Method for establishing relationship model between soil type and soil-forming environmental factors

Abstract

The present invention relates to the technical field of electrical digital data processing, and relates to a method for establishing a relationship model between soil types and soil-forming environmental factors, including: demarcating a plurality of research areas in a simulation system, and classifying the plurality of research areas into a conventional soil environment and a changing soil environment; simulating the changing soil environment in a specific form and simulating the conventional soil environment in a standardized manner until the simulation duration reaches a threshold and then stopping; detecting the attribute data of the two soil types, and extracting the data bands of the conventional soil environment and the changing soil environment; respectively combining the data bands of the conventional soil environment and the changing soil environment, and analyzing the conventional soil environment and the changing soil environment in combination with the attribute data to obtain the difference between the changing soil environment and the conventional soil environment; screening out the data with a difference lower than the set index, and importing it into a preset soil type relationship model for iterative update to complete the training of the soil type relationship model.

Description

Technical Field

[0001] The present invention relates to the technical field of digital data processing, and more particularly, to a method for establishing a relationship model between soil types and soil-forming environmental factors. Background Art

[0002] Soil refers to a layer of loose material on the earth's surface, composed of various granular minerals, organic matter, moisture, air, microorganisms, etc. Soil types are closely related to the environmental factors where the soil is located. At present, due to the relatively dispersed spatial distribution of soil, there are corresponding growing vegetation for soil types in each space. As the vegetation grows, they will also change some properties of the surrounding soil types. In existing research, mainly by changing environmental factors to study the impact on pure soil types, ignoring that different soil types often have different vegetation. Only studying the impact of environmental factors on pure soil types has poor reliability, and it is also impossible to judge how the soil types with vegetation will be affected when environmental factors change. Therefore, it is particularly important to study the relationship between the soil environment with vegetation and environmental factors. Based on this, in view of the above problems, we propose a method for establishing a relationship model between soil types and soil-forming environmental factors. Summary of the Invention

[0003] The purpose of the present invention is to provide a method for establishing a relationship model between soil types and soil-forming environmental factors. By setting two types of soil environments, namely conventional soil environment and changing soil environment, it respectively studies the impact of changing environmental factors on the soil environment with vegetation and the state of the normal soil environment with vegetation. Based on the combined spectral data and the attribute data detected before and after the research process, it studies the difference between the two soil types, and finally iteratively updates and verifies the soil type relationship model, which not only improves the reliability of establishing the soil type relationship model, but also provides an important scientific basis for the relationship between the soil environment with vegetation and environmental factors.

[0004] The embodiments of the present invention are implemented by the following technical solutions:

[0005] A method for establishing a relationship model between soil types and soil-forming environmental factors, the steps of the method include:

[0006] Designate multiple research areas within the simulation system, and classify the multiple research areas into a conventional soil environment and a changing soil environment;

[0007] Simulate the changing soil environment in a specific form and simulate the conventional soil environment in a standardized manner until the simulation duration reaches the threshold and then stop;

[0008] Detect the attribute data of the two soil types, and extract the data bands of the conventional soil environment and the changing soil environment;

[0009] Combine the data bands of the conventional soil environment and the changing soil environment respectively, and analyze the conventional soil environment and the changing soil environment in combination with the attribute data to obtain the difference between the changing soil environment and the conventional soil environment;

[0010] Filter out the data with differences lower than the set index, and import them into the preset soil type relationship model for iterative update to complete the training of the soil type relationship model.

[0011] Optionally, both the conventional soil environment and the changing soil environment are soil environments with vegetation. Among them, the conventional soil environment is specifically the normal soil environment based on the preset vegetation planting standard area, and the changing soil environment is specifically the soil environment obtained by simulating the change of the conventional soil environment through independent variables.

[0012] Optionally, simulate the changing soil environment in a specific form, specifically: simulate the changing soil environment by setting at least one independent variable.

[0013] Optionally, simulate by setting at least one independent variable, where the independent variables include: land use type, parent material lithology, geomorphic type, elevation, slope, aspect, slope position, plane curvature, profile curvature, rainfall, terrain wetness index, remote sensing spectrum, rotation method.

[0014] Optionally, before combining the data bands of the conventional soil environment and the changing soil environment, a preprocessing operation is also included, and the steps are as follows:

[0015] Remove noise from the data bands of the conventional soil environment and the changing soil environment through the wavelet transform algorithm;

[0016] Perform dimensionality reduction processing on the denoised data bands through the principal component analysis algorithm;

[0017] Obtain the spectral data of the conventional soil environment and the spectral data of the changing soil environment, and obtain the vegetation indices of the two soil environments by converting the spectral data.

