Method and system for simultaneously measuring pore ice content and unfrozen water content of frozen soil

Abstract

The invention provides a method and system for simultaneously measuring frozen soil pore ice content and unfrozen water content, and belongs to the technical field of frozen soil parameter measurement based on machine learning, and the method comprises the following steps: acquiring a temperature response curve and specific frozen soil parameters of frozen soil under a pulse heat source; the specific frozen soil parameters comprise frozen soil particle density, frozen soil particle heat conductivity coefficient, frozen soil particle specific heat capacity and frozen soil particle specific gravity; and processing the obtained temperature response curve and the specific frozen soil parameters by using a pre-trained inversion analysis model to obtain the frozen soil ice content and the unfrozen water content. The ice content and the unfrozen water content of the frozen soil can be measured at the same time within one minute, the measurement efficiency is greatly improved, and assumed conditions of an empirical model are avoided at the same time.

Description

Technical Field

[0001] The present invention relates to the technical field of frozen soil parameter measurement based on machine learning, and in particular to a method and system for simultaneously measuring the ice content and unfrozen water content in frozen soil pores. Background Art

[0002] A short and constant pulse heat source is applied to the frozen soil. The curve of temperature change over time at a certain location is measured (called temperature response curve, TRC). TRC is directly controlled by the thermal parameters of the frozen soil (i.e., λ and C). When the ice and water content in the frozen soil changes, the λ and C of the frozen soil also change, causing the characteristics of TRC to change. Therefore, TRC contains information about ice and water. In other words, if TRC can be accurately measured, it can be inverted and analyzed to obtain ice content information.

[0003] At present, the methods for measuring unfrozen water content and ice content have their own limitations. The methods for testing soil ice content can be divided into direct methods and indirect methods. Indirect methods usually require the total moisture content of the soil to be known first, and then measure the unfrozen water content, and calculate the ice content through the two. For example, neutron scattering method and nuclear magnetic resonance method. This type of method is only applicable to closed systems or when the total moisture content is known. However, frozen soil in nature is an open system, and the continuous migration of unfrozen water will cause the total moisture content to change continuously. Therefore, the scope of application of indirect methods is limited. Direct methods do not require the measurement of total moisture content, such as expansion method and transient heat source method. These methods rely heavily on complex instruments, which reduces measurement efficiency and makes it difficult to quickly measure ice content and unfrozen water content in field environments.

[0004] The transient heat source method has certain advantages in measuring ice content due to the sensitivity of frozen soil to temperature. The basic principle of the transient heat source method is to place a short pulse heat source in the soil and interpret the frozen soil parameters by analyzing the temperature response. This method can easily cause ice in the frozen soil to melt, thus affecting the measurement results. However, the ice / water phase change requires a large and continuous energy input. As long as the power / time of the input energy is controlled, the impact of the phase change can be reduced to a certain extent. Li Xu [1] et al. proposed a pulse heat probe method based on the transient heat source method. By measuring the temperature response in the frozen soil, the ice content of the frozen soil was calculated. Wu Bing [2] et al. developed an actively heated fiber Bragg grating sensor based on the transient heat source method for measuring ice content.

[0005] The existing calculation models based on transient heat source method are all empirical models, which not only lack theoretical basis, but also contain many empirical parameters. In addition, these parameters need to be calibrated through a large number of laboratory experiments, which further reduces the measurement efficiency. There are two problems with the two measurement methods of Li Xu et al. [1] and Wu Bing et al. [2]: 1) The calculation models of ice content are all empirical models, which contain many assumptions and empirical parameters and lack a solid theoretical basis, resulting in certain calculation errors; 2) The undetermined parameters in the calculation model need to be calibrated through a large number of laboratory experiments, which limits the measurement accuracy and efficiency. Summary of the invention

[0006] The object of the present invention is to provide a method and system for simultaneously measuring the ice content and unfrozen water content in frozen soil pores, so as to solve at least one technical problem existing in the above-mentioned background technology.

[0007] In order to achieve the above object, the present invention adopts the following technical solutions:

[0008] In a first aspect, the present invention provides a method for simultaneously measuring the pore ice content and unfrozen water content of frozen soil, comprising:

[0009] Obtaining a temperature response curve of frozen soil under a pulse heat source and specific frozen soil parameters; the specific frozen soil parameters include frozen soil particle density, frozen soil particle thermal conductivity, frozen soil particle specific heat capacity and frozen soil particle specific gravity;

[0010] The obtained temperature response curve and specific frozen soil parameters are processed by a pre-trained inversion analysis model to obtain the frozen soil ice content and unfrozen water content; wherein, the inversion analysis model is trained by a training set, and the training set includes multiple groups of data, each group of data includes a temperature response curve, and each temperature response curve is labeled with the frozen soil ice content, the frozen soil unfrozen water content, the frozen soil particle density, the thermal conductivity of the frozen soil particles, the specific heat capacity of the frozen soil particles and the specific gravity of the frozen soil particles; a machine learning algorithm is used, with the temperature response curve and different frozen soil parameters as input, and the frozen soil ice content and the unfrozen water content of the frozen soil as output, to train the training set and extract features, compare the prediction effects when different frozen soil parameters are used as input, and obtain the optimal inversion analysis model.

