Rock ore specimen frequency spectrum induced polarization multi-parameter measurement method based on geophysical metal factors

Abstract

The invention discloses a rock and ore specimen spectrum induced polarization multi-parameter measurement method based on geophysical metal factors, and relates to the technical field of mineral parameter detection. Different types of rock and ore samples are selected; the method comprises the following steps: placing a rock ore sample in a measurement area of a spectrum induced polarization instrument, applying current under multiple frequencies, and gradually obtaining spectrum data of the sample; calculating to obtain a geophysical metal factor according to the spectrum data; analyzing the measured spectrum induced polarization multiple parameters to obtain the change conditions of the spectrum induced polarization multiple parameters of different types of rock ore samples; and distinguishing different rock types and stratigraphic units according to the change conditions of the spectrum induced polarization multi-parameters of different types of rock ore samples and the calculated value of the geophysical metal factor. According to the invention, more comprehensive spectrum data is obtained through spectrum induced polarization measurement; by constructing geophysical metal factors, the distribution and related characteristics of metal minerals in underground rocks are evaluated more accurately, and accurate distinguishing of different rock types and stratigraphic units is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of mineral parameter detection, and more specifically to a spectrum induced polarization multi-parameter measurement method of rock and ore samples based on geophysical metal factors. Background Art

[0002] With the development of geological science, the requirements for accurate identification of rock types and stratigraphic units and research on geological structural evolution are becoming increasingly higher. Different rock types and stratigraphic units have different physical and chemical properties, and their electrical properties are also different. Traditional rock classification and stratigraphic division methods may rely on macroscopic geological observations and limited physical parameter measurements, which are not effective in distinguishing rocks and strata with similar properties or in complex geological environments. Therefore, a method based on more comprehensive electrical parameter measurements and comprehensive analysis is needed to meet the needs of modern geological research for fine identification of rocks and strata.

[0003] Theoretical models such as the Cole-Cole model are widely used to explain the spectral induced polarization response of rocks and minerals, and are constantly being improved and expanded. Through these models, the measured spectral data can be linked to the electrical parameters of rocks and minerals (such as resistivity, polarizability, etc.), thereby inferring the structure and properties of the underground medium. Although the Cole-Cole model is a commonly used model to describe the spectral induced polarization response, it is an empirical model and may not accurately fit the actual measured data for some complex rock and mineral structures and mineral compositions. For example, when there are multiple conductive minerals in the rock and mineral and their distribution is uneven, the assumptions of the model may not be met, resulting in deviations between the calculated parameters and the actual situation.

[0004] Therefore, how to provide a spectral induced polarization multi-parameter measurement method for rock and mineral samples based on the geophysical metal factor, and construct a coefficient that can effectively evaluate the relationship between underground media and metal minerals, namely the geophysical metal factor, so as to more accurately evaluate the distribution of metal minerals in underground rocks is an urgent problem that technicians in this field need to solve. Summary of the invention

[0005] In view of this, the present invention provides a spectral induced polarization multi-parameter measurement method for rock and mineral samples based on geophysical metal factors. The electrical response of the rock and mineral at multiple frequencies is obtained through spectral induced polarization measurement to obtain more comprehensive spectral data, reflecting information such as the physical structure and mineral composition inside the rock and mineral; by constructing the geophysical metal factor, the distribution and related characteristics of metal minerals in underground rocks are more accurately evaluated; and by utilizing spectral induced polarization multi-parameter measurement and geophysical metal factor calculation, accurate distinction between different rock types and stratigraphic units is achieved.

[0006] In order to achieve the above-mentioned purpose, the present invention adopts the following technical scheme: a method for measuring multi-parameters of rock and ore samples by spectrum induced polarization based on geophysical metal factors, comprising:

[0007] Select different types of rock and mineral samples;

[0008] Place the rock sample in the measurement area of ​​the spectrum induced polarization instrument to ensure good electrode contact; apply current at multiple frequencies, record the voltage response at different frequencies, and gradually obtain the spectrum data of the sample;

[0009] Obtain spectrum induced polarization multi-parameters according to spectrum data;

[0010] Obtaining geophysical metal factor based on the spectrum induced polarization multi-parameter calculation;

[0011] The measured spectrum induced polarization multi-parameters are analyzed to obtain the changes of spectrum induced polarization multi-parameters of different types of rock and mineral samples;

[0012] Different rock types and stratigraphic units are distinguished based on the changes in multiple parameters of spectral induced polarization of different types of rock and mineral samples and the calculated values ​​of geophysical metal factors.

