Soil organic pollution concentration prediction method and system based on resistivity parameters

Abstract

The present application provides a method and system for predicting soil organic pollution concentration based on resistivity parameters, wherein the method includes: in an organic pollution detection scenario under a saline-alkali soil environment, by synchronously acquiring the dielectric constant distribution data, conductivity spatial distribution data, resistivity parameters and polarizability parameters of the target area, a bimodal feature matrix reflecting the electrical characteristics of the soil is constructed. The bimodal feature matrix is ​​multi-source coupled with the dielectric constant and conductivity data to generate a dynamic data set, and then the frequency domain dielectric spectrum analysis is used to extract the polarization attenuation characteristic differences between organic pollutants and salt ions. Combined with a machine learning model, the spatial concentration distribution prediction results of organic pollutants in a highly saline-alkali environment are finally output. The present application improves the accuracy and spatial positioning accuracy of organic pollutant detection in highly saline-alkali soils.

Description

Technical Field

[0001] The present application relates to the field of resistivity parameters and machine learning technology, and in particular to a method and system for predicting soil organic pollution concentration based on resistivity parameters. Background Art

[0002] In salinized soil environments, the accurate detection of organic pollutants presents a significant challenge. Due to the cross-interference between the electrochemical properties of salt ions and organic pollutants, traditional single-parameter detection methods struggle to effectively distinguish between the two, leading to inaccurate pollution assessments.

[0003] A soil contamination detection method based on near-infrared spectroscopy has been proposed. This method collects soil near-infrared spectral data, extracts the spectral reflectance of characteristic bands, and uses a support vector machine algorithm to establish a predictive model linking spectral characteristics with pollutant concentrations, enabling quantitative analysis of pollutants.

[0004] While this method utilizes spectroscopy to achieve non-contact detection, it lacks effective means to address the spectral interference caused by high salt levels in salinized soils, which can easily mask the characteristic wavelengths of organic pollutants. Furthermore, the method struggles to distinguish between the responses of salt ions and organic pollutants, significantly reducing detection accuracy in areas with high salinity. Summary of the Invention

[0005] The present application provides a soil organic pollution concentration prediction method and system based on resistivity parameters, which is used to solve the problems of low accuracy and poor spatial positioning accuracy in the detection of organic pollutants in highly saline and alkaline soils in the prior art.

[0006] In a first aspect, the present application provides a method for predicting soil organic pollution concentration based on resistivity parameters, comprising:

[0007] In the organic pollution detection scenario under saline-alkali soil environment, the dielectric constant distribution data, conductivity spatial distribution data, resistivity parameters and polarizability parameters of the soil in the target area are obtained;

[0008] generating a bimodal characteristic matrix based on the resistivity parameter and the polarizability parameter;

[0009] Performing multi-source coupling correlation on the bimodal feature matrix, the dielectric constant distribution data, and the conductivity spatial distribution data to generate a dynamic data set;

[0010] Combined with the frequency domain dielectric spectrum analysis method, the dynamic data set is subjected to frequency band screening to extract the polarization attenuation characteristic differences between organic pollutants and salt ions within the preset frequency band;

[0011] Based on the polarization attenuation characteristic differences and the bimodal characteristic matrix, a trained machine learning model is used to output prediction results of the spatial concentration distribution of organic pollutants in a saline-alkali soil environment.

[0012] Optionally, the frequency domain dielectric spectrum analysis method is combined to perform frequency band screening on the dynamic data set to extract the polarization attenuation characteristic differences between organic pollutants and salt ions within a preset frequency band, including:

[0013] Converting the polarization response signal in the dynamic data set into frequency domain distribution data by frequency domain dielectric spectrum analysis;

[0014] Based on the differences in polarization characteristics of organic pollutants and salt ions in saline-alkali soil environments, the boundary between the low-frequency band and the mid-frequency band within the preset frequency band is determined;

[0015] Within the boundary interval, based on the frequency domain distribution data, frequency points whose signal strength first increases and then decreases sharply as the frequency increases are screened out to form a frequency point set;

[0016] The characteristic frequency band corresponding to the organic pollutants and the interference frequency band corresponding to the salt ions are extracted from the frequency point set, and the signal attenuation slope difference between the characteristic frequency band and the interference frequency band is calculated, and the signal attenuation slope difference is used as the polarization attenuation characteristic difference.

[0017] Optionally, converting the polarization response signal in the dynamic data set into frequency domain distribution data by frequency domain dielectric spectrum analysis includes:

[0018] extracting a polarization response signal from the dynamic data set;

[0019] intercepting the polarization response signal according to a preset segment length to generate a plurality of time domain signal sequences;

[0020] Perform frequency domain conversion processing on each time domain signal sequence to generate a target signal strength distribution sequence corresponding to each time domain signal sequence;

[0021] determining a frequency range of the target signal intensity distribution sequence according to a proportional relationship between a conductivity parameter and the polarizability parameter in the conductivity spatial distribution data;

[0022] The target signal intensity distribution sequences corresponding to all time domain signal sequences are superimposed to generate frequency domain distribution data covering the frequency band range.

[0023] Optionally, extracting a characteristic frequency band corresponding to organic pollutants and an interference frequency band corresponding to salt ions from the frequency point set, and calculating a signal attenuation slope difference between the characteristic frequency band and the interference frequency band, includes:

[0024] According to the difference in the steep drop time sequence of the signal strength in the frequency point set, N consecutive frequency points in the frequency point set whose steep drop time sequence is located in the starting area of ​​the boundary interval are divided into an organic pollutant characteristic frequency band, and consecutive frequency points whose steep drop time sequence is located in the end area of ​​the boundary interval are divided into a salt ion interference frequency band, wherein N is greater than or equal to 3;

[0025] In the characteristic frequency band of organic pollutants, a turning point where the signal strength changes from increasing to dropping sharply is selected as a first reference point, and a first analysis window is generated by extending a preset number of frequency points to both sides of the first reference point.

[0026] In the salt ion interference frequency band, a frequency point at which the signal intensity drops sharply and exceeds a preset amplitude threshold is selected as a second reference point, and a second analysis window is generated by extending the same preset number of frequency points to both sides of the second reference point as the center;

[0027] Calculate respectively a first average decreasing rate of the signal strength of the frequency points after the turning frequency point in the first analysis window and a second average decreasing rate of the signal strength of the frequency points after the turning frequency point in the second analysis window;

[0028] The absolute value difference between the first average decreasing rate and the second average decreasing rate is used as the signal attenuation slope difference.

[0029] Optionally, performing frequency domain conversion processing on each time domain signal sequence to generate a target signal strength distribution sequence corresponding to each time domain signal sequence includes:

[0030] Multiply each time domain signal sequence by a preset window function sequence point by point to obtain a time domain signal subsequence;

[0031] Performing spectrum analysis on the time domain signal subsequence to generate an initial signal strength distribution sequence;

[0032] Frequency units in the initial signal strength distribution sequence that are lower than a preset environmental noise threshold are set to zero to generate a target signal strength distribution sequence.

[0033] Optionally, generating a bimodal characteristic matrix based on the resistivity parameter and the polarizability parameter includes:

[0034] In the saline-alkali soil environment, synchronously aligning the resistivity parameter and the polarizability parameter at the same spatial position to generate a parameter pair set;

[0035] Converting the resistivity parameter in each parameter pair in the parameter pair set into a first eigenvalue, and converting the polarizability parameter into a second eigenvalue;

[0036] Combining the first eigenvalue and the second eigenvalue at the same spatial position into a two-dimensional eigenvector;

[0037] The two-dimensional feature vectors of all spatial positions are arranged to form a bimodal feature matrix.

[0038] Optionally, the outputting of a prediction result of the spatial concentration distribution of organic pollutants in a salinized soil environment using a trained machine learning model based on the polarization attenuation characteristic difference and the bimodal characteristic matrix includes:

[0039] Combining each two-dimensional eigenvector in the dual-modal feature matrix with the polarization attenuation characteristic difference at a corresponding spatial position to generate an enhanced eigenvector;

[0040] performing dynamic weighting processing on the enhanced feature vector according to a salt ion concentration gradient generated by a conductivity parameter in the conductivity spatial distribution data;

[0041] The weighted enhanced feature vectors are input into the trained machine learning model according to the preset spatial distribution order to generate the initial pollution concentration value for each spatial location;

[0042] Performing dielectric correction processing on the initial contamination concentration value based on the dielectric constant distribution data;

[0043] The corrected pollution concentration value and the corresponding spatial position coordinates are associated to output the spatial concentration distribution prediction results.