[0011] Furthermore, a set of parameters including the ice content of frozen soil, the unfrozen water content of frozen soil, the density of frozen soil particles, the thermal conductivity of frozen soil particles, the specific heat capacity of frozen soil particles and the specific gravity of frozen soil particles are set, and a set of thermal conductivity and specific heat capacity of frozen soil are calculated according to the set parameters; based on the transient heat conduction numerical simulation model of frozen soil, the thermal conductivity and specific heat capacity of frozen soil under heat pulse are set to obtain the corresponding temperature response curve.

[0012] Furthermore, Monte Carlo sampling was performed on the set frozen soil ice content, frozen soil unfrozen water content, frozen soil particle density, frozen soil particle thermal conductivity, frozen soil particle specific heat capacity and frozen soil particle specific gravity, and multiple sets of frozen soil thermal conductivity and frozen soil specific heat capacity were calculated. The obtained multiple sets of frozen soil thermal conductivity and frozen soil specific heat capacity were input into the transient heat conduction numerical simulation model to obtain the corresponding multiple sets of temperature response curves.

[0013] Furthermore, the following four equations are used to calculate the thermal conductivity parameters and density of frozen soil:

[0014]

[0015] ρ f =ρ s +0.92θ i +θ w

[0016]

[0017] Among them, ρ f is the density of frozen soil, C f is the specific heat capacity of frozen soil, λ f is the thermal conductivity of frozen soil; ρ s is the density of soil particles; θ is the volume fraction; φ is the mass fraction, and Gs is the specific gravity of soil particles.

[0018] Furthermore, in the transient heat conduction numerical simulation model of frozen soil, the following equation is used as the control equation for heat conduction simulation: Where T is temperature, t is time, u is velocity, and Q is the heat source.

[0019] Furthermore, the BP model is used to i ,θu) to perform inversion analysis and establish the following mapping relationship:

[0020]

[0021] Among them, [θ i ,θ u ],TRC=[T f0 ,T f1 ,T f2 …T f60 ],S=[ρ s ,λ s ,C s ].T fi is the temperature response data of frozen soil within 60s, with a measurement frequency of 1s; θ i represents the ice content of frozen soil, θ u represents water content; ρ s represents the density of frozen soil particles; s represents the thermal conductivity of frozen soil particles; Cs represents the specific heat capacity of frozen soil particles.

[0022] In a second aspect, the present invention provides a system for simultaneously measuring the pore ice content and unfrozen water content of frozen soil, comprising:

[0023] An acquisition module is used to obtain the temperature response curve of frozen soil under the pulse heat source and specific frozen soil parameters; the specific frozen soil parameters include frozen soil particle density, frozen soil particle thermal conductivity, frozen soil particle specific heat capacity and frozen soil particle specific gravity;

[0024] A processing module is used to process the acquired temperature response curve and specific frozen soil parameters using a pre-trained inversion analysis model to obtain the frozen soil ice content and unfrozen water content; wherein, the inversion analysis model is trained using a training set, the training set includes multiple groups of data, each group of data includes a temperature response curve, and each temperature response curve is labeled with the frozen soil ice content, the frozen soil unfrozen water content, the frozen soil particle density, the frozen soil particle thermal conductivity, the frozen soil particle specific heat capacity and the frozen soil particle specific gravity; a machine learning algorithm is used, with the temperature response curve and different frozen soil parameters as input, and the frozen soil ice content and the frozen soil unfrozen water content as output, to train the training set and extract features, compare the prediction effects when different frozen soil parameters are used as input, and obtain the optimal inversion analysis model.

[0025] In a third aspect, the present invention provides a non-transitory computer-readable storage medium, which is used to store computer instructions. When the computer instructions are executed by a processor, the method for simultaneously measuring the pore ice content and unfrozen water content of frozen soil as described in the first aspect is implemented.

[0026] In a fourth aspect, the present invention provides a computer device comprising a memory and a processor, wherein the processor and the memory communicate with each other, the memory stores program instructions executable by the processor, and the processor calls the program instructions to execute the method for simultaneously measuring the pore ice content and unfrozen water content of frozen soil as described in the first aspect.