[0013] Preferably, the geophysical metal factor is obtained based on the spectrum induced polarization multi-parameter calculation, and the calculation formula of the geophysical metal factor is as follows:

[0014]

[0015] In the formula, δ s Represents the geophysical metal factor at a certain underground location s; M e is the chemical content of a metal element e at that position (in %, ppm or ppb, etc.), and takes the value of 1 when unknown; K e is the adjustment coefficient of the stratum or rock layer, which is taken as 1 when unknown; c is the frequency correlation coefficient of the Cole-Cole model, dimensionless; τ is the time constant of the Cole-Cole model (in seconds), m i is the polarizability (charging rate, in % or other units) of each time window, n is the number of time windows, and ρ is the resistivity of a certain position s underground. The geophysical metal factor is constructed based on geophysical parameters, combining geochemical element content and strata, lithology and other factors. It is a coefficient for evaluating underground media related to metal minerals.

[0016] Preferably, the spectral induced polarization multi-parameters include: resistivity at zero frequency, limiting polarizability, time constant and frequency correlation coefficient. The spectral induced polarization multi-parameters are Cole-Cole model parameters of rock specimens; these parameters can quantitatively describe the electrical properties of rocks and minerals.

[0017] Preferably, the spectrum induced polarization multi-parameters are obtained according to the spectrum data, including:

[0018] Preprocessing the spectrum data to ensure that there are no missing values ​​or abnormal values ​​in the spectrum data;

[0019] A Cole-Cole model is constructed, and the expression of the Cole-Cole model is:

[0020]

[0021] Where ρ0 is the resistivity at zero frequency, m is the limiting polarization rate (charging rate), τ and c are the time constant and frequency correlation coefficient, respectively, which are parameters used to evaluate induced polarization anomalies, and i is an imaginary unit;

[0022] The preprocessed spectrum data is substituted into the Cole-Cole model, and each model parameter is obtained by solving the nonlinear fitting algorithm.

[0023] Preferably, the model parameters are obtained by solving the nonlinear fitting algorithm, including:

[0024] Obtaining sample values ​​of the model parameters to be fitted according to the preprocessed spectrum data, and taking multiple groups of sample values ​​of the model parameters to be fitted as model parameter sample sets;

[0025] The model parameter sample set is used as the input of the Cole-Cole model, and the Cole-Cole model calculates a corresponding complex resistivity simulation set according to the model parameter sample set;

[0026] Setting neural network training parameters, taking the complex resistivity simulation set as the input of the neural network, taking the model parameter sample set as the output of the neural network, and training to obtain a parameter fitting model;

[0027] The measured actual value of complex resistivity is used as the input of the parameter fitting model to obtain the fitting value of the model parameter, and the fitting value of the model parameter is used as the actual value of the model parameter.

[0028] Preferably, the neural network training parameters include the number of hidden nodes of the neural network, a training target, a maximum number of training times, a training algorithm, an activation function, and a node transfer function.

[0029] Preferably, the measured spectral induced polarization multi-parameters are analyzed to obtain the changes of spectral induced polarization multi-parameters of different types of rock and ore samples, including:

[0030] For each rock or mineral type, select the spectral IP multi-parameters to be analyzed;

[0031] Draw a curve with frequency as the horizontal axis and each spectrum IP multi-parameter as the vertical axis;

[0032] Plot the same parameter-frequency curves of different types of rocks and minerals in the same coordinate system;

[0033] For each spectral IP multi-parameter of each rock and ore type, the mean value and standard deviation are calculated at different frequencies; the correlation between different spectral IP multi-parameters is analyzed, and the trend of each spectral IP multi-parameter changing with frequency is determined.