[0044] In a second aspect, the present application provides a soil organic pollution concentration prediction system based on resistivity parameters, comprising:

[0045] The acquisition module is used to obtain the dielectric constant distribution data, conductivity spatial distribution data, resistivity parameters, and polarizability parameters of the soil in the target area in the organic pollution detection scenario under salinized soil environment;

[0046] A first generating module is configured to generate a bimodal characteristic matrix based on the resistivity parameter and the polarizability parameter;

[0047] A second generating module is configured to perform multi-source coupling correlation on the bimodal feature matrix, the dielectric constant distribution data, and the conductivity spatial distribution data to generate a dynamic data set;

[0048] A screening module, configured to perform frequency band screening on the dynamic data set by frequency domain dielectric spectrum analysis to extract the polarization attenuation characteristic differences between organic pollutants and salt ions within a preset frequency band;

[0049] The output module is used to output the spatial concentration distribution prediction results of organic pollutants in salinized soil environment based on the polarization attenuation characteristic difference and the bimodal characteristic matrix using a trained machine learning model.

[0050] In a third aspect, the present application provides a computing device comprising a processor and a memory, wherein the memory stores a computer program, and the processor is configured to run the computer program to execute a soil organic pollution concentration prediction method based on resistivity parameters as described in any one of the first aspects.

[0051] In a fourth aspect, the present application provides a computer storage medium having computer program instructions stored thereon, which, when executed by a processor, implements a soil organic pollution concentration prediction method based on resistivity parameters as described in any one of the first aspects.

[0052] In the present application, a method for predicting soil organic pollution concentration based on resistivity parameters is provided, which includes: in an organic pollution detection scenario under a salinized soil environment, obtaining dielectric constant distribution data, conductivity spatial distribution data, resistivity parameters and polarizability parameters of the soil in the target area; generating a bimodal feature matrix based on the resistivity parameters and the polarizability parameters; performing multi-source coupling correlation on the bimodal feature matrix, the dielectric constant distribution data and the conductivity spatial distribution data to generate a dynamic data set; combining the frequency domain dielectric spectrum analysis method, performing frequency band screening on the dynamic data set to extract the polarization attenuation characteristic differences between organic pollutants and salt ions within a preset frequency band; based on the polarization attenuation characteristic differences and the bimodal feature matrix, using a trained machine learning model to output the spatial concentration distribution prediction results of organic pollutants in the salinized soil environment.

[0053] The technical solution provided by this application has the following beneficial effects:

[0054] This application establishes a comprehensive soil electrical characteristic database to provide multi-dimensional data support for subsequent analysis. It constructs a feature expression that simultaneously reflects soil conductivity and polarization properties, enhancing feature characterization capabilities. It enables collaborative analysis of multi-source electrical parameters, improving data complementarity and integrity. It effectively separates the characteristic frequency bands of organic pollutants and salt ions, suppressing background interference from salinization. It also achieves high-precision visualization of the spatial distribution of organic pollutants.

[0055] Furthermore, the present application also converts the polarization response signal into frequency domain distribution data through frequency domain dielectric spectrum analysis, determines the boundary interval based on the difference in polarization characteristics of organic pollutants and salt ions in a salinized environment, screens the frequency point set of specific signal change patterns, and then extracts the characteristic frequency band and interference frequency band and calculates the difference in their signal attenuation slope as the characteristic difference.

[0056] In addition, this technical solution effectively distinguishes the differences in polarization characteristics between organic pollutants and salt ions through frequency domain characteristic analysis of polarization response signals, and uses quantitative comparison of signal attenuation slopes to improve the identification accuracy of organic pollutants and the reliability of feature extraction in the context of high salinity and alkali.

[0057] These and other aspects of the present application will become more readily apparent from the description of the following embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0059] Figure 1 A flow chart of a method for predicting soil organic pollution concentration based on resistivity parameters provided in an embodiment of the present application;

[0060] Figure 2 A schematic diagram of the structure of a soil organic pollution concentration prediction system based on resistivity parameters provided in an embodiment of the present application;

[0061] Figure 3 A schematic diagram of the structure of a computing device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0062] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application.

[0063] In some of the processes described in the specification and claims of this application and the above-mentioned figures, multiple operations that appear in a specific order are included, but it should be clearly understood that these operations may not be executed in the order in which they appear in this document or may be executed in parallel. The serial numbers of the operations, such as 101, 102, etc., are only used to distinguish between different operations, and the serial numbers themselves do not represent any order of execution. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions of "first", "second", etc. in this document are used to distinguish different messages, devices, modules, etc., and do not represent a sequential order, nor do they limit "first" and "second" to being different types.

[0064] Researchers have found that in saline-alkali soil environments, traditional organic pollutant detection methods have difficulty effectively distinguishing between salt ion interference and organic pollution characteristics, resulting in insufficient detection accuracy. Based on this, a soil organic pollution concentration prediction method based on resistivity parameters is provided. This method synchronously collects soil dielectric constant, conductivity, resistivity and polarizability parameters, constructs a dual-modal feature matrix and performs multi-source data coupling, and combines frequency domain dielectric spectrum analysis to extract the polarization attenuation characteristic differences between organic pollutants and salt ions. Finally, a machine learning model is used to achieve accurate spatial distribution prediction of organic pollutants in highly saline-alkali environments. The technical solution of this application can be applied to organic pollution detection and assessment scenarios in highly saline-alkali soil environments such as saline-alkali land and coastal mudflats.

[0065] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without making creative efforts are within the scope of protection of this application.

[0066] Figure 1 A flow chart of a method for predicting soil organic pollution concentration based on resistivity parameters provided in an embodiment of the present application is shown in FIG. Figure 1 As shown, the method includes:

[0067] Step 101: In an organic pollution detection scenario in a saline-alkali soil environment, obtain dielectric constant distribution data, conductivity spatial distribution data, resistivity parameters, and polarizability parameters of the soil in the target area.

[0068] In this step, the dielectric constant distribution data reflects the soil's response to electric fields at different locations and is used to characterize the soil's electrical properties. The spatial distribution data of electrical conductivity describes the spatial variation of soil electrical conductivity and is used to assess the degree of salinization. The resistivity parameter, which has an inverse relationship with electrical conductivity, describes the soil's ability to resist the flow of electric current. The polarizability parameter reflects the degree to which the soil polarizes under an applied electric field.

[0069] In an embodiment of the present application, the target detection area is first determined, and professional soil electrical parameter measurement equipment is used to synchronously measure the four parameters of dielectric constant, conductivity, resistivity and polarizability at each sampling point according to a preset grid sampling scheme. The discrete point measurement data is converted into continuous spatial distribution data through spatial interpolation method, and finally a complete regional soil electrical parameter data set is obtained.

[0070] For example, in a salinized farmland pollution investigation, the target area was divided into a 10-meter x 10-meter grid. At each grid center, a quadrupole soil resistivity meter was used to measure resistivity parameters. A frequency-domain dielectric spectrometer was used to measure dielectric constant and polarizability parameters, and an electromagnetic induction meter was used to obtain conductivity data. All measured data were interpolated using kriging to generate a continuous spatial distribution map, providing the foundation for subsequent analysis.

[0071] Step 102: Generate a bimodal characteristic matrix based on the resistivity parameter and the polarizability parameter.

[0072] In this step, the bimodal characteristic matrix represents a matrix composed of characteristic data of two modes, resistivity and polarizability, which is used to comprehensively characterize the electrical characteristics of the soil.

[0073] In an embodiment of the present application, the resistivity and polarizability data at the same spatial position are first normalized to eliminate dimensional differences, and then the two parameters are converted into eigenvalues ​​with physical meaning through a feature fusion algorithm. Finally, the eigenvalues ​​of all points are arranged and combined in order of spatial position to form a feature matrix containing dual information of resistivity and polarizability.