[0027] In a fifth aspect, the present invention provides an electronic device comprising: a processor, a memory and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory so that the electronic device executes instructions for implementing the method for simultaneously measuring the pore ice content and unfrozen water content of frozen soil as described in the first aspect.

[0028] Terminology explanation:

[0029] 1. Ice content and unfrozen water content of frozen soil (θ i and θ u): Due to the suction in the pores of frozen soil, ice and water coexist. The phase change of ice and water in frozen soil is one of the important factors that lead to its unstable properties. Ice content and unfrozen water content are basic parameters to describe the characteristics of frozen soil. They directly affect key processes such as thermal properties, mechanical behavior and water migration of frozen soil. Rapid and accurate measurement of ice content and unfrozen water content in frozen soil is of great significance to engineering construction in cold regions.

[0030] 2. Temperature response curve: In geotechnical materials, if constant time (t0) and constant power (P0) heat is input, the geotechnical materials at a constant distance (r0) from the heat source will gradually heat up. Measure the temperature of this point at different times (t i ,T i ), where the time of starting heating is t=0. Further according to ΔT i =T i -T0, (where T0 is the initial temperature), calculate the temperature change at different times (t i ,ΔT i ). i As the horizontal axis, ΔT i The curve described by the ordinate is the thermal response curve. As the geothermal parameters of the geotechnical materials change, the curve will change synchronously.

[0031] 3. Inverse analysis: Starting from the measured structural response (such as displacement, strain, temperature field, etc.), the input conditions or model parameters are inferred. In the present invention, the ice content and unfrozen water content are inversely calculated by measuring the temperature response curve.

[0032] 4. Monte Carlo numerical simulation: It is a method for performing large-scale, random numerical simulations. This method can be used to perform a large number of numerical simulations and generate a large number of data sets required for machine learning.

[0033] 5. Backpropagation Algorithm: Backpropagation Algorithm, referred to as BP algorithm, is a machine learning algorithm. BP algorithm calculates the gradient of the loss function with respect to the network parameters and uses these gradients to reversely update the weights and biases of the network to minimize the prediction error of the loss function.

[0034] The beneficial effects of the present invention are as follows: the ice content and the unfrozen water content of frozen soil can be measured simultaneously within one minute, which greatly improves the measurement efficiency and avoids the assumptions of the empirical model.

[0035] Additional advantages of the present invention will be more clearly presented in the following description or learned through the practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0037] Figure 1 A schematic diagram of the numerical simulation model described in the embodiment of the present invention is established. Figure 1 (a) is a schematic diagram under laboratory conditions; Figure 1 (b) is a schematic diagram of the numerical simulation model.

[0038] Figure 2 This is a schematic diagram of Monte Carlo sampling according to an embodiment of the present invention.

[0039] Figure 3 This is a structural diagram of the BP model described in an embodiment of the present invention. DETAILED DESCRIPTION

[0040] The embodiments of the present invention are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements with the same or similar functions. The embodiments described below by the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be interpreted as limiting the present invention.

[0041] It should be understood by those skilled in the art that unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by those skilled in the art to which this invention belongs.

[0042] It should also be understood that terms, such as those defined in commonly used dictionaries, should be understood to have a meaning consistent with that in the context of the prior art and will not be interpreted in an idealized or overly formal sense unless as defined herein.

[0043] Those skilled in the art will appreciate that, unless otherwise stated, the singular forms "a", "an", "said" and "the" used herein may also include plural forms. It should be further understood that the term "comprising" used in the specification of the present invention refers to the presence of the features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements and / or groups thereof.

[0044] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. Different embodiments or examples described in this specification and features of different embodiments or examples may be combined and combined by those skilled in the art without contradiction.

[0045] To facilitate understanding of the present invention, the present invention is further explained below with reference to specific embodiments in conjunction with the accompanying drawings, and the specific embodiments do not constitute a limitation on the embodiments of the present invention.

[0046] Those skilled in the art should understand that the drawings are merely schematic diagrams of embodiments, and the components in the drawings are not necessarily necessary for implementing the present invention.

[0047] Example 1

[0048] In this embodiment 1, a system for simultaneously measuring the ice content and unfrozen water content in frozen soil pores is first provided, including: an acquisition module for acquiring a temperature response curve and specific frozen soil parameters of frozen soil under a pulse heat source; the specific frozen soil parameters include frozen soil particle density, frozen soil particle thermal conductivity, frozen soil particle specific heat capacity and frozen soil particle specific gravity; a processing module for processing the acquired temperature response curve and specific frozen soil parameters using a pre-trained inversion analysis model to obtain the frozen soil ice content and unfrozen water content; wherein the inversion model is trained using a training set. The analysis model is provided, wherein the training set includes multiple groups of data, each group of data includes a temperature response curve, and each temperature response curve is labeled with the ice content of frozen soil, the unfrozen water content of frozen soil, the density of frozen soil particles, the thermal conductivity of frozen soil particles, the specific heat capacity of frozen soil particles and the specific gravity of frozen soil particles; a machine learning algorithm is used, with the temperature response curve and different frozen soil parameters as input, and the ice content of frozen soil and the unfrozen water content of frozen soil as output, to train the training set and extract features, compare the prediction effects when different frozen soil parameters are used as input, and obtain the optimal inversion analysis model.