[0034] Through the above technical scheme, it can be known that compared with the prior art, the present invention discloses a method for measuring the multi-parameter spectrum induced polarization of rock and ore samples based on geophysical metal factors, which relates to the field of mineral parameter detection technology, including: selecting different types of rock and ore samples; placing the rock and ore samples in the measurement area of ​​the spectrum induced polarization instrument, applying current at multiple frequencies, recording the voltage response at different frequencies, and gradually obtaining the spectrum data of the samples; obtaining the spectrum induced polarization multi-parameters according to the spectrum data; obtaining the geophysical metal factor based on the spectrum induced polarization multi-parameter calculation; analyzing the measured spectrum induced polarization multi-parameters to obtain the changes in the spectrum induced polarization multi-parameters of different types of rock and ore samples; according to the changes in the spectrum induced polarization multi-parameters of different types of rock and ore samples and the calculated values ​​of the geophysical metal factors, different rock types and stratigraphic units are distinguished. The present invention obtains the electrical response of rock and ore at multiple frequencies through spectrum induced polarization measurement to obtain more comprehensive spectrum data. These data can reflect the physical structure and mineral composition inside the rock and ore, and can provide more details about the properties of the rock and ore compared to single frequency or simple electrical parameter measurement, which is helpful for in-depth study of the characteristics of the rock and ore and its relationship with metal minerals. By constructing a coefficient that can effectively evaluate the relationship between underground media and metal minerals, namely the geophysical metal factor, the distribution and related characteristics of metal minerals in underground rocks can be more accurately evaluated; using spectrum induced polarization multi-parameter measurement and geophysical metal factor calculation, accurate distinction between different rock types and stratigraphic units can be achieved. Combined with the geophysical metal factor, the characteristics of rock types and stratigraphic units can be identified more finely, solving the problem of inaccurate distinction between rock and mineral types and stratigraphic units under complex geological conditions in traditional methods. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0036] Figure 1A schematic flow chart of a method for spectral induced polarization multi-parameter measurement of rock and mineral samples based on geophysical metal factors provided by the present invention. DETAILED DESCRIPTION

[0037] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0038] The embodiment of the present invention discloses a spectral induced polarization multi-parameter measurement method for rock and mineral samples based on geophysical metal factors. The electrical response of the rock and mineral at multiple frequencies is obtained through spectral induced polarization measurement to obtain more comprehensive spectral data. These data can reflect the physical structure and mineral composition inside the rock and mineral, and can provide more details about the properties of the rock and mineral compared to single frequency or simple electrical parameter measurement, which is helpful for in-depth research on the characteristics of the rock and mineral and its relationship with metal minerals. By constructing a coefficient that can effectively evaluate the underground medium related to metal minerals, namely the geophysical metal factor, the distribution and related characteristics of metal minerals in underground rocks can be more accurately evaluated; by using spectral induced polarization multi-parameter measurement and geophysical metal factor calculation, accurate distinction between different rock types and stratigraphic units can be achieved. Combined with the geophysical metal factor, the characteristics of rock types and stratigraphic units can be identified more finely, solving the problem of inaccurate distinction between rock and mineral types and stratigraphic units under complex geological conditions in traditional methods.

[0039] In a specific embodiment of the present invention, a method for measuring multi-parameters of rock and ore samples using spectrum induced polarization based on geophysical metal factors is provided. Figure 1 As shown, including:

[0040] Select different types of rock and mineral samples; select representative rock samples, such as granite, mudstone, limestone, basalt, etc., and ensure that the physical and chemical properties of the samples are diverse.

[0041] Place the rock sample in the measurement area of ​​the spectrum induced polarization instrument to ensure good electrode contact; apply current at multiple frequencies, record the voltage response at different frequencies, and gradually obtain the spectrum data of the sample;

[0042] Obtain spectrum induced polarization multi-parameters according to spectrum data;

[0043] Based on the spectrum induced polarization multi-parameter calculation, the geophysical metal factor is obtained; and a coefficient capable of evaluating the relationship between underground media and metal minerals is constructed by comprehensively utilizing various data and parameters.

[0044] The measured spectrum induced polarization multi-parameters are analyzed to obtain the changes of spectrum induced polarization multi-parameters of different types of rock and mineral samples;

[0045] Different rock types and stratigraphic units are distinguished based on the changes in multiple parameters of spectral induced polarization of different types of rock and mineral samples and the calculated values ​​of geophysical metal factors.