[0074] For example, the resistivity and polarizability data of each sampling point in the farmland are standardized respectively, and the principal component analysis method is used to extract the main characteristic components of the two parameters. The processed eigenvalues ​​are arranged in the order of the sampling point coordinates to construct a 100×2 bimodal feature matrix, in which each row represents a sampling point and the two columns correspond to the resistivity and polarizability eigenvalues, respectively.

[0075] Step 103: performing multi-source coupling correlation on the bimodal feature matrix, the dielectric constant distribution data, and the conductivity spatial distribution data to generate a dynamic data set.

[0076] In this step, the dynamic dataset refers to the dataset with spatiotemporal characteristics formed by integrating multi-source data.

[0077] In an embodiment of the present application, the bimodal feature matrix is ​​first spatially aligned with the dielectric constant distribution data to ensure that the data positions correspond. Then, the conductivity spatial distribution data is introduced as a weight factor, and the three types of data are correlated and integrated through a data fusion algorithm to finally generate a dynamic data set containing complete electrical characteristics.

[0078] For example, the spatial coordinates of the bimodal feature matrix and the dielectric constant distribution map are matched, the conductivity data is used to calculate the fusion weight of each point, and the three types of data are integrated using a weighted fusion algorithm to form a dynamic data set containing resistivity, polarizability, dielectric constant and conductivity information, providing a data basis for subsequent frequency domain analysis.

[0079] Step 104: frequency-domain dielectric spectrum analysis is combined to perform frequency band screening on the dynamic data set to extract the polarization attenuation characteristic differences between organic pollutants and salt ions within a preset frequency band.

[0080] In this step, the difference in polarization attenuation characteristics represents the distinguishing features of the polarization attenuation characteristics between organic pollutants and salt ions.

[0081] In an embodiment of the present application, the polarization response signal in the dynamic data set is first transformed in the frequency domain, and the characteristic analysis frequency band is determined according to the characteristics of salinized soil. Then, the characteristic points of signal change are identified within the specific frequency band, and the characteristic frequency bands of organic pollutants and salt ions are extracted respectively. Finally, the signal attenuation difference between the two frequency bands is calculated as the distinguishing feature.

[0082] For example, for the farmland dynamic data set, the 50-500 Hz frequency band was selected for frequency domain analysis, the signal turning point was identified within this frequency band, 80-200 Hz was divided into the organic matter characteristic frequency band, and 300-450 Hz was divided into the salt ion interference frequency band. The difference in the signal attenuation slope of the two frequency bands was calculated as the characteristic difference indicator.

[0083] Step 105: Based on the polarization attenuation characteristic difference and the bimodal characteristic matrix, the trained machine learning model is used to output the prediction result of the spatial concentration distribution of organic pollutants in the salinized soil environment.

[0084] In this step, the trained machine learning model refers to a prediction model that has been pre-trained using a large amount of sample data. The spatial concentration distribution prediction result refers to the spatial distribution information of organic pollutants generated by the model calculation.

[0085] In an embodiment of the present application, the polarization attenuation feature difference is first fused with the bimodal feature matrix at the feature level, and then input into a pre-trained machine learning model. The model outputs a predicted value based on the learned relationship between the feature and the concentration, and finally generates a concentration distribution map in combination with the spatial position information.

[0086] For example, after fusing the feature differences in farmland data with bimodal features, the data is input into a trained random forest model. The model outputs the predicted concentration value of each sampling point and generates a heat map of the organic pollutant concentration distribution in the entire farmland area through spatial interpolation.

[0087] This method effectively overcomes the interference problem in the detection of organic pollutants in saline-alkali soil through the comprehensive collection and analysis of multi-source electrical parameters, realizes the accurate identification and spatial distribution visualization of organic pollutants, and provides a reliable technical means for saline-alkali land pollution control.

[0088] To address the technical challenge of accurately distinguishing organic pollutants from salt ion interference in a highly saline and alkaline soil environment, in some embodiments, step 104: combining frequency domain dielectric spectroscopy analysis to perform frequency band screening on the dynamic data set to extract the polarization attenuation characteristic differences between organic pollutants and salt ions within a preset frequency band includes:

[0089] Step 201: converting the polarization response signal in the dynamic data set into frequency domain distribution data by frequency domain dielectric spectrum analysis.

[0090] In step 201, the polarization response signal represents the time-varying electrical response signal generated by the soil under the action of an applied alternating electric field. The frequency domain distribution data represents the frequency-signal intensity distribution spectrum obtained by spectrally converting the time domain polarization response signal. This data is used to analyze the electrical response characteristics at different frequencies.

[0091] In this embodiment, the collected time-domain polarization signals are converted into frequency-domain signals using discrete Fourier transform technology. The specific process involves first windowing the original time-domain signals in the dynamic dataset, then performing a fast Fourier transform (FFT), ultimately generating frequency-domain distribution data containing amplitude and phase information. This data is presented as a spectrogram with frequency as the horizontal axis and signal strength as the vertical axis.

[0092] Step 202: Based on the difference in polarization characteristics between organic pollutants and salt ions in a saline-alkali soil environment, a boundary interval between a low frequency band and a mid frequency band within a preset frequency band is determined.

[0093] In step 202, the difference in polarization characteristics represents the difference between the low-frequency relaxation characteristics of organic matter and the mid-frequency migration characteristics of salt ions. For example, in highly saline-alkali soil, the polarizability attenuation of sodium chloride solution at 10 kHz is 3–5 times greater than that of petroleum hydrocarbon pollutants, and the peak frequencies of the two polarizability characteristics are separated by at least an order of magnitude. The boundary region represents the transition frequency band where the polarization response characteristics of organic pollutants and salt ions intersect. For example, the polarization response of organic pollutants (such as petroleum hydrocarbons and benzene series) is primarily in the low-frequency range (e.g., 1 Hz–100 Hz) due to their longer molecular polarization relaxation time. The polarization response of salt ions (such as Na⁺ and Cl⁻) is concentrated in the mid-frequency range (e.g., 100 Hz–10 kHz) due to their faster ion migration polarization rate.

[0094] In this application's examples, based on statistical analysis of extensive experimental data, the characteristic analysis frequency range of 50-500 Hz was determined. Within this range, 80-200 Hz was designated as the low-frequency band (dominated by organic matter), 200-450 Hz as the mid-frequency band (dominated by salt ions), and the overlapping range of 150-250 Hz as the critical boundary. This division is based on the fundamental differences in the frequency response characteristics between the molecular polarization relaxation of organic pollutants and the ionic migration of salt ions.

[0095] Step 203: Within the boundary interval, based on the frequency domain distribution data, filter out frequency points where the signal strength first increases and then decreases sharply as the frequency increases, to form a frequency point set.

[0096] In step 203, the frequency point set represents a set of discrete frequency points that satisfy a specific signal variation rule within the boundary interval.

[0097] In this embodiment, an automatic peak detection algorithm scans the frequency domain data within the boundary interval to identify frequency points that meet the following characteristics: 1. The presence of clear rising and falling signal edges; 2. The falling slope exceeds a set threshold; 3. The peak amplitude reaches a specific multiple of the background noise. Frequency points that meet all these conditions are arranged in chronological order to form a characteristic frequency point set.

[0098] Step 204: extracting the characteristic frequency band corresponding to the organic pollutants and the interference frequency band corresponding to the salt ions from the frequency point set, calculating the signal attenuation slope difference between the characteristic frequency band and the interference frequency band, and using the signal attenuation slope difference as the polarization attenuation characteristic difference.

[0099] In step 204, the characteristic frequency band represents a continuous frequency interval dominated by the polarization response of organic pollutants. The interference frequency band represents a continuous frequency interval dominated by the polarization response of salt ions. The signal attenuation slope difference represents the difference in the signal decay rate of the two frequency bands.

[0100] In this embodiment, a cluster analysis is first performed based on the temporal characteristics of the frequency set. The frequency points that experience the earliest steep drop are classified as the characteristic frequency band (organic matter), while the later ones are classified as the interference frequency band (salt ions). The average attenuation slopes of the two frequency bands are then calculated, and the arithmetic difference between the two slopes is finally used as the polarization attenuation characteristic difference value.