[0049] In this embodiment, the above-mentioned system is used to implement a method for simultaneously measuring the pore ice content and unfrozen water content of frozen soil, including: using an acquisition module to obtain the temperature response curve and specific frozen soil parameters of frozen soil under a pulse heat source; the specific frozen soil parameters include the density of frozen soil particles, the thermal conductivity of frozen soil particles, the specific heat capacity of frozen soil particles and the specific gravity of frozen soil particles; using a processing module to process the acquired temperature response curve and specific frozen soil parameters using a pre-trained inversion analysis model to obtain the frozen soil ice content and unfrozen water content; wherein, the inversion analysis model is trained using a training set. The invention discloses an inversion analysis model, wherein the training set comprises multiple groups of data, each group of data comprises a temperature response curve, and each temperature response curve is labeled with the ice content of frozen soil, the unfrozen water content of frozen soil, the density of frozen soil particles, the thermal conductivity of frozen soil particles, the specific heat capacity of frozen soil particles and the specific gravity of frozen soil particles; a machine learning algorithm is adopted, with the temperature response curve and different frozen soil parameters as inputs, and the ice content of frozen soil and the unfrozen water content of frozen soil as outputs, to train the training set and extract features, compare the prediction effects when different frozen soil parameters are used as inputs, and obtain the optimal inversion analysis model.

[0050] Among them, a group of parameters are set including the ice content of frozen soil, the unfrozen water content of frozen soil, the density of frozen soil particles, the thermal conductivity of frozen soil particles, the specific heat capacity of frozen soil particles and the specific gravity of frozen soil particles, and a group of thermal conductivity and specific heat capacity of frozen soil are calculated according to the set parameters; based on the transient heat conduction numerical simulation model of frozen soil, the thermal conductivity and specific heat capacity of frozen soil under heat pulse are set to obtain the corresponding temperature response curve; Monte Carlo sampling is performed on the set frozen soil ice content, unfrozen water content of frozen soil, the density of frozen soil particles, the thermal conductivity of frozen soil particles, the specific heat capacity of frozen soil particles and the specific gravity of frozen soil particles, and multiple groups of thermal conductivity and specific heat capacity of frozen soil are calculated, and the obtained multiple groups of thermal conductivity and specific heat capacity of frozen soil are input into the transient heat conduction numerical simulation model to obtain the corresponding multiple groups of temperature response curves.

[0051] The following four equations are used to calculate the thermal conductivity parameters and density of frozen soil:

[0052]

[0053] ρ f =ρ s +0.92θ i +θ w

[0054]

[0055] Among them, ρ f is the density of frozen soil, C f is the specific heat capacity of frozen soil, λ f is the thermal conductivity of frozen soil; ρ s is the density of soil particles; θ is the volume fraction; φ is the mass fraction, and Gs is the specific gravity of soil particles.

[0056] In the transient heat conduction numerical simulation model of frozen soil, the following equation is used as the control equation for heat conduction simulation: Where T is temperature, t is time, u is velocity, and Q is the heat source.

[0057] Using BP model to (θ i ,θu) to perform inversion analysis and establish the following mapping relationship:

[0058]

[0059] Among them, [θ i ,θ u ],TRC=[T f0 ,T f1 ,T f2 …T f60 ],s=[ρ s ,λ s ,C s ].T fi is the temperature response data of frozen soil within 60s, with a measurement frequency of 1s; θ i represents the ice content of frozen soil, θ u represents water content; ρ s represents the density of frozen soil particles; s represents the thermal conductivity of frozen soil particles; C s represents the specific heat capacity of frozen soil particles.

[0060] Example 2

[0061] In this embodiment 2, a method for quickly and simultaneously measuring the pore ice content and unfrozen water content of frozen soil is provided. First, the TRC and specific frozen soil parameters of the frozen soil under a pulse heat source are measured and obtained, and then they are input into the inversion analysis model to quickly predict the ice content θi and the unfrozen water content θu.

[0062] First, select the calculation model of thermal conductivity and specific heat capacity of frozen soil. Set a set of (θ i ,θ u , s , s , C s , G s ) parameters and input them into the calculation model to obtain a set of frozen soil (λ f and C f ).