[0018] Optionally, based on the vegetation indices of the two soil environments, take the vegetation index of the conventional soil environment as the reference group, and at the same time take the vegetation index of the changing soil environment as the calculation group. Calculate the reference group and the calculation group through the index fitting function to obtain the fitting coefficient between the independent variable and the vegetation index of the conventional soil environment, and verify whether the coincidence degree of the fitting coefficient reaches the expected value. If so, output the fitting coefficient to construct the preset soil type relationship model; if not, return to obtain the fitting coefficient again until the coincidence degree of the fitting coefficient reaches the expected value.

[0019] Optionally, after the soil type relationship model is trained, the accuracy of the soil type relationship model is further verified by a preset test set.

[0020] The technical solution of the embodiment of the present invention has at least the following advantages and beneficial effects:

[0021] In the embodiment of the present invention, by setting two types of conventional soil environments and changing soil environments, the effects of changing environmental factors on the soil environment with vegetation and the state of the normal soil environment with vegetation are respectively studied. Based on the combined spectral data and the attribute data detected before and after the research process, the difference between the two soil types is studied. Finally, the soil type relationship model is iteratively updated and verified, which not only improves the reliability of the establishment of the soil type relationship model, but also provides an important scientific basis for the relationship between the soil environment with vegetation and environmental factors. Description of the Drawings

[0022] Figure 1 It is a schematic diagram of the overall process of the method for establishing a relationship model between soil types and soil-forming environmental factors provided by the embodiment of the present invention. Detailed Embodiments

[0023] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Components of the embodiments of the present invention described and illustrated herein can be arranged and designed in various different configurations.

[0024] Refer to Figure 1 as shown Figure 1 It is a schematic diagram of the overall process of the method for establishing a relationship model between soil types and soil-forming environmental factors provided by the embodiment of the present invention.

[0025] In some embodiments, a method for establishing a relationship model between soil types and soil-forming environmental factors, the steps of the method include:

[0026] Designate multiple research areas in the simulation system, and classify the multiple research areas into a conventional soil environment and a changing soil environment;

[0027] Simulate the changing soil environment in a specific form and simulate the conventional soil environment in a standardized manner until the simulation duration reaches the threshold and then stop;

[0028] Detect the attribute data of the two soil types, and extract the data bands of the conventional soil environment and the changing soil environment;

[0029] Combine the data bands of the conventional soil environment and the changing soil environment respectively, and analyze the conventional soil environment and the changing soil environment in combination with the attribute data to obtain the difference between the changing soil environment and the conventional soil environment;

[0030] Screen out the data with differences lower than the set index, and import them into the preset soil type relationship model for iterative update to complete the training of the soil type relationship model.

[0031] In the above implementation process, this embodiment selects the sample data combined with existing research and uses a simulation system to study the relationship between soil types and environmental factors. Taking the simulation system of this embodiment as an example, multiple research areas are divided in the simulation system. Among them, a set number of research areas are used as the cultivation areas for the (soil environment with vegetation) conventional soil environment, a set number of research areas are used as the cultivation areas for the soil environment without vegetation, and a set of multiple research areas are used as the areas of the soil environment with vegetation to simulate the changes in environmental factors. Among them, the environmental factors are in a specific form, that is, the independent variables described later. In this embodiment, the independent variables are specifically land use type, parent material lithology, geomorphic type, elevation, slope, aspect, slope position, plane curvature, profile curvature, rainfall, terrain wetness index, remote sensing spectrum, rotation method, etc., and at least one of the corresponding independent variables exists in multiple research areas. This embodiment sets the threshold of the simulation duration. When the threshold is reached, the soil type simulation of multiple research areas is stopped in the simulation system. Among them, by comparing the data of the soil environment without vegetation and the conventional soil environment, it can be clearly seen that vegetation has a certain impact on the soil environment, and the attribute data therein is different from the soil type environment without vegetation. The attribute data is the component data such as soil pH, organic matter content, and soil spectrum. After studying the certain impact of vegetation on the soil environment, this embodiment extracts the data bands of the conventional soil environment and the soil environment with vegetation that has undergone environmental factor changes, and then combines them into spectral data, and analyzes and compares them in combination with the conventional soil environment attribute data and the soil environment attribute data after environmental factor transformation to obtain the difference between the transformed soil environment and the conventional soil environment. Here, the difference is affected by environmental factors due to vegetation factors. In order to ensure the accuracy and reliability of the research, we screen out the differences with lower impacts, that is, the lower differences, and then iterate the screened differences on the preset soil type relationship model to complete the training of the soil type relationship model.