[0101] Here's a specific example:

[0102] Taking the detection of oil pollution in a salinized farmland as an example, after completing 10m×10m grid sampling and obtaining resistivity, polarizability, dielectric constant and conductivity data, the dynamic data set was subjected to frequency band screening: first, the polarization response signal of each sampling point was converted into frequency domain data through fast Fourier transform. According to the historical test data of the area, the characteristic frequency band of organic pollutants (petroleum hydrocarbons) was determined to be 80-200Hz (low frequency band), and the salt ion interference frequency band was 300-450Hz (mid frequency band). The 150-250Hz between the two frequency bands was selected as the The frequency domain data within this interval was analyzed to identify 18 characteristic frequency points where the signal intensity first peaked and then rapidly decreased (with a slope exceeding 5 decibels per second). Nine frequency points in the 160-190 Hz range (with an average slope of 1.2 units per minute) were identified as oil-specific frequencies, and nine frequency points in the 210-240 Hz range (with an average slope of 2.8 units per minute) were identified as salt interference frequencies. The polarization attenuation characteristic difference value was calculated by taking the difference between the average slopes of the two frequency bands (2.8-1.2=1.6). This characteristic difference value was calculated by first taking the arithmetic average of the instantaneous slopes of each frequency point and then taking the difference between the two average values ​​to ensure that the quantified results accurately reflect the polarization response differences between organic matter and salt ions.

[0103] In the embodiment of the present application, this technical solution effectively separates the electrical response characteristics of organic pollutants and salt ions through precise frequency band division and feature extraction, providing a reliable characteristic basis for subsequent accurate prediction of pollutant concentrations and improving the accuracy of organic pollution detection in high-salt environments.

[0104] To further improve the accuracy and anti-interference capability of frequency domain analysis, in some embodiments, step 201: converting the polarization response signal in the dynamic data set into frequency domain distribution data by frequency domain dielectric spectrum analysis includes:

[0105] Step 301: extracting a polarization response signal from the dynamic data set.

[0106] In step 301, the polarization response signal represents the electrical response data of the soil that varies with time under the action of the alternating electric field, and includes amplitude and phase information.

[0107] In an embodiment of the present application, polarization response signal data is located and read from a dynamic data set storage structure. The signal is waveform data with time as the horizontal coordinate and voltage value as the vertical coordinate. The sampling frequency must meet the requirements of the Nyquist sampling theorem to ensure signal integrity.

[0108] Step 302: The polarization response signal is segmented according to a preset segment length to generate a plurality of time domain signal sequences.

[0109] In step 302, the preset segment length represents the time window length set according to the signal characteristics and analysis requirements. The time domain signal sequence represents discrete signal segments divided into fixed time lengths.

[0110] In the embodiment of the present application, overlapping framing is used to segment the continuous polarization signal. The length of each segment is set to an integer multiple of the signal fundamental wave period (usually 5-10 periods). Partial overlapping areas are retained between adjacent segments to reduce spectrum leakage caused by truncation effects.

[0111] Step 303: Perform frequency domain conversion processing on each time domain signal sequence to generate a target signal strength distribution sequence corresponding to each time domain signal sequence.

[0112] In step 303, the frequency domain conversion process represents a digital signal processing method for converting a time domain signal into a frequency domain representation. The target signal strength distribution sequence represents the frequency-energy distribution data obtained after the frequency domain conversion.

[0113] In an embodiment of the present application, each time domain signal is first windowed (using a Hanning window function), and then the spectrum is calculated using a fast Fourier transform algorithm. Finally, the modulus value is squared to obtain a signal strength spectrum, thereby forming a target signal strength distribution sequence corresponding to each segment.

[0114] Step 304: Determine the frequency range of the target signal intensity distribution sequence according to the proportional relationship between the conductivity parameter and the polarizability parameter in the conductivity spatial distribution data.

[0115] In step 304, the conductivity-polarizability ratio represents a characteristic parameter reflecting the degree of soil salinization. This ratio is established through statistical analysis of historical sample data. The specific process involves collecting a large number of sample points in a typical salinized soil area, simultaneously measuring the conductivity and polarizability values ​​at each point. The two parameters are normalized and their ratio is calculated. Cluster analysis reveals a correlation between this ratio and the characteristic frequency bands of organic pollutants. A ratio greater than 1 indicates strong salt ion interference, necessitating analysis of the mid- to high-frequency bands (200-500 Hz). A ratio less than 0.5 indicates significant organic signatures, necessitating focus on the low-frequency bands (50-200 Hz). A ratio between 0.5 and 1 utilizes full-band analysis. Finally, these empirical relationships are compiled into a lookup table. During actual testing, the optimal analysis frequency range is determined based on the real-time conductivity and polarizability values ​​measured. The frequency range represents the effective analysis frequency band of the signal intensity distribution sequence.

[0116] In this embodiment, the analysis frequency range for each signal segment is dynamically determined based on a table of conductivity-polarizability ratios and characteristic frequency bands, established using measured data. A higher ratio prioritizes mid- and high-frequency analysis, while a lower ratio expands the low-frequency range.

[0117] Step 305: performing superposition processing on the target signal strength distribution sequences corresponding to all time domain signal sequences to generate frequency domain distribution data covering a frequency band range.

[0118] In step 305 , the superposition process represents a method of fusing frequency domain results of multiple signal segments.

[0119] In an embodiment of the present application, the target signal intensity distribution sequences of each segment are incoherently superimposed, that is, the maximum intensity value of the same frequency point is taken as the final result, the valid information of all frequency bands is retained, and the complete frequency domain distribution data is generated.

[0120] Here's a specific example:

[0121] Taking the detection of petroleum pollution in a salinized farmland as an example, after completing grid sampling and establishing a dynamic data set, the signal frequency domain conversion was performed on sampling point 3 (conductivity value 1200μS / cm, polarizability value 0.15): first, the polarization response signal of this point was extracted from the dynamic data set (sampling rate 1000Hz, duration 2 seconds), and framed into 200 milliseconds per segment (containing 10 complete cycles), with adjacent frames overlapping by 100 milliseconds, resulting in a total of 19 time domain signal sequences; after adding a Hamming window to each signal segment, a 1024-point Fourier transform was performed to obtain the signal within the range of 0-500Hz. The signal intensity spectrum at this point was calculated. Based on the conductivity-polarizability ratio of 8 (1200 / 150 = 8, where the polarizability value of 0.15 is converted to 150 for uniform dimension), a pre-established ratio-frequency comparison table was used to determine the analysis frequency range of 100-400 Hz. The intensity spectra of 19 segments within the 100-400 Hz range were summed, taking the maximum values ​​at each frequency point. This generated the frequency domain distribution data for this point, revealing a distinct peak at 180 Hz (the oil characteristic peak) and a secondary peak at 350 Hz (the salt ion interference peak). This provided accurate frequency domain data for subsequent feature difference analysis. The frame length of 200 milliseconds was determined based on 10 times the period of the signal's fundamental frequency of 50 Hz (20 millisecond period). The frequency range of 100-400 Hz was determined based on the "5-10" range, which corresponds to the "focus on mid- and high-frequency analysis" rule.

[0122] In the embodiment of the present application, this technical solution effectively extracts spectral features with physical significance through scientific and reasonable signal segmentation and frequency domain conversion methods, combined with dynamic adjustment of analysis frequency bands of soil electrical parameters, laying a solid foundation for the subsequent accurate distinction between organic pollutants and salt ion interference, and improving the reliability of detection results.

[0123] To further improve the accuracy of identifying characteristic frequency bands of organic pollutants and salt ions, in some embodiments, step 204: extracting the characteristic frequency band corresponding to the organic pollutants and the interference frequency band corresponding to the salt ions from the frequency point set, and calculating the signal attenuation slope difference between the characteristic frequency band and the interference frequency band, includes:

[0124] Step 401: Based on the difference in the steep drop timing of the signal strength in the frequency point set, N consecutive frequency points in the frequency point set whose steep drop timing is located in the starting area of ​​the boundary interval are divided into an organic pollutant characteristic frequency band, and consecutive frequency points whose steep drop timing is located in the end area of ​​the boundary interval are divided into a salt ion interference frequency band, where N is greater than or equal to 3.