[0063] Secondly, a transient heat conduction numerical simulation model of frozen soil was established based on laboratory conditions. In this model, a heating probe was inserted into the frozen soil to provide a short heat pulse. By setting the geothermal parameters (λf and Cf) of the frozen soil, the TRC under the parameter conditions can be obtained.

[0064] Then, Monte Carlo sampling is performed on (θi, θu, ρs, λs, Cs, Gs), tens of thousands of sets of (λf and Cf) are calculated, and input into the numerical model. Each time (λf and Cf) is input, a set of TRCs can be obtained. Finally, a database consisting of tens of thousands of TRCs is constructed, each of which is labeled with (θi, θu, ρs, λs, Cs, Gs).

[0065] Finally, a machine learning algorithm was used to train the data set and extract features, with TRC and different frozen soil parameters (ρs, λs, Cs, Gs) as input and θu and θi as output. The prediction effects when different frozen soil parameters were used as input were compared to obtain the optimal inversion analysis model.

[0066] In this embodiment, constructing a numerical simulation model includes the following steps:

[0067] The following four equations are used to calculate the thermal conductivity parameters and density of frozen soil:

[0068]

[0069] ρ f =ρ s +0.92θ i +θ w (3)

[0070]

[0071] Among them, ρ f is the density of frozen soil; C f is the specific heat capacity of frozen soil; f is the thermal conductivity of frozen soil; ρ s is the density of soil particles; θ is the volume fraction; φ is the mass fraction, and Gs is the specific gravity of the particles.

[0072] In the above four equations, six independent variables or parameters in frozen soil need to be determined: θ i (Ice content), θ u (unfrozen water content), ρ s (soil particle density), μ s (Parameters related to soil texture, including λ s (particle thermal conductivity), C s (Specific heat capacity of frozen soil particles), G s (Specific gravity of particles). Once the above six variables or parameters are determined, the thermal parameters of frozen soil can be determined to simulate the transient heat conduction process of frozen soil.

[0073] In this embodiment, the laboratory conditions of the numerical simulation model are as follows: an acrylic cylinder is used as a soil container, and the container dimensions are inner diameter Φ = 10 cm, thickness δ = 5 mm, and height h = 10 cm. The interior of the container is frozen soil, and a temperature sensor is inserted into it. Figure 1 (a) shown.

[0074] The sensor has three probes, with a heating needle on the left, which provides a short pulse heat source. The heating needle is composed of polypropylene filler and steel shell. There is a pt100 temperature measuring needle (called the temperature measuring needle) on the right side 1 cm away from the heating needle, and the material is platinum. The left side of the heating needle is close to another temperature measuring needle, which is used to monitor the temperature of the heating needle. The radius of the three needles is 2 mm. The thermophysical parameters used for each material in the numerical model are shown in Table 1. Large pulse heat sources can cause drastic phase changes in frozen soil. Appropriate pulse heat source time and power are crucial to accurately simulate the heat conduction of frozen soil. The pulse heat source power used in this study is Q = 15W / m, and the heating time is t = 10s.

[0075] Table 1 Thermophysical parameters of each material in the numerical model

[0076]

[0077] The following equation is used as the governing equation for heat conduction simulation:

[0078]

[0079] Among them, T is temperature; t is time; u is speed; Q is heat source.

[0080] The area between the three probes is the key area for simulating transient heat conduction. The cylindrical area with a radius of 15 mm in the middle of the frozen soil was refined, and the grid unit size was 0.022 mm to 2.2 mm. The outer area of ​​the cylinder was coarsened, and the grid unit size was 1.1 mm to 8.8 mm. Figure 1 The advantage of this treatment is that it reduces the amount of calculation for a single simulation while ensuring the accuracy of TRC simulation.

[0081] Using the above numerical simulation model, after inputting the six parameters, the transient heat conduction in frozen soil can be simulated to obtain the TRC curve of the heating needle. After each simulation, the TRC time range derived from the model is 0 to 60 seconds, and the sampling frequency is 1Hz.

[0082] Monte Carlo sampling of six parameters includes:

[0083] Six parameters (θi, θu, ρs, λs, Cs, Gs) need to be input to obtain a simulated TRC in the numerical model. Among these six parameters, (λs, Cs, Gs, ρs) are parameters related to soil texture; (θi, θu, ρs) are parameters related to the volume fraction of the three phases of frozen soil. Monte Carlo sampling is performed on the six parameters (λs, Cs, Gs, θi, θu, ρs).