[0046] Specifically, the geophysical metal factor is obtained based on the spectrum induced polarization multi-parameter calculation, and the calculation formula of the geophysical metal factor is as follows:

[0047]

[0048] In the formula, δ s Represents the geophysical metal factor at a certain underground location s; M e is the chemical content of a metal element e at that position (in %, ppm or ppb, etc.), and takes the value of 1 when unknown; K e is the adjustment coefficient of the stratum or rock layer, which is taken as 1 when unknown; c is the frequency correlation coefficient of the Cole-Cole model, dimensionless; τ is the time constant of the Cole-Cole model (in seconds), m i is the polarizability (charging rate, in % or other units) of each time window, n is the number of time windows, and ρ is the resistivity of a certain position s underground. The geophysical metal factor is constructed based on geophysical parameters, combining geochemical element content and strata, lithology and other factors. It is a coefficient for evaluating underground media related to metal minerals.

[0049] Specifically, the spectral induced polarization multi-parameters include: resistivity at zero frequency, limiting polarizability, time constant and frequency correlation coefficient. The measurement results of spectral induced polarization multi-parameters and geophysical metal factors are multi-solutions in geological interpretation. Even the same measurement data may correspond to a variety of different geological conditions due to the complexity and diversity of geological conditions. For example, high polarizability may be caused by the presence of metal minerals, but it may also be due to the influence of clay minerals or other conductive substances in the rock. Therefore, when conducting geological interpretation, it is necessary to comprehensively consider multiple factors. It is often difficult to accurately judge the rock type and stratigraphic unit based on a single parameter or indicator.

[0050] The spectral induced polarization multi-parameters are the Cole-Cole model parameters of the rock specimen: resistivity at zero frequency, limiting polarizability, time constant and frequency correlation coefficient; these parameters can quantitatively describe the electrical properties of the rock ores: the resistivity at zero frequency reflects the electrical conductivity of the rock ores, the limiting polarizability reflects the polarization degree of the rock ores under the action of the electric field, and the time constant and frequency correlation coefficient further reveal the relationship between the electrical response characteristics of the rock ores and the frequency.

[0051] Specifically, spectrum induced polarization multi-parameters are obtained according to the spectrum data, including:

[0052] Preprocess the spectrum data to ensure that there are no missing values ​​or outliers in the spectrum data. If there are missing values, fill them with interpolation methods; for outliers, clean or remove them. Data smoothing: Since there may be noise interference during the measurement process, smoothing the spectrum data can improve the quality of the data. Smoothing is performed using methods such as moving average or median filtering.

[0053] A Cole-Cole model is constructed, and the expression of the Cole-Cole model is:

[0054]

[0055] Where ρ0 is the resistivity at zero frequency, m is the limiting polarization rate (charging rate), τ and c are the time constant and frequency correlation coefficient, respectively, which are parameters used to evaluate induced polarization anomalies, and i is an imaginary unit;

[0056] The preprocessed spectrum data is substituted into the Cole-Cole model, and each model parameter is obtained by solving the nonlinear fitting algorithm.

[0057] Specifically, the model parameters are obtained by solving the nonlinear fitting algorithm, including:

[0058] Obtaining sample values ​​of the model parameters to be fitted according to the preprocessed spectrum data, and taking multiple groups of sample values ​​of the model parameters to be fitted as model parameter sample sets;

[0059] The model parameter sample set is used as the input of the Cole-Cole model, and the Cole-Cole model calculates a corresponding complex resistivity simulation set according to the model parameter sample set;

[0060] Setting neural network training parameters, taking the complex resistivity simulation set as the input of the neural network, taking the model parameter sample set as the output of the neural network, and training to obtain a parameter fitting model;

[0061] The measured actual value of complex resistivity is used as the input of the parameter fitting model to obtain the fitting value of the model parameter, and the fitting value of the model parameter is used as the actual value of the model parameter.

[0062] Specifically, the neural network training parameters include the number of hidden nodes of the neural network, training objectives, maximum number of training times, training algorithms, activation functions, and node transfer functions.

[0063] Specifically, a neural network model including an input layer, a hidden layer and an output layer is established, the complex resistivity simulation set is used as an input parameter for neural network training, and the corresponding sample set of model parameters to be fitted is used as an output parameter for artificial neural network training to train the network; including:

[0064] Set the number of hidden nodes of the neural network;

[0065] Training goal, let the objective function be

[0066]

[0067] Where N is the number of samples, C i is the fitted value of the model parameters during the i-th training, M i is the sample value of the model parameter to be fitted during the i-th training;

[0068] Set the maximum number of training times. When the number of training times reaches the maximum number of training times, stop training.

[0069] Select the optimization algorithm, set the activation function, and set the node transfer function.