[0125] In step 401, the steep drop timing difference indicates the relative order of the falling edges of the signal strength on the time axis. The boundary interval is divided into only two regions: the starting region (corresponding to the characteristic frequency band of organic matter) and the ending region (corresponding to the salt ion interference frequency band). The relationship and difference between the boundary interval and the starting region: The starting region is the first half of the boundary interval (low-frequency side) and is used to capture the polarization response of organic matter. The boundary interval is the entire range (e.g., 50Hz–200Hz), and the starting region is a subset of it (e.g., 50Hz–120Hz). The relationship and difference between the boundary interval and the ending region: The ending region is the second half of the boundary interval (high-frequency side) and is used to capture the salt ion polarization response. The ending region (e.g., 120Hz–200Hz) does not overlap with the starting region; together, they constitute the complete boundary interval.

[0126] In an embodiment of the present application, all frequency points in the boundary interval are first sorted from low to high according to frequency, and the starting time of the signal strength decline of each frequency point is calculated. The five consecutive frequency points that begin to decline in the first third of the frequency band are designated as the characteristic frequency band of organic pollutants, and the five consecutive frequency points that begin to decline in the last third of the frequency band are designated as the salt ion interference frequency band, ensuring that the two types of characteristic frequency bands are completely separated in frequency distribution.

[0127] Step 402: Within the characteristic frequency band of organic pollutants, select a turning point where the signal strength changes from increasing to dropping sharply as a first reference point, and extend a preset number of frequency points to both sides with the first reference point as the center to generate a first analysis window.

[0128] In step 402, the turning frequency point represents the critical frequency point where the signal strength change trend changes from increasing to decreasing. The first analysis window represents the local frequency band analysis range centered at the turning point.

[0129] In an embodiment of the present application, the signal strength change rate of each frequency point in the characteristic frequency band is calculated by the differential method. When the change rate turns from positive to negative, it is determined to be a turning frequency point. With this point as the center, two adjacent frequency points are taken on both sides of the high and low frequencies to form a first analysis window containing 5 frequency points.

[0130] Step 403: In the salt ion interference frequency band, a frequency point where the signal strength drops sharply above a preset amplitude threshold is selected as a second reference point, and a second analysis window is generated by extending the same preset number of frequency points to both sides of the second reference point.

[0131] In step 403, the preset amplitude threshold represents the minimum signal strength change for determining a valid steep drop. The second analysis window represents an interference frequency band analysis range that is symmetrical to the first window.

[0132] In an embodiment of the present application, the signal strength drop amplitude of each frequency point is calculated within the interference frequency band, and the frequency point with an amplitude exceeding 3 times the background noise level is selected as the second reference point. Similarly, two frequency points are extended to both sides to form a second analysis window, maintaining the same analysis scale as the first window.

[0133] Step 404: Calculate a first average decreasing rate of the signal strength of the frequency points after the turning frequency point in the first analysis window and a second average decreasing rate of the signal strength of the frequency points after the turning frequency point in the second analysis window.

[0134] In step 404, the average decreasing rate represents the average decreasing speed of the signal strength as a function of frequency within the analysis window.

[0135] In an embodiment of the present application, the signal intensity change of the frequency band after the turning point is calculated for each of the two analysis windows, the linear part of the intensity-frequency curve is fitted using the least squares method, and the absolute value of the slope of the fitting line is used as the average decrease rate of the corresponding frequency band.

[0136] Step 405: taking the absolute value difference between the first average decreasing rate and the second average decreasing rate as the signal attenuation slope difference.

[0137] In step 405 , the absolute value difference represents the absolute value of the difference between the two descent rate values.

[0138] In the embodiment of the present application, the organic matter decrease rate obtained in the first analysis window is subtracted from the salt ion decrease rate in the second analysis window, and the absolute value is taken as the final characteristic difference index.

[0139] Here's a specific example:

[0140] Taking the detection of petroleum pollution in a salinized farmland as an example, the 18 characteristic frequency points (signal intensity first increases and then decreases, and the decline slope exceeds 5 decibels per second) screened out from sampling point 7 in the 150-250Hz boundary range are processed: first, the first 9 frequency points (in the range of 160-190Hz) are divided into the petroleum characteristic frequency band according to the order of frequency point appearance, and the last 9 frequency points (in the range of 210-240Hz) are divided into the salt interference frequency band; in the petroleum characteristic frequency band, the turning point at 175Hz where the signal intensity reaches the peak and then begins to decrease is selected as the first reference point, and two frequency points are extended to both sides to form the first analysis window of 170-180Hz; in the salt interference frequency band, the frequency point at 230Hz where the decline reaches the preset threshold (more than 3 times the background noise) is selected as the second reference point, and two frequency points are extended to both sides to form the second analysis window of 225-235Hz; the two frequency points are calculated respectively. The signal strength decrease rate of the frequency band after the turning point in each window is calculated. The decrease rates of the three frequency points (176Hz, 178Hz, and 180Hz) in the first analysis window are 1.1, 1.2, and 1.3 units per minute, respectively. The arithmetic mean is taken to obtain a first average decrease rate of 1.2 units. The decrease rates of the three frequency points (231Hz, 233Hz, and 235Hz) in the second analysis window are 2.7, 2.8, and 2.9 units, respectively. The average is taken to obtain a second average decrease rate of 2.8 units. Finally, the two rates are subtracted and the absolute value is taken to obtain the signal attenuation slope difference of 1.6 (2.8-1.2=1.6). In this numerical calculation process, the instantaneous slope of each frequency point is obtained by dividing the signal strength difference of adjacent frequency points by the frequency difference. For example, the slope of the 176Hz frequency point = (176Hz signal strength - 174Hz signal strength) / (176Hz - 174Hz).

[0141] In the embodiment of the present application, this technical solution effectively captures the essential difference in polarization attenuation characteristics between organic pollutants and salt ions by accurately dividing characteristic frequency bands and scientifically calculating attenuation differences, providing stable and reliable characteristic indicators for subsequent pollution identification, and improving the accuracy of organic pollution detection in high saline and alkaline environments.

[0142] To further improve the accuracy and anti-interference capability of frequency domain conversion, in some embodiments, step 303: performing frequency domain conversion on each time domain signal sequence to generate a target signal strength distribution sequence corresponding to each time domain signal sequence includes:

[0143] Step 501: perform point-by-point multiplication processing on each time domain signal sequence and a preset window function sequence to obtain a time domain signal subsequence.

[0144] In step 501, the window function sequence represents a weighted coefficient sequence for suppressing signal truncation effects, and the time domain signal subsequence represents a finite length signal segment after windowing.

[0145] In this embodiment, a Hanning window function is used to window each time-domain signal segment. The window function length is strictly consistent with the signal segment length. Through point-by-point multiplication, smooth attenuation is achieved at both ends of the signal, effectively reducing spectral leakage. The window function coefficients are generated according to a centrosymmetric cosine-squared distribution, ensuring a concentrated mainlobe and rapid sidelobe decay.

[0146] Step 502: Perform spectrum analysis on the time domain signal subsequence to generate an initial signal strength distribution sequence.

[0147] In step 502, spectrum analysis processing represents a digital signal processing method for converting a time domain signal into a frequency domain representation. The initial signal strength distribution sequence represents an unoptimized original spectrum analysis result.

[0148] In an embodiment of the present application, a discrete Fourier transform is performed on the windowed signal subsequence, and the number of transformation points is an integer power of 2 that is greater than or equal to the minimum signal length. After the complex spectrum is calculated, the modulus value is squared to generate an initial signal intensity distribution sequence containing amplitude information. The frequency resolution is determined by the sampling rate and the number of transformation points.

[0149] Step 503: performing zeroing processing on frequency units in the initial signal strength distribution sequence that are lower than a preset environmental noise threshold, to generate a target signal strength distribution sequence.

[0150] In step 503, the environmental noise threshold represents a signal strength threshold value set according to the background noise level.

[0151] In an embodiment of the present application, the average intensity of the lowest frequency band in the initial spectrum is first calculated as the noise reference, a specific multiple of the reference value is set as the noise threshold, the intensity values ​​of all frequency units are threshold judged, the units below the threshold are set to zero, the effective signal components are retained, and finally the denoised target signal intensity distribution sequence is generated.