[0084] Due to the limitation of soil porosity, the following boundary conditions are set:

[0085] ρ s / G s +θ i +θ w ≤1 (6)

[0086] Considering the numerical range of frozen soil parameters in nature, the range of λs is 1-5W / m·K, the range of Cs is 600-1500J / kg·K, and the range of Gs is 2.6-2.8. The sampling times of these three parameters are 5 times. The range of ρs is 0.6-1.9g / cm 3 , the value range of θu is 0-50%, and the sampling times of both are more than 10 times. The sampling ranges of the six parameters are shown in Table 2. Finally, (θ i ,θu,ρ s ,λ s ,C s ,G s ) is as follows: Figure 2 shown.

[0087] Table 2 Ranges of dry density, ice content, and unfrozen water content of three soils

[0088]

[0089]

[0090] Therefore, by Monte Carlo sampling of the six parameters, tens of thousands of parameter combinations with different values ​​of (θi, θu, ρs, λs, Cs, Gs) were established. Based on the tens of thousands of parameter combinations (θi, θu, ρs, λs, Cs, Gs) in Section 3.1, tens of thousands of (λf, Cf, ρf) can be obtained by calculation through equations (1) to (4) and input into the numerical model. After tens of thousands of simulations, tens of thousands of TRCs with (θi, θu, ρs, λs, Cs, Gs) and (λf, Cf, ρf) as labels can be obtained.

[0091] It should be noted that the TRC derived from the numerical model only includes the temperature data of the first 60 seconds of heating needle, i.e., T0, T1, T2…T 60(According to the frequency of 1 second, a total of 61 data points).

[0092] In this embodiment, constructing a machine learning model includes: based on a numerical model, establishing a i ,θ u ,ρ s ,λ s ,C s ,G s ) six parameters as labels of the TRC database. Then, the BP model is used to i ,θu) to perform inversion analysis and establish the following mapping relationship:

[0093]

[0094] Among them, [θ i ,θ u ],TRC=[T f0 ,T f1 ,T f2 …T f60 ],s=[ρ s ,λ s ,C s ].T fi It is the temperature response data of frozen soil within 60s, and the measurement frequency is 1s.

[0095] This embodiment selects the commonly used and classic machine learning algorithm - Back Propagation (BP), and the construction steps are as follows: Figure 3 The constructed model contains 2 hidden layers, each with 100 neurons. The dimension of the input feature ranges from 61 to 127, and the output dimension is 1 or 2. MLPRegressor is used as the feedforward neural network and is wrapped with MultiOutputRegressor. The early stopping method is used to stop training when the performance on the validation set no longer improves to prevent overfitting. Other parameters when constructing the BP model are shown in Table 3.

[0096] Table 3 Other parameters of BP model

[0097]

[0098]

[0099] In this embodiment, in a specific application, a pt100 temperature sensor is used to measure TRC in real time in the field, and the pulse heat source parameters are (Q=15W / m, t=10s). A DC linear power supply and a time control switch are used to ensure the duration and power of the heat pulse. A data acquisition device (DT85) is used to collect data from the temperature sensor. In order to reduce the influence of randomness, three consecutive measurements should be performed, with an interval of more than 12 hours. The average of the three measurement results is taken, and the curve is denoised using a median filter algorithm. The denoised curve is regarded as the final measured TRC.

[0100] Samples of frozen soil in the field are taken and transported to the laboratory to measure other parameters. For dry density, wax sealing method or drying method should be used for measurement.

[0101] The thermal conductivity of soil particles is determined by the following method. The transported soil is made into a sample with a dry density of ρ0 and dried. At this time, the soil body consists of two phases: soil and gas. The thermophysical parameters of the dry soil sample are tested, and the thermophysical parameters of the soil particles can be obtained by inversion:

[0102]

[0103] C s =C0 (8)

[0104] Among them, s, a, and 0 represent the thermophysical parameters of soil particles, air, and soil respectively. Gs is the specific gravity of soil particles, and the recommended value is 2.7; ρ0 is the dry density.

[0105] By inputting the (ρs, λs, Cs) measured in the laboratory and the TRC within one minute measured in the field into the machine learning model in real time, (θi, θu) at different times can be obtained in real time.

[0106] Example 3

[0107] This embodiment 3 provides a non-transitory computer-readable storage medium, wherein the non-transitory computer-readable storage medium is used to store computer instructions. When the computer instructions are executed by a processor, the method for simultaneously measuring the pore ice content and the unfrozen water content of frozen soil as described above is implemented. The method includes:

[0108] Obtaining a temperature response curve of frozen soil under a pulse heat source and specific frozen soil parameters; the specific frozen soil parameters include frozen soil particle density, frozen soil particle thermal conductivity, frozen soil particle specific heat capacity and frozen soil particle specific gravity;

[0109] The obtained temperature response curve and specific frozen soil parameters are processed by a pre-trained inversion analysis model to obtain the frozen soil ice content and unfrozen water content; wherein, the inversion analysis model is trained by a training set, and the training set includes multiple groups of data, each group of data includes a temperature response curve, and each temperature response curve is labeled with the frozen soil ice content, the frozen soil unfrozen water content, the frozen soil particle density, the thermal conductivity of the frozen soil particles, the specific heat capacity of the frozen soil particles and the specific gravity of the frozen soil particles; a machine learning algorithm is used, with the temperature response curve and different frozen soil parameters as input, and the frozen soil ice content and the unfrozen water content of the frozen soil as output, to train the training set and extract features, compare the prediction effects when different frozen soil parameters are used as input, and obtain the optimal inversion analysis model.