[0032] In the above implementation process, for the simulation in this embodiment, soil virtual simulation technology or a corresponding cultivation chamber can be used for cultivation experiments, and the change parameters in both simulation experiments are the independent variables described in this embodiment.

[0033] In the above implementation process, for the standardization in this embodiment, it specifically refers to the normal cultivation of the conventional soil environment, that is, the cultivation is carried out in an environment without the interference of independent variable changes.

[0034] In the above implementation process, for the combination in this embodiment, specifically, the purpose of combining the data bands of the conventional soil environment and the data bands of the changing soil environment is to form the spectral data of the conventional soil environment and the spectral data of the changing data. Specifically, on the premise that any one-dimensional property except the band channel needs to be kept consistent, the data bands are read out and superimposed in sequence through the Toolbox / Layer tacking tool. Finally, after being exported, the spectral data of the conventional soil environment and the changing soil environment are formed.

[0035] In the above implementation process, for the difference in this embodiment, specifically, under the interference of different independent variables, the spectral data of the changing soil environment obtained is combined with the attribute data detected in the above steps, and the spectral data of the conventional soil environment not affected by the independent variable is combined with the attribute data detected in the above steps to form a data table. The difference obtained by the one-by-one horizontal comparison of the two sets of data is the difference.

[0036] In the above implementation process, the set index in this embodiment specifically refers to the data with a difference lower than 1-5%. These are the corresponding data that may be less susceptible to environmental factor interference in the soil. Here, in this embodiment, the data lower than 1%-5% is screened out, and the remaining data with a difference higher than 5% is identified as the data that will cause the soil environment to be susceptible to environmental factor interference. We use this part of the data as the update amount, import the update amount into the preset soil type relationship model, and update the parameters of the preset soil type relationship model to form a trained soil type relationship model.

[0037] More specifically, both the conventional soil environment and the changing soil environment are soil environments with vegetation. Among them, the conventional soil environment is specifically the normal soil environment based on the preset vegetation planting standard area, and the changing soil environment is specifically the soil environment obtained by simulating the change of the conventional soil environment through independent variables.

[0038] Among them, in this embodiment, the above-mentioned vegetation soil environment is actually a soil environment with crops in this embodiment.

[0039] More specifically, the changing soil environment is simulated in a specific form, specifically: the changing soil environment is simulated by setting at least one independent variable.

[0040] More specifically, at least one form of independent variable is set for simulation, where the independent variables include: land use type, parent material lithology, geomorphic type, elevation, slope, aspect, slope position, plane curvature, profile curvature, rainfall, terrain humidity index, remote sensing spectrum, and rotation method.

[0041] In some embodiments, before combining the data bands of the conventional soil environment and the changing soil environment, a preprocessing operation is also included, and the steps are as follows:

[0042] Remove noise from the data bands of the conventional soil environment and the changing soil environment through the wavelet transform algorithm;

[0043] Perform dimensionality reduction on the denoised data bands through the principal component analysis algorithm;

[0044] Obtain the spectral data of the conventional soil environment and the spectral data of the changing soil environment, and obtain the vegetation indices of the two soil environments by converting the spectral data.

[0045] In the above implementation process, the data bands of the conventional soil environment extracted and the data bands of the changing soil environment extracted may cause relatively large noise at the edges of the data bands of the two soil type environments due to the influence of the simulation system environment. Therefore, in order to avoid the influence that noise may bring to the research of the two soil type environments, the wavelet transform has characteristics such as multi-resolution, low entropy, flexible selection of basis functions, and decorrelation. It has high resolution in the high-frequency range and low frequency resolution in the low-frequency range, and is suitable for analyzing signals at any scale. Therefore, in this embodiment, the wavelet transform algorithm is used to compress and denoise the spectral data of the two soil type environments within the set range; in addition, since the soil spectral data has the characteristics of large data volume and high resolution, but large data redundancy and large calculation amount, therefore, in this embodiment, the principal component analysis algorithm is used to perform dimensionality reduction on the denoised data bands to avoid situations such as large data redundancy and large calculation amount. Specifically, the principal component analysis algorithm and the wavelet transform algorithm are both existing algorithms, and will not be explained in this embodiment.