[0152] Here's a specific example:

[0153] The fifth segment of the time domain signal (200 milliseconds, 200 sampling points) of sampling point 3 of a salinized farmland is taken as an example for frequency domain conversion processing: first, a 200-point Hamming window function is used (the window function coefficient is calculated according to the formula 0.54-0.46*cos(2πn / 199)) to multiply the original signal point by point to obtain a windowed time domain signal subsequence; the subsequence is padded with zeros to 1024 points and then fast Fourier transform is performed to obtain the initial signal intensity spectrum of 512 frequency points in the range of 0-500Hz, where the intensity value of each frequency point is the square of the modulus value of the corresponding frequency component; Taking the average intensity of 0.02 in the 120-150Hz frequency band (which theoretically contains only ambient noise) as a baseline, the preset ambient noise threshold was set to 2.5 times the baseline value, or 0.05. All frequency points in the initial spectrum with intensities below 0.05 (such as 0.03 at 80Hz and 0.04 at 420Hz) were zeroed. The resulting target signal intensity distribution sequence retains the oil-related peak at 180Hz (intensity 0.78) and the salt ion interference peak at 350Hz (intensity 0.12), while all other noise frequencies are effectively filtered out. The Hamming window coefficient calculation formula ensures a smooth transition between the two ends of the signal. The noise threshold of 0.05 is a reasonable threshold determined through statistical analysis of typical noise frequency bands, effectively distinguishing useful signals from background noise.

[0154] In the embodiment of the present application, this technical solution improves the accuracy and signal-to-noise ratio of frequency domain analysis through scientific windowing processing and noise suppression means, provides high-quality spectral data for subsequent feature extraction, and effectively ensures the accuracy of organic pollutant detection.

[0155] To further improve the characterization capability of soil electrical characteristics, in some embodiments, step 102: generating a bimodal characteristic matrix based on the resistivity parameter and the polarizability parameter includes:

[0156] Step 601: In the saline-alkali soil environment, synchronously align the resistivity parameter and the polarizability parameter at the same spatial position to generate a parameter pair set.

[0157] In step 601, spatial location refers to dividing the target soil area into an equally spaced square / hexagonal grid, with each grid center point representing a spatial location (e.g., a 1m×1m grid). Synchronous alignment is the data matching process that ensures strict spatial alignment of different parameters. A parameter pair set represents a paired dataset consisting of resistivity and polarizability data from the same location.

[0158] In an embodiment of the present application, a unified spatial coordinate index system is first established, and the resistivity and polarizability measurement data of each sampling point are positionally matched through a geographic information system, and data points whose position offset exceeds the allowable error are eliminated, ultimately forming a set of resistivity-polarizability parameter pairs that strictly correspond to each sampling point.

[0159] Step 602: Convert the resistivity parameter in each parameter pair in the parameter pair set into a first eigenvalue, and convert the polarizability parameter into a second eigenvalue.

[0160] In step 602, the first eigenvalue is used to reflect the normalized value of the soil resistivity characteristic. The second eigenvalue represents the normalized value of the soil polarization characteristic.

[0161] In the embodiment of the present application, the resistivity parameter is logarithmized to the base 10 and linearly normalized, converted to the range of 0-1 as the first eigenvalue; the polarizability parameter is angle-normalized, and its phase angle is divided by 90 degrees as the second eigenvalue to ensure that the two types of parameters are comparable.

[0162] Step 603: Combine the first eigenvalue and the second eigenvalue at the same spatial position into a two-dimensional eigenvector.

[0163] In step 603, the two-dimensional feature vector contains two-element data units of both resistivity and polarizability features.

[0164] In an embodiment of the present application, the first eigenvalue and the second eigenvalue of the same sampling point are combined in a fixed order (resistivity first, polarizability last) to form a two-dimensional vector representing the comprehensive electrical characteristics of the point, and the vector elements retain their original physical meanings.

[0165] Step 604: Arrange the two-dimensional feature vectors of all spatial positions to form a bimodal feature matrix.

[0166] In an embodiment of the present application, the two-dimensional feature vectors of each point are arranged in sequence as row vectors of the matrix according to the spatial distribution order of the sampling points (such as from left to right, from top to bottom), and an m×2 bimodal feature matrix is ​​constructed, where m is the number of sampling points and the two columns correspond to the resistivity and polarizability characteristics, respectively.

[0167] Here's a specific example:

[0168] Taking the pollution detection of a salinized farmland (200 m × 150 m) as an example, based on the completed 10 m × 10 m grid sampling (a total of 315 sampling points), the collected resistivity and polarizability data were processed: first, the built-in GPS module of the measuring equipment was used to ensure that the resistivity data of each sampling point (for example, the measured value of point 15 was 92 Ω·m) and polarizability data (for example, the measured value of point 15 was 0.22) were strictly spatially aligned; after taking the common logarithm (lg92≈1.96) of the resistivity value of each point, linear normalization was performed according to the regional resistivity range (50-300 Ω·m). The calculation formula is ( 1.96 - lg50) / (lg300 - lg50) ≈ 0.72; the polarizability value was directly divided by the regional maximum value of 0.3 to obtain 0.73, which generated the two-dimensional feature vector of point 15 [0.72, 0.73]. Following the order of the farmland sampling points (starting from the northwest corner and arranging from west to east and north to south), the feature vectors of all 315 points were used as row vectors to construct a 315-row × 2-column bimodal feature matrix. The first column (resistivity features) was obtained through the aforementioned logarithmic normalization, and the second column (polarizability features) was obtained through linear normalization. This matrix, when fused with the dielectric constant and conductivity data, provides standardized feature input for subsequent frequency domain analysis and machine learning prediction. Normalization eliminates dimensional differences between parameters, ensuring that all features have equal importance in model training.

[0169] In the embodiment of the present application, this technical solution constructs a comprehensive feature matrix that can simultaneously reflect the conductivity and polarization properties of the soil through scientific data alignment and feature conversion methods, providing a multi-dimensional electrical feature basis for the identification of organic pollutants and improving the feature characterization capability of the detection model.

[0170] To further improve the accuracy of organic pollutant concentration prediction, in some embodiments, step 105: outputting a prediction result of the spatial concentration distribution of organic pollutants in a saline-alkali soil environment using a trained machine learning model based on the polarization attenuation characteristic difference and the bimodal characteristic matrix, includes:

[0171] Step 701: Combine each two-dimensional eigenvector in the dual-modal feature matrix with the polarization attenuation feature difference at the corresponding spatial position to generate an enhanced eigenvector.

[0172] In step 701 , the enhanced feature vector refers to a multi-dimensional feature representation that integrates electrical features and spectral features.

[0173] In an embodiment of the present application, the two-dimensional electrical characteristic vector (resistivity + polarizability) of each sampling point is spliced ​​with the polarization attenuation characteristic difference value to form a three-dimensional enhanced characteristic vector, ensuring that both time domain and frequency domain characteristic information are included.

[0174] Step 702: Dynamically weighting the enhanced feature vector according to the salt ion concentration gradient generated by the conductivity parameter in the conductivity spatial distribution data.

[0175] In step 702, the salt ion concentration gradient is generated as follows: First, the spatial distribution of conductivity data is smoothed using a Gaussian filter to eliminate measurement noise. Then, based on the linear relationship model between soil conductivity and salt ion concentration (salt ion concentration = 0.8 × conductivity value + 20, where conductivity is expressed in μS / cm and concentration is expressed in mg / kg), the conductivity parameters at each point are converted to salt ion concentration values. The spatial rate of change of the concentration values ​​is then calculated to obtain gradient components in the east-west and north-south directions. Finally, these gradient components are vector-synthesized to obtain a gradient value representing the intensity of the spatial variation in salt ion concentration. The gradient direction indicates the direction of the fastest concentration increase, and the gradient magnitude reflects the severity of the concentration change. This gradient value is used in dynamic weighting to ensure that the contribution of spectral features is enhanced in areas where salt ion concentration varies. Dynamic weighting is a method that adjusts feature importance based on the intensity of salt ion interference.

[0176] In an embodiment of the present application, the salt ion concentration gradient is calculated using the conductivity parameter, and the polarization attenuation characteristic difference component in the enhanced feature vector is weighted. The higher the salt ion concentration, the greater the feature weight (weight coefficient = 1 + 0.5 × normalized salt ion concentration value).