[0110] Example 4

[0111] This embodiment 4 provides a computer device, including a memory and a processor, the processor and the memory communicate with each other, the memory stores program instructions that can be executed by the processor, and the processor calls the program instructions to execute the method for simultaneously measuring the pore ice content and unfrozen water content of frozen soil as described above, the method comprising:

[0112] Obtaining a temperature response curve of frozen soil under a pulse heat source and specific frozen soil parameters; the specific frozen soil parameters include frozen soil particle density, frozen soil particle thermal conductivity, frozen soil particle specific heat capacity and frozen soil particle specific gravity;

[0113] The obtained temperature response curve and specific frozen soil parameters are processed by a pre-trained inversion analysis model to obtain the frozen soil ice content and unfrozen water content; wherein, the inversion analysis model is trained by a training set, and the training set includes multiple groups of data, each group of data includes a temperature response curve, and each temperature response curve is labeled with the frozen soil ice content, the frozen soil unfrozen water content, the frozen soil particle density, the thermal conductivity of the frozen soil particles, the specific heat capacity of the frozen soil particles and the specific gravity of the frozen soil particles; a machine learning algorithm is used, with the temperature response curve and different frozen soil parameters as input, and the frozen soil ice content and the unfrozen water content of the frozen soil as output, to train the training set and extract features, compare the prediction effects when different frozen soil parameters are used as input, and obtain the optimal inversion analysis model.

[0114] Example 5

[0115] This embodiment 5 provides an electronic device, including: a processor, a memory, and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory to enable the electronic device to execute instructions for implementing the method for simultaneously measuring the pore ice content and unfrozen water content of frozen soil as described above, the method comprising:

[0116] Obtaining a temperature response curve of frozen soil under a pulse heat source and specific frozen soil parameters; the specific frozen soil parameters include frozen soil particle density, frozen soil particle thermal conductivity, frozen soil particle specific heat capacity and frozen soil particle specific gravity;

[0117] The obtained temperature response curve and specific frozen soil parameters are processed by a pre-trained inversion analysis model to obtain the frozen soil ice content and unfrozen water content; wherein, the inversion analysis model is trained by a training set, and the training set includes multiple groups of data, each group of data includes a temperature response curve, and each temperature response curve is labeled with the frozen soil ice content, the frozen soil unfrozen water content, the frozen soil particle density, the thermal conductivity of the frozen soil particles, the specific heat capacity of the frozen soil particles and the specific gravity of the frozen soil particles; a machine learning algorithm is used, with the temperature response curve and different frozen soil parameters as input, and the frozen soil ice content and the unfrozen water content of the frozen soil as output, to train the training set and extract features, compare the prediction effects when different frozen soil parameters are used as input, and obtain the optimal inversion analysis model.

[0118] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0119] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0120] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0121] These computer program instructions can also be loaded onto a computer or other programmable data processing device, and a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, so that the instructions executed on the computer or other programmable device provide for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0122] Although the above describes the specific implementation mode of the present invention in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art without creative work on the basis of the technical solution disclosed in the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for simultaneously measuring the pore ice content and unfrozen water content of frozen soil, characterized in that: include: Obtaining a temperature response curve of frozen soil under a pulse heat source and specific frozen soil parameters; the specific frozen soil parameters include frozen soil particle density, frozen soil particle thermal conductivity, frozen soil particle specific heat capacity and frozen soil particle specific gravity; The obtained temperature response curve and specific frozen soil parameters are processed by a pre-trained inversion analysis model to obtain the frozen soil ice content and unfrozen water content; wherein, the inversion analysis model is trained by a training set, and the training set includes multiple groups of data, each group of data includes a temperature response curve, and each temperature response curve is labeled with the frozen soil ice content, the frozen soil unfrozen water content, the frozen soil particle density, the thermal conductivity of the frozen soil particles, the specific heat capacity of the frozen soil particles and the specific gravity of the frozen soil particles; a machine learning algorithm is used, with the temperature response curve and different frozen soil parameters as input, and the frozen soil ice content and the unfrozen water content of the frozen soil as output, to train the training set and extract features, compare the prediction effects when different frozen soil parameters are used as input, and obtain the optimal inversion analysis model.