[0070] The use of advanced spectrum induced polarization equipment and precise data processing methods can achieve high-precision measurements and parameter calculations. The measured spectrum data is carefully analyzed, and the Cole-Cole model parameters are solved through nonlinear fitting and other algorithms, reducing the impact of measurement errors and model fitting errors on the results. This makes the obtained rock and mineral electrical parameters more accurate and reliable, and can provide a high-quality data foundation for subsequent geological research and mineral exploration. When studying the microstructural changes of rocks and minerals or distinguishing between rock and mineral types with slight differences, high-precision parameters can provide more discerning information.

[0071] Specifically, the measured spectrum induced polarization multi-parameters are analyzed to obtain the changes of spectrum induced polarization multi-parameters of different types of rock and ore samples, including:

[0072] First, the measured spectral induced polarization multi-parameter data are classified according to the type of rock ores. For example, all sample data of granite are placed in one group, mudstone in another group, and so on. This makes it easier to analyze different rock types separately in the future.

[0073] Record the basic information of each rock and mineral sample, such as the collection location, depth, geological age, etc. These factors may affect the multi-parameters of spectral induced polarization.

[0074] For each type of rock or ore, select the spectrum IP multi-parameters to be analyzed, such as resistivity at zero frequency, limiting polarization, time constant and frequency correlation coefficient, etc. Determine the appropriate frequency range, which can reflect the range of changes in the electrical properties of the rock or ore.

[0075] With frequency as the horizontal axis and each spectrum IP multi-parameter as the vertical axis, a curve is drawn; for example, for the resistivity at zero frequency, a curve showing the resistivity of different rock and mineral samples changing with frequency is drawn. The changing trend of each parameter at different frequencies can be intuitively seen.

[0076] Plot the same parameter-frequency curves of different types of rocks and minerals in the same coordinate system for comparison. For example, put the resistivity-frequency curves of granite and mudstone together to observe their differences in shape, value range, change trend, etc.

[0077] For each spectral IP multi-parameter of each rock and mineral type, calculate the mean value and standard deviation at different frequencies; analyze the correlation between different spectral IP multi-parameters, and determine the trend of each spectral IP multi-parameter with frequency; the mean value can reflect the typical value of the parameter in this rock type, and the standard deviation can reflect the discrete degree of the parameter value. For example, calculate the mean value and standard deviation of the limiting polarizability of limestone samples at a certain frequency, and find the differences between them by comparing these statistics of different rock types. Study the correlation between different spectral IP multi-parameters. For example, analyze whether there is a linear or nonlinear relationship between resistivity and limiting polarizability at zero frequency. And quantify this relationship through correlation coefficients (such as Pearson correlation coefficient). The different correlations of different rock types further reveal the intrinsic connection between the electrical properties of rocks and minerals.

[0078] Some parameters change slowly at low frequencies and rapidly at high frequencies, while some parameters show monotonic changes or complex fluctuations over the entire frequency range. For example, the limiting polarizability may gradually decrease with increasing frequency in some rocks and minerals, and this trend helps identify the type of rock or mineral.

[0079] The embodiment of the present invention can obtain a wealth of electrical parameters, including resistivity at zero frequency, limiting polarizability, time constant, and frequency correlation coefficient, by performing spectral induced polarization measurement on rock and mineral samples at multiple frequencies. These parameters reflect the electrical properties of rocks and minerals from different angles, such as conductivity, degree of polarization, and response characteristics to electric fields of different frequencies. Compared with the traditional single electrical parameter measurement method, this multi-parameter measurement can more comprehensively describe the influence of the internal structure and mineral composition of rocks and minerals on their electrical properties. For example, zero-frequency resistivity can reflect the overall content and connectivity of conductive minerals in rocks and minerals, while limiting polarizability is more sensitive to the presence of polarizable minerals (such as metal minerals such as sulfides) therein.

[0080] Specifically, the correlation between different spectrum IP multi-parameters is analyzed, and the trend of each spectrum IP multi-parameter changing with frequency is determined, including:

[0081] (1) Determine whether the model parameters are monotonically increasing or monotonically decreasing with frequency; if the derivative of the fitting function is greater than zero in the entire frequency range, the model parameters are monotonically increasing; if the derivative is less than zero, the model parameters are monotonically decreasing.