[0177] Step 703: Input the weighted enhanced feature vectors into the trained machine learning model according to the preset spatial distribution order to generate the initial pollution concentration value for each spatial location.

[0178] In step 703, the initial pollution concentration value represents the original prediction result directly output by the machine learning model.

[0179] In an embodiment of the present application, the weighted feature vectors are input into a pre-trained gradient boosting tree model in the spatial order of the sampling points. The model outputs the predicted pollutant concentration value of each point based on the learned feature-concentration mapping relationship.

[0180] Step 704: Perform dielectric correction processing on the initial contamination concentration value based on the dielectric constant distribution data.

[0181] In step 704 , the dielectric correction process represents a post-processing method for adjusting the concentration prediction value according to the dielectric characteristics.

[0182] In the embodiment of the present application, a correction coefficient table of dielectric constant and pollutant concentration is established to compensate the initial prediction value (correction coefficient = 1 + 0.3 × (dielectric constant - regional average) / standard deviation) to improve the prediction accuracy of high dielectric constant areas.

[0183] Step 705: Associating the corrected pollution concentration value with the corresponding spatial position coordinates, and outputting the spatial concentration distribution prediction result.

[0184] In an embodiment of the present application, the corrected concentration value is associated with the sampling point coordinates, and a concentration distribution grid map with consistent resolution is generated through spatial interpolation, and different colors are used to represent the pollution level.

[0185] Here's a specific example:

[0186] Taking the pollution detection of a salinized farmland (100 m × 100 m) as an example, based on the data collection and processing of 100 sampling points: first, the bimodal eigenvector of each sampling point (such as point 23 [0.68, 0.71]) and the corresponding polarization attenuation characteristic difference value (1.5) are combined into a three-dimensional enhanced eigenvector [0.68, 0.71, 1.5]. Based on the conductivity value of 1500 μS / cm at this point, the salt ion concentration of 1220 mg / kg is calculated using the formula "salt ion concentration = 0.8 × conductivity + 20". The gradient size is then calculated by combining the concentration values ​​of the eight surrounding points (using the central difference method, the east-west gradient = (1260-1180) / 20 = 4, the north-south gradient = (1240-1200) / 20 = 2, and the composite gradient value is √(4²+2²)≈4.47), to determine the dynamic The weight coefficient is 1+0.2×4.47≈1.89, and after weighting the third dimension features, it is [0.68, 0.71, 2.84]. The vector is input into the pre-trained gradient boosting decision tree model, and the output is an initial predicted concentration value of 18.6 mg / kg. According to the dielectric constant of 9.2 at this point (regional mean 8.0, standard deviation 1.5), the correction coefficient 1+0.25×(9.2-8.0) / 1.5=1.2 is calculated to obtain a corrected concentration of 22.3 mg / kg. Finally, the corrected concentration values ​​and coordinate information of all 100 sampling points are imported into the geographic information system, and a spatial concentration distribution map with a 10-meter resolution is generated by inverse distance weighted interpolation. The gradient calculation uses a 20-meter distance between adjacent sampling points as the differential step size, and the 0.25 in the dielectric correction coefficient is an empirical adjustment parameter to ensure that the pollutant concentration in the high dielectric area is reasonably corrected.

[0187] In the examples of this application, this technical solution, through innovative processes such as feature fusion, dynamic weighting, and dielectric correction, improves the accuracy and spatial resolution of organic pollution prediction in highly saline and alkaline environments, providing a reliable basis for precise remediation. This entire approach fully considers the specific characteristics of salinized soils and has significant practical value.

[0188] Figure 2 A structural diagram of a soil organic pollution concentration prediction system based on resistivity parameters provided in an embodiment of the present application is shown in FIG. Figure 2 As shown, the system includes:

[0189] The acquisition module 21 is used to obtain the dielectric constant distribution data, conductivity spatial distribution data, resistivity parameters and polarizability parameters of the soil in the target area in the organic pollution detection scenario in the saline-alkali soil environment.

[0190] The first generating module 22 is configured to generate a bimodal characteristic matrix based on the resistivity parameter and the polarizability parameter.

[0191] The second generating module 23 is configured to perform multi-source coupling correlation on the bimodal characteristic matrix, the dielectric constant distribution data, and the conductivity spatial distribution data to generate a dynamic data set.

[0192] The screening module 24 is used to perform frequency band screening on the dynamic data set through frequency domain dielectric spectrum analysis to extract the difference in polarization attenuation characteristics of organic pollutants and salt ions within a preset frequency band.

[0193] The output module 25 is used to output the spatial concentration distribution prediction results of organic pollutants in salinized soil environment based on the polarization attenuation characteristic difference and the dual-modal characteristic matrix using a trained machine learning model.

[0194] Figure 2 The soil organic pollution concentration prediction system based on resistivity parameters can be executed Figure 1 The implementation principles and technical effects of the resistivity-based soil organic contamination concentration prediction method described in the illustrated embodiment are not further elaborated. The specific manner in which the various modules and units in the resistivity-based soil organic contamination concentration prediction system described in the aforementioned embodiment perform their operations has been described in detail in the related embodiments of the method and will not be further elaborated here.

[0195] In one possible design, Figure 2 The soil organic pollution concentration prediction system based on resistivity parameters of the embodiment shown can be implemented as a computing device, such as Figure 3 As shown, the computing device may include a storage component 31 and a processing component 32;

[0196] The storage component 31 stores one or more computer instructions, wherein the one or more computer instructions are called and executed by the processing component 32 .

[0197] The processing component 32 is as follows Figure 1 The embodiment provides a method for predicting soil organic pollution concentration based on resistivity parameters.

[0198] The processing component 32 may include one or more processors to execute computer instructions to complete all or part of the steps in the above method. Of course, the processing component may also be implemented as one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the above method.

[0199] The storage component 31 is configured to store various types of data to support operations on the terminal. The storage component can be implemented by any type of volatile or non-volatile memory device, or a combination thereof, such as static random-access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk, or optical disk.

[0200] Of course, a computing device may also include other components, such as input / output interfaces, display components, communication components, etc.

[0201] The input / output interface provides an interface between the processing component and the peripheral interface module, which can be an output device, an input device, etc.

[0202] The communication component is configured to facilitate, among other things, wired or wireless communications between the computing device and other devices.

[0203] Among them, the computing device can be a physical device or an elastic computing host provided by a cloud computing platform, etc. In this case, the computing device can refer to a cloud server, and the above-mentioned processing components, storage components, etc. can be basic server resources rented or purchased from the cloud computing platform.

[0204] The present application also provides a computer storage medium storing a computer program, wherein the computer program can achieve the above-mentioned Figure 1 The embodiment shown is a method for predicting soil organic pollution concentration based on resistivity parameters.

[0205] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0206] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, i.e., they may be located in one location or distributed across multiple network units. Some or all of the modules may be selected based on actual needs to achieve the objectives of the present embodiment. Persons of ordinary skill in the art will be able to understand and implement the present invention without inventive effort.

[0207] Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented using software plus a necessary general-purpose hardware platform, or of course, hardware. Based on this understanding, the essence of the above technical solution, or the portion 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, a magnetic disk, or an optical disk, and includes a number of instructions for causing a computer device (such as a personal computer, server, or network device) to execute the methods described in each embodiment or certain portions of the embodiments.