2. The method for simultaneously measuring the pore ice content and unfrozen water content of frozen soil according to claim 1, characterized in that: A set of parameters including ice content in frozen soil, unfrozen water content in frozen soil, density of frozen soil particles, thermal conductivity of frozen soil particles, specific heat capacity of frozen soil particles and specific gravity of frozen soil particles are set, and a set of thermal conductivity and specific heat capacity of frozen soil are calculated according to the set parameters; based on the transient heat conduction numerical simulation model of frozen soil, the thermal conductivity and specific heat capacity of frozen soil under heat pulse are set to obtain the corresponding temperature response curve.

3. The method for simultaneously measuring the pore ice content and unfrozen water content of frozen soil according to claim 2, characterized in that: Monte Carlo sampling was performed on the set frozen soil ice content, frozen soil unfrozen water content, frozen soil particle density, frozen soil particle thermal conductivity, frozen soil particle specific heat capacity and frozen soil particle specific gravity, and multiple sets of frozen soil thermal conductivity and frozen soil specific heat capacity were calculated. The obtained multiple sets of frozen soil thermal conductivity and frozen soil specific heat capacity were input into the transient heat conduction numerical simulation model to obtain the corresponding multiple sets of temperature response curves.

4. The method for simultaneously measuring the pore ice content and unfrozen water content of frozen soil according to claim 3, characterized in that: The following four equations are used to calculate the thermal conductivity parameters and density of frozen soil: r f =ρ s +0.92θ i +θ w Among them, ρ f is the density of frozen soil, C f is the specific heat capacity of frozen soil, λ f is the thermal conductivity of frozen soil; ρ s is the density of soil particles; θ is the volume fraction; φ is the mass fraction, and Gs is the specific gravity of soil particles.

5. The method for simultaneously measuring the pore ice content and unfrozen water content of frozen soil according to claim 3, characterized in that: In the transient heat conduction numerical simulation model of frozen soil, the following equation is used as the control equation for heat conduction simulation: Where T is temperature, t is time, u is velocity, and Q is the heat source.

6. The method for simultaneously measuring the pore ice content and unfrozen water content of frozen soil according to claim 1, characterized in that: Using BP model to (θ i , θu) to perform inversion analysis and establish the following mapping relationship: Among them, [θ i ,θ u ],TRC=[T f0 , T f1 , T f2 …T f60 ],s=[ρ s ,λ s , C s ].T fi is the temperature response data of frozen soil within 60s, with a measurement frequency of 1s; θ i represents the ice content of frozen soil, θ u represents water content; ρ s represents the density of frozen soil particles; s represents the thermal conductivity of frozen soil particles; C s represents the specific heat capacity of frozen soil particles.

7. A system for simultaneously measuring the ice content and unfrozen water content in frozen soil pores, characterized in that: include: An acquisition module is used to obtain the temperature response curve of frozen soil under the pulse heat source and specific frozen soil parameters; the specific frozen soil parameters include frozen soil particle density, frozen soil particle thermal conductivity, frozen soil particle specific heat capacity and frozen soil particle specific gravity; A processing module is used to process the acquired temperature response curve and specific frozen soil parameters using a pre-trained inversion analysis model to obtain the frozen soil ice content and unfrozen water content; wherein, the inversion analysis model is trained using a training set, the training set includes multiple groups of data, each group of data includes a temperature response curve, and each temperature response curve is labeled with the frozen soil ice content, the frozen soil unfrozen water content, the frozen soil particle density, the frozen soil particle thermal conductivity, the frozen soil particle specific heat capacity and the frozen soil particle specific gravity; a machine learning algorithm is used, with the temperature response curve and different frozen soil parameters as input, and the frozen soil ice content and the frozen soil unfrozen water content as output, to train the training set and extract features, compare the prediction effects when different frozen soil parameters are used as input, and obtain the optimal inversion analysis model.

8. A non-transitory computer-readable storage medium, characterized in that: The non-transitory computer-readable storage medium is used to store computer instructions. When the computer instructions are executed by the processor, the method for simultaneously measuring the pore ice content and unfrozen water content of frozen soil as described in any one of claims 1 to 6 is implemented.

9. A computer device, characterized in that: It includes a memory and a processor, the processor and the memory communicate with each other, the memory stores program instructions that can be executed by the processor, and the processor calls the program instructions to execute the method for simultaneously measuring the pore ice content and unfrozen water content of frozen soil as described in any one of claims 1-6.

10. An electronic device, characterized in that: include: A processor, a memory and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory so that the electronic device executes instructions for implementing the method for simultaneously measuring the pore ice content and unfrozen water content of frozen soil as described in any one of claims 1 to 6.

Images