[0082] (2) Analyze the speed of change of each model parameter through the derivative of the fitting function. The larger the absolute value of the derivative, the faster the model parameter changes within the frequency range. For example, for two types of rocks and minerals, although one of their parameters decreases monotonically with frequency, the change rate of the parameter of one type is faster, which reflects the difference in their internal mineral structure or composition.

[0083] (3) Analyze the asymptotic values ​​of each model parameter in the low-frequency and high-frequency limit cases. For example, some parameters tend to a constant at high frequencies.

[0084] In a specific embodiment of the present invention, different rock types and stratigraphic units are distinguished according to the changes in the spectrum induced polarization multi-parameters of different types of rock and ore samples and the calculated values ​​of geophysical metal factors, specifically including:

[0085] 1. Data Collection and Processing

[0086] Acquiring Spectral IP Multi-parameter Data

[0087] Spectral induced polarization measurements were performed on various selected rock and mineral samples to obtain multi-parameter data including resistivity at zero frequency, limiting polarization rate, time constant and frequency correlation coefficient. To ensure the accuracy and completeness of the data, abnormal data were processed and repeated measurements or data corrections were performed when necessary.

[0088] The multi-parameter data of each sample are recorded in detail according to the sample number, rock type, measurement frequency and other information to form the original data set.

[0089] Calculating Geophysical Metal Factors

[0090] According to the existing calculation formulas and methods, the geophysical metal factor of each rock and ore sample is calculated in combination with spectrum induced polarization multi-parameters.

[0091] The calculated geophysical metal factors are added to the data records of the corresponding samples.

[0092] 2. Feature Analysis and Extraction

[0093] Parametric statistical analysis

[0094] For each rock type, the spectrum IP multi-parameters are statistically analyzed at different frequencies. The mean, standard deviation, maximum, minimum and other statistical quantities of each parameter are calculated to facilitate the subsequent central tendency and dispersion degree of the parameter in different rock types.

[0095] For example, comparing the average zero-frequency resistivity values ​​of different rock types, it is found that some rock types have higher average values, while others have lower values. This can be used as a preliminary distinguishing feature.

[0096] Parameter change trend analysis

[0097] Observe the trend of the multi-parameters of each spectrum changing with frequency. The parameter change curves of different rock types have different shapes and characteristics.

[0098] For example, the polarizability of one type of rock may gradually decrease with increasing frequency, and the rate of decrease is relatively stable; while the polarizability of another type of rock may fluctuate or change suddenly within a certain frequency range. This change trend can serve as an important basis for distinguishing different rock types.

[0099] Analysis of Geophysical Metal Factor Characteristics

[0100] Study the distribution of geophysical metal factors in different rock types, and analyze their average value, range, and relationship with other parameters. Some rock types may have unique value ranges or variation patterns of geophysical metal factors due to their richness in specific metal minerals or specific geological structures.

[0101] 3. Establish a classification model or standard

[0102] Threshold-based classification

[0103] Based on the results of parameter statistical analysis and characteristic analysis, the threshold or range of key parameters is determined to preliminarily distinguish different rock types and stratigraphic units.

[0104] Multivariate statistical analysis methods

[0105] Cluster analysis method is used to comprehensively analyze spectral induced polarization multi-parameters and geophysical metal factors.

[0106] Cluster analysis groups samples with similar parameter characteristics into one category. By continuously adjusting the parameters and methods of the clustering algorithm, the clustering results are matched with the known rock types and stratigraphic unit information, thus establishing a clustering-based classification standard.

[0107] 4. Verification and Optimization

[0108] Cross Validation

[0109] Use the cross-validation method to evaluate the accuracy and reliability of a classification model or standard. Randomly divide the sample data into several groups, one of which is used as a test set and the rest as training sets. Repeat this process multiple times, use different groups for training and testing, and calculate evaluation indicators such as average accuracy, recall, and F1 value.

[0110] Comparison and verification with geological data

[0111] Verify the classification results by comparing them with known geological data. For example, check whether the classification results are consistent with the actual rock types and stratigraphic unit distribution according to geological exploration reports or geological maps.

[0112] Model Optimization

[0113] According to the verification results, the classification model or standard is optimized. The algorithm parameters are adjusted, feature variables are added, or data preprocessing methods are improved. The verification and optimization process is repeated until a satisfactory classification effect is achieved.