[0208] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for predicting soil organic pollution concentration based on resistivity parameters, characterized in that: include: In the organic pollution detection scenario under saline-alkali soil environment, the dielectric constant distribution data, conductivity spatial distribution data, resistivity parameters and polarizability parameters of the soil in the target area are obtained; generating a bimodal characteristic matrix based on the resistivity parameter and the polarizability parameter; Performing multi-source coupling correlation on the bimodal feature matrix, the dielectric constant distribution data, and the conductivity spatial distribution data to generate a dynamic data set; Combined with the frequency domain dielectric spectrum analysis method, the dynamic data set is subjected to frequency band screening to extract the polarization attenuation characteristic differences between organic pollutants and salt ions within the preset frequency band; Based on the polarization attenuation characteristic differences and the bimodal characteristic matrix, a trained machine learning model is used to output prediction results of the spatial concentration distribution of organic pollutants in a saline-alkali soil environment; The frequency domain dielectric spectrum analysis method is combined to perform frequency band screening on the dynamic data set to extract the polarization attenuation characteristic differences between organic pollutants and salt ions within a preset frequency band, including: Converting the polarization response signal in the dynamic data set into frequency domain distribution data by frequency domain dielectric spectrum analysis; Based on the differences in polarization characteristics of organic pollutants and salt ions in saline-alkali soil environments, the boundary between the low-frequency band and the mid-frequency band within the preset frequency band is determined; Within the boundary interval, based on the frequency domain distribution data, frequency points whose signal strength first increases and then decreases sharply as the frequency increases are screened out to form a frequency point set; Extracting a characteristic frequency band corresponding to organic pollutants and an interference frequency band corresponding to salt ions from the frequency point set, calculating a signal attenuation slope difference between the characteristic frequency band and the interference frequency band, and using the signal attenuation slope difference as a polarization attenuation characteristic difference; The method outputs the spatial concentration distribution prediction results of organic pollutants in a saline-alkali soil environment using a trained machine learning model based on the polarization attenuation characteristic difference and the bimodal characteristic matrix, including: Combining each two-dimensional eigenvector in the dual-modal feature matrix with the polarization attenuation characteristic difference at a corresponding spatial position to generate an enhanced eigenvector; performing dynamic weighting processing on the enhanced feature vector according to a salt ion concentration gradient generated by a conductivity parameter in the conductivity spatial distribution data; The weighted enhanced feature vectors are input into the trained machine learning model according to the preset spatial distribution order to generate the initial pollution concentration value for each spatial location; Performing dielectric correction processing on the initial contamination concentration value based on the dielectric constant distribution data; The corrected pollution concentration value and the corresponding spatial position coordinates are associated to output the spatial concentration distribution prediction results.

2. The method according to claim 1, characterized in that The converting the polarization response signal in the dynamic data set into frequency domain distribution data by frequency domain dielectric spectrum analysis includes: extracting a polarization response signal from the dynamic data set; intercepting the polarization response signal according to a preset segment length to generate a plurality of time domain signal sequences; Perform frequency domain conversion processing on each time domain signal sequence to generate a target signal strength distribution sequence corresponding to each time domain signal sequence; determining a frequency range of the target signal intensity distribution sequence according to a proportional relationship between a conductivity parameter and the polarizability parameter in the conductivity spatial distribution data; The target signal intensity distribution sequences corresponding to all time domain signal sequences are superimposed to generate frequency domain distribution data covering the frequency band range.

3. The method according to claim 1, characterized in that The step of extracting a characteristic frequency band corresponding to organic pollutants and an interference frequency band corresponding to salt ions from the frequency point set, and calculating a signal attenuation slope difference between the characteristic frequency band and the interference frequency band, includes: According to the difference in the steep drop time sequence of the signal strength in the frequency point set, N consecutive frequency points in the frequency point set whose steep drop time sequence is located in the starting area of ​​the boundary interval are divided into an organic pollutant characteristic frequency band, and consecutive frequency points whose steep drop time sequence is located in the end area of ​​the boundary interval are divided into a salt ion interference frequency band, wherein N is greater than or equal to 3; In the characteristic frequency band of organic pollutants, a turning point where the signal strength changes from increasing to dropping sharply is selected as a first reference point, and a first analysis window is generated by extending a preset number of frequency points to both sides of the first reference point. In the salt ion interference frequency band, a frequency point at which the signal intensity drops sharply and exceeds a preset amplitude threshold is selected as a second reference point, and a second analysis window is generated by extending the same preset number of frequency points to both sides of the second reference point as the center; Calculate respectively a first average decreasing rate of the signal strength of the frequency points after the turning frequency point in the first analysis window and a second average decreasing rate of the signal strength of the frequency points after the turning frequency point in the second analysis window; The absolute value difference between the first average decreasing rate and the second average decreasing rate is used as the signal attenuation slope difference.

4. The method according to claim 2, characterized in that The performing frequency domain conversion processing on each time domain signal sequence to generate a target signal strength distribution sequence corresponding to each time domain signal sequence includes: Multiply each time domain signal sequence by a preset window function sequence point by point to obtain a time domain signal subsequence; Performing spectrum analysis on the time domain signal subsequence to generate an initial signal strength distribution sequence; Frequency units in the initial signal strength distribution sequence that are lower than a preset environmental noise threshold are set to zero to generate a target signal strength distribution sequence.

5. The method according to claim 1, wherein The generating of a bimodal characteristic matrix based on the resistivity parameter and the polarizability parameter includes: In the saline-alkali soil environment, synchronously aligning the resistivity parameter and the polarizability parameter at the same spatial position to generate a parameter pair set; Converting the resistivity parameter in each parameter pair in the parameter pair set into a first eigenvalue, and converting the polarizability parameter into a second eigenvalue; Combining the first eigenvalue and the second eigenvalue at the same spatial position into a two-dimensional eigenvector; The two-dimensional feature vectors of all spatial positions are arranged to form a bimodal feature matrix.

6. A soil organic pollution concentration prediction system based on resistivity parameters, characterized in that: include: The acquisition module is used to obtain the dielectric constant distribution data, conductivity spatial distribution data, resistivity parameters, and polarizability parameters of the soil in the target area in the organic pollution detection scenario under salinized soil environment; A first generating module is configured to generate a bimodal characteristic matrix based on the resistivity parameter and the polarizability parameter; A second generating module is configured to perform multi-source coupling correlation on the bimodal feature matrix, the dielectric constant distribution data, and the conductivity spatial distribution data to generate a dynamic data set; A screening module, configured to perform frequency band screening on the dynamic data set by frequency domain dielectric spectrum analysis to extract the polarization attenuation characteristic differences between organic pollutants and salt ions within a preset frequency band; An output module is used to output a prediction result of the spatial concentration distribution of organic pollutants in a salinized soil environment using a trained machine learning model based on the polarization attenuation characteristic difference and the bimodal characteristic matrix; Combined with frequency domain dielectric spectrum analysis, the dynamic data set is subjected to frequency band screening to extract the differences in polarization attenuation characteristics of organic pollutants and salt ions within the preset frequency band, including: Converting the polarization response signal in the dynamic data set into frequency domain distribution data by frequency domain dielectric spectrum analysis; Based on the differences in polarization characteristics of organic pollutants and salt ions in saline-alkali soil environments, the boundary between the low-frequency band and the mid-frequency band within the preset frequency band is determined; Within the boundary interval, based on the frequency domain distribution data, frequency points whose signal strength first increases and then decreases sharply as the frequency increases are screened out to form a frequency point set; Extracting a characteristic frequency band corresponding to organic pollutants and an interference frequency band corresponding to salt ions from the frequency point set, calculating a signal attenuation slope difference between the characteristic frequency band and the interference frequency band, and using the signal attenuation slope difference as a polarization attenuation characteristic difference; The method outputs the spatial concentration distribution prediction results of organic pollutants in a saline-alkali soil environment using a trained machine learning model based on the polarization attenuation characteristic difference and the bimodal characteristic matrix, including: Combining each two-dimensional eigenvector in the dual-modal feature matrix with the polarization attenuation characteristic difference at a corresponding spatial position to generate an enhanced eigenvector; performing dynamic weighting processing on the enhanced feature vector according to a salt ion concentration gradient generated by a conductivity parameter in the conductivity spatial distribution data; The weighted enhanced feature vectors are input into the trained machine learning model according to the preset spatial distribution order to generate the initial pollution concentration value for each spatial location; Performing dielectric correction processing on the initial contamination concentration value based on the dielectric constant distribution data; The corrected pollution concentration value and the corresponding spatial position coordinates are associated to output the spatial concentration distribution prediction results.

7. A computing device, characterized in that It comprises a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement a soil organic pollution concentration prediction method based on resistivity parameters as described in any one of claims 1 to 5.

8. A computer storage medium, characterized in that A computer program is stored, and when the computer program is executed by a computer, the method for predicting soil organic pollution concentration based on resistivity parameters according to any one of claims 1 to 5 is implemented.

Images