-diffusion equation and the temperature data of the soil to be measured, construct a soil thermodynamics constraint, and train the water flux monitoring model; Construct a soil water-heat coupling constraint, train the water flux monitoring model, and obtain a target water flux monitoring model; Obtain the spatio-temporal coordinates of the position to be measured of the soil to be measured, and determine the soil water flux at the position to be measured of the soil to be measured through the target water flux monitoring model.

[0064] In addition, when the logical instructions in the above-mentioned memory 803 are implemented in the form of software functional units and sold or used as an independent product, they may be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of the technical solution, may be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods in various embodiments of the present invention. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs that can store program codes.

[0065] On the other hand, the present invention also provides a computer program product. The computer program product includes a computer program. The computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the method for monitoring soil water flux provided by the above-mentioned various methods. The method includes: Deploy an active heat source and monitoring devices, and obtain the water content data and temperature data of the soil to be measured; Based on the Richards equation and the water content data of the soil to be measured, construct a soil hydrodynamics constraint and train the pre-constructed water flux monitoring model; Based on the convection-dispersion equation and the temperature data of the soil to be measured, construct a soil thermodynamics constraint and train the water flux monitoring model; Construct a soil water-heat coupling constraint, train the water flux monitoring model, and obtain the target water flux monitoring model; Obtain the spatio-temporal coordinates of the position to be measured of the soil to be measured, and determine the soil water flux at the position to be measured of the soil to be measured through the target water flux monitoring model.

[0066] In another aspect, the present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it is used to execute the soil water flux monitoring method provided by the above-mentioned various methods. The method includes: Deploy an active heat source and monitoring equipment, and obtain the water content data and temperature data of the soil to be measured; Based on the Richards equation and the water content data of the soil to be measured, construct a soil hydrodynamics constraint and train the pre-constructed water flux monitoring model; Based on the convection-dispersion equation and the temperature data of the soil to be measured, construct a soil thermodynamics constraint and train the water flux monitoring model; Construct a soil water-heat coupling constraint, train the water flux monitoring model, and obtain the target water flux monitoring model; Obtain the spatio-temporal coordinates of the position to be measured of the soil to be measured, and determine the soil water flux at the position to be measured of the soil to be measured through the target water flux monitoring model.

[0067] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated. The components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art can understand and implement it without creative labor.

[0068] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on such an understanding, the essence of the above technical solution, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0069] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for monitoring soil water flux, characterized in that, Including: Deploy an active heat source and monitoring equipment, and obtain the moisture content data and temperature data of the soil to be measured; Based on the Richards equation and the moisture content data of the soil to be measured, construct a soil hydrodynamics constraint, and train a pre-constructed water flux monitoring model; Based on the convection-diffusion equation and the temperature data of the soil to be measured, construct a soil thermodynamics constraint, and train the water flux monitoring model; Construct a soil water-heat coupling constraint, train the water flux monitoring model, and obtain a target water flux monitoring model; Obtain the spatio-temporal coordinates of the position to be measured of the soil to be measured, and determine the soil water flux at the position to be measured of the soil to be measured through the target water flux monitoring model.

2. The soil water flux monitoring method according to claim 1, wherein Deploy an active heat source and monitoring equipment, and obtain the moisture content data and temperature data of the soil to be measured, including: Determine the monitoring area of the soil to be measured, and deploy an active heat source and monitoring equipment in the monitoring area of the soil to be measured; Based on the monitoring area and the monitoring positions of the monitoring equipment, determine a set of measurement points and a set of collocation points for training; the set of measurement points contains several measurement point coordinates, and the set of collocation points contains several collocation point coordinates; Obtain the moisture content data and temperature data at the positions where the measurement point coordinates of the soil to be measured are located.

3. The soil water flux monitoring method according to claim 2, characterized in that Before obtaining the moisture content data and temperature data at the positions where the measurement point coordinates of the soil to be measured are located, it includes: normalizing the measurement point coordinates and collocation point coordinates of the soil to be measured.

4. The soil water flux monitoring method according to claim 2, characterized in that, Based on the Richards equation and the moisture content data of the soil to be measured, construct a soil hydrodynamics constraint, and train a pre-constructed water flux monitoring model, including: Predict the moisture content, matrix potential, and hydraulic conductivity at the positions where the measurement point coordinates and collocation point coordinates of the soil to be measured are located through the water flux monitoring model, and obtain predicted moisture content values; Based on the set of measurement points and the set of collocation points of the soil to be measured and the predicted moisture content values, construct a soil hydrodynamics constraint, and determine the loss of the soil hydrodynamics constraint of the water flux monitoring model; With the goal of minimizing the loss of the soil hydrodynamics constraint of the water flux monitoring model, train the water flux monitoring model and update the parameters.

5. The soil water flux monitoring method according to claim 4, characterized in that Based on the convection-diffusion equation and the temperature data of the soil to be measured, construct a soil thermodynamics constraint, and train the water flux monitoring model, including: Obtain the temperature, thermal dispersion coefficient, thermal conductivity, and volumetric heat capacity at the positions where the measurement point coordinates and collocation point coordinates of the soil to be measured are located through the water flux monitoring model, and obtain predicted temperature values; Based on the set of measurement points and the set of collocation points of the soil to be measured and the predicted temperature values, construct a soil thermodynamics constraint, and determine the loss of the soil thermodynamics constraint of the water flux monitoring model; With the goal of minimizing the loss of the soil thermodynamics constraint of the water flux monitoring model, train the water flux monitoring model and update the parameters.

6. The soil water flux monitoring method according to claim 5, wherein, Construct a soil water-heat coupling constraint, train the water flux monitoring model, and obtain a target water flux monitoring model, including: Determine the water-heat coupling constraint loss based on the loss of the soil hydrodynamics constraint and the loss of the soil thermodynamics constraint of the water flux monitoring model; The water flux monitoring model is trained and its parameters are updated with the goal of minimizing the loss of the hydrothermal coupling constraint of the water flux monitoring model, to obtain a target water flux monitoring model.

7. The soil water flux monitoring method according to claim 1, characterized in that, Determining the soil water flux at the position to be measured of the soil to be measured through the target water flux monitoring model includes: Inputting the spatio-temporal coordinates of the position to be measured of the soil to be measured into the target water flux monitoring model to obtain the water content and soil temperature of the soil to be measured at the position to be measured; Based on the water content and soil temperature of the soil to be measured at the position to be measured, determining the soil water flux at the position to be measured of the soil to be measured.

8. A soil water flux monitoring device, characterized in that, It includes: A data acquisition module, configured to deploy an active heat source and monitoring equipment, and acquire the water content data and temperature data of the soil to be measured; A hydrodynamic training module, configured to construct a soil hydrodynamics constraint based on the Richards equation and the water content data of the soil to be measured, and train a pre-constructed water flux monitoring model; A thermodynamics training module, configured to construct a soil thermodynamics constraint based on the convection-diffusion equation and the temperature data of the soil to be measured, and train the water flux monitoring model; A coupling training module, configured to construct a soil hydrothermal coupling constraint, train the water flux monitoring model, and obtain a target water flux monitoring model; A monitoring module, configured to acquire the spatio-temporal coordinates of the position to be measured of the soil to be measured, and determine the soil water flux at the position to be measured of the soil to be measured through the target water flux monitoring model.

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

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the soil water flux monitoring method according to any one of claims 1 to 7.

Images