[0114] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0115] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for measuring multi-parameters of rock and ore samples using spectrum induced polarization based on geophysical metal factors, characterized in that: include: Select different types of rock and mineral samples; Place the rock sample in the measurement area of ​​the spectrum induced polarization instrument, apply current at multiple frequencies, record the voltage response at different frequencies, and gradually obtain the spectrum data of the sample; Obtain spectrum induced polarization multi-parameters according to spectrum data; Obtaining geophysical metal factor based on the spectrum induced polarization multi-parameter calculation; The measured spectrum induced polarization multi-parameters are analyzed to obtain the changes of spectrum induced polarization multi-parameters of different types of rock and mineral samples; Different rock types and stratigraphic units are distinguished based on the changes in multiple parameters of spectral induced polarization of different types of rock and mineral samples and the calculated values ​​of geophysical metal factors.

2. The method for measuring multi-parameters of rock and ore samples by spectrum induced polarization based on geophysical metal factors according to claim 1 is characterized in that: The geophysical metal factor is obtained based on the spectrum induced polarization multi-parameter calculation, and the calculation formula of the geophysical metal factor is as follows: In the formula, δ s Represents the geophysical metal factor at a certain underground location s; M e is the chemical content of a metal element e at that location; K e is the adjustment coefficient; c is the frequency correlation coefficient of the cole-cole model; τ is the time constant of the cole-cole model, m i is the polarizability of each time window, n is the number of time windows, and ρ is the resistivity at a location s in the subsurface.

3. The method for measuring multi-parameters of rock and ore samples by spectrum induced polarization based on geophysical metal factors according to claim 1, characterized in that: The spectrum induced polarization multi-parameters include: resistivity at zero frequency, limiting polarization rate, time constant and frequency correlation coefficient.

4. The method for measuring multi-parameters of rock and ore samples by spectrum induced polarization based on geophysical metal factors according to claim 1 is characterized in that: According to the spectrum data, multiple spectrum induced polarization parameters are obtained, including: Preprocessing the spectrum data; A Cole-Cole model is constructed, and the expression of the Cole-Cole model is: Where ρ0 is the resistivity at zero frequency, m is the limiting polarizability, τ and c are the time constant and frequency-dependent coefficient, respectively, and i is the imaginary unit; The preprocessed spectrum data is substituted into the Cole-Cole model, and each model parameter is obtained by solving the nonlinear fitting algorithm.

5. The method for measuring multi-parameters of rock and ore samples by spectrum induced polarization based on geophysical metal factors according to claim 4 is characterized in that: The model parameters are obtained by solving the nonlinear fitting algorithm, including: Obtaining sample values ​​of the model parameters to be fitted according to the preprocessed spectrum data, and taking multiple groups of sample values ​​of the model parameters to be fitted as model parameter sample sets; The model parameter sample set is used as the input of the Cole-Cole model, and the Cole-Cole model calculates a corresponding complex resistivity simulation set according to the model parameter sample set; Setting neural network training parameters, taking the complex resistivity simulation set as the input of the neural network, taking the model parameter sample set as the output of the neural network, and training to obtain a parameter fitting model; The measured actual value of complex resistivity is used as the input of the parameter fitting model to obtain the fitting value of the model parameter, and the fitting value of the model parameter is used as the actual value of the model parameter.

6. The method for measuring multi-parameters of rock and ore samples by spectrum induced polarization based on geophysical metal factors according to claim 5, characterized in that: The neural network training parameters include the number of hidden nodes of the neural network, training objectives, maximum training times, training algorithms, activation functions and node transfer functions.

7. The method for measuring multi-parameters of rock and ore samples by spectrum induced polarization based on geophysical metal factors according to claim 1 is characterized in that: The measured spectrum induced polarization multi-parameters are analyzed to obtain the changes of spectrum induced polarization multi-parameters of different types of rock and ore samples, including: For each rock or mineral type, select the spectral IP multi-parameters to be analyzed; Draw a curve with frequency as the horizontal axis and each spectrum IP multi-parameter as the vertical axis; Plot the same parameter-frequency curves of different types of rocks and minerals in the same coordinate system; For each spectral IP multi-parameter of each rock and ore type, the mean value and standard deviation are calculated at different frequencies; the correlation between different spectral IP multi-parameters is analyzed, and the trend of each spectral IP multi-parameter changing with frequency is determined.

Images