[0046] In some embodiments, based on the vegetation indices of the two soil environments, the vegetation index of the conventional soil environment is used as the reference group, and at the same time, the vegetation index of the changing soil environment is used as the calculation group. The reference group and the calculation group are calculated through the index fitting function to obtain the fitting coefficient between the independent variable and the vegetation index of the conventional soil environment, and verify whether the coincidence degree of the fitting coefficient reaches the expected value. If so, the fitting coefficient is output to construct the preset soil type relationship model; if not, return to obtain the fitting coefficient again until the coincidence degree of the fitting coefficient reaches the expected value.

[0047] In the above implementation process, the vegetation index of the conventional soil type environment is used as the test data of the reference group, and the vegetation index of the soil type environment affected by changes in environmental factors is used as the experimental data of the calculation group, that is, the comparison group. In order to better interpret the relationship between the soil type with vegetation and environmental factors, this embodiment uses an exponential fitting function to perform fitting calculations on the two types of test data, and determines whether the obtained correlation coefficient, that is, the fitting coefficient, reaches the expected degree of coincidence, and uses the output fitting coefficient as the most preferred for constructing the soil type relationship model, improving the accuracy of the soil type relationship model.

[0048] In the above implementation process, the exponential fitting function in this embodiment is specifically an existing algorithm. By calculating the fitting function or the correlation coefficient, it is verified whether the degree of coincidence (the degree of coincidence is based on being close to the historical experimental data) between the fitting function or the correlation coefficient and the historical experimental data reaches the expectation, so as to judge whether the calculated fitting function or the correlation coefficient is reasonable.

[0049] More specifically, after the soil type relationship model is completed training, the accuracy of the soil type relationship model is also verified through a preset test set.

[0050] The above are only the preferred embodiments of the present invention and are not used to limit the present invention. For those skilled in the art, the present invention can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A method for establishing a relationship model between soil types and environmental factors, characterized in that, The steps of the method include: Define multiple research areas within the simulation system, and classify the multiple research areas into a conventional soil environment and a changing soil environment; Simulate the changing soil environment in a specific form, and simulate the conventional soil environment in a standardized manner until the simulation duration reaches the threshold and then stop; Detect the attribute data of the two land types, and extract the band data of the conventional soil environment and the changing soil environment; Combine the band data of the conventional soil environment and the changing soil environment respectively, and analyze the conventional soil environment and the changing soil environment in combination with the attribute data to obtain the difference between the changing soil environment and the conventional soil environment; Screen out the data with a difference lower than the set index, and import it into the preset soil type relationship model for iterative update to complete the training of the soil type relationship model; Both the conventional soil environment and the changing soil environment are soil environments with vegetation. Among them, the conventional soil environment is specifically the normal soil environment based on the preset vegetation planting standard area, and the changing soil environment is specifically the soil environment obtained by simulating the change of the conventional soil environment through independent variables; Simulating the changing soil environment in a specific form specifically means: simulating the changing soil environment by setting at least one independent variable; Before combining the band data of the conventional soil environment and the changing soil environment, a preprocessing operation is also included, and the steps are as follows: Remove the noise of the band data of the conventional soil environment and the changing soil environment through the wavelet transform algorithm; Perform dimensionality reduction processing on the denoised band data through the principal component analysis algorithm; Obtain the spectral data of the conventional soil environment and the spectral data of the changing soil environment, and obtain the vegetation indices of the two soil environments by converting the spectral data; Based on the vegetation indices of the two soil environments, take the vegetation index of the conventional soil environment as the reference group, and at the same time take the vegetation index of the changing soil environment as the calculation group. Calculate the reference group and the calculation group through the index fitting function to obtain the fitting coefficient between the independent variable and the vegetation index of the conventional soil environment, and verify whether the coincidence degree of the fitting coefficient reaches the expected value. If so, output the fitting coefficient to construct the preset soil type relationship model; if not, return to obtain the fitting coefficient again until the coincidence degree of the fitting coefficient reaches the expected value.

2. The method for establishing a relationship model between soil types and environmental factors according to claim 1, characterized in that, Set at least one form of independent variable for simulation, where the independent variables include: altitude, atmospheric temperature, surface temperature, precipitation, atmospheric humidity, and surface humidity.

3. The method for establishing a relationship model between soil types and environmental factors according to claim 2, characterized in that After the soil type relationship model is trained, the accuracy of the soil type relationship model is also verified through a preset test set.

Images