Regional lithology interpretation method, system, device and medium based on multi-property characteristics

Abstract

The application discloses a regional lithology interpretation method, system, device and medium based on multiple physical characteristics, and the method comprises the following steps: acquiring regional logging data and geophysical data; preprocessing the regional logging data to obtain a target regional logging data set; establishing multiple classification models associated with the input of multiple physical characteristics; training and verifying the multiple classification models by using the target regional logging data set to determine the best classification model; performing multidimensional inversion on the geophysical data to obtain an inversion result; and inputting the inversion result into the best classification model for processing to obtain a lithology interpretation result. Therefore, by establishing multiple classification models associated with the input of multiple physical characteristics, and determining the model with the optimal classification effect, i.e. the best classification model, from the multiple classification models, the best classification model can be applied to the lithology interpretation of the geophysical inversion result, so that the lithology interpretation result with relatively better accuracy, efficiency and intelligence can be obtained.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of geological exploration, and in particular to a regional lithology interpretation method, system, device and medium based on multi-property characteristics. BACKGROUND

[0002] In the existing oil geological exploration process, lithology interpretation is given by geologists through field reconnaissance, and reference is made to geophysical exploration results and logging results, which depends on the experience and subjective judgment of geologists, is low in efficiency, and is different in different regional geological conditions. In some newly developed areas without prior information, accurate geological and lithological interpretation cannot be quickly given. Traditional lithology prediction is based on logging data with existing lithology annotation, and the logging data of unannotated lithology of the same well is predicted, that is, one-dimensional prediction, and a multi-dimensional lithology and facies interpretation method with intelligence, such as two-dimensional or three-dimensional, has not been developed. SUMMARY

[0003] The present application aims to at least solve one of the technical problems existing in the prior art. To this end, the present application provides a regional lithology interpretation method based on multi-property characteristics, which solves the problems of relatively low accuracy, efficiency and intelligence of the current traditional regional lithology interpretation method.

[0004] The present application also provides a regional lithology interpretation system based on multi-property characteristics, a regional lithology interpretation device based on multi-property characteristics and a computer readable storage medium.

[0005] The regional lithology interpretation method based on multi-property characteristics according to the first aspect of the present application comprises the following steps:

[0006] Obtaining regional logging data and geophysical data;

[0007] Preprocessing the regional logging data to obtain a target regional logging data set, the target regional logging data set having multi-property characteristics;

[0008] Establishing a plurality of classification models associated with the input of the multi-property characteristics;

[0009] Training and verifying a plurality of the classification models using the target regional logging data set to determine the best classification model, the best classification model being the model with the best classification effect in the plurality of the classification models;

[0010] Performing multi-dimensional inversion on the geophysical data to obtain an inversion result;

[0011] Inputting the inversion result into the best classification model for processing to obtain a lithology interpretation result.

[0012] The data-driven multi-property feature-based regional lithology interpretation method according to the embodiments of the present application has at least the following beneficial effects:

[0013] By establishing multiple classification models associated with the input of the multi-property features and determining the model with the optimal verification classification effect, i.e., the best classification model, to be applied to the lithology interpretation of the geophysical inversion result, a lithology interpretation result with relatively better accuracy, efficiency and intelligence can be obtained. Therefore, for the method of the embodiments of the present application, by using multiple property intersection features as model inputs, the multi-solution property of the lithology interpretation result is greatly reduced; and a data set is established based on the logging data of the target region, so that the intelligent classification model trained has strong adaptability to the target region and does not rely on human experience; and multiple mathematical or statistical models can be selected as the intelligent classification model. Since the data characteristics of the target region are different, the classifier effect is also different, so multiple classification models are trained synchronously and the best model is selected to ensure the accuracy of the lithology interpretation result; and based on the multi-dimensional inversion result, multi-dimensional intelligent lithology interpretation is realized.

[0014] According to some embodiments of the present application, the preprocessing of the regional logging data to obtain the target regional logging data set comprises the following steps:

[0015] The regional logging data is subjected to principal component analysis, and outlier elimination processing and abnormal data recovery processing are performed according to the lithology label, the principal component condition and the clustering condition to obtain the target regional logging data set.

[0016] According to some embodiments of the present application, the outlier elimination processing is implemented by using a DBscan algorithm.

[0017] According to some embodiments of the present application, the establishment of multiple classification models associated with the input of the multi-property features comprises the following steps:

[0018] Based on the multi-class property intersection feature analysis of the regional logging data, multiple classification models are established, and the multiple classification models at least include a probabilistic neural network model, a support vector machine model and a K-nearest neighbor algorithm model. Each of the classification models at least uses the resistivity, velocity, density and magnetic susceptibility of the core as model input and uses the lithology as model output.

[0019] According to some embodiments of the present application, the training and verification of multiple classification models using the target regional logging data set to determine the best classification model comprises the following steps:

[0020] The target regional logging data set is divided into a training set and a test set.

[0021] Based on a cross-validation method, the training set is used to train a plurality of classification models respectively, and the test set is used to verify the classification effect of the trained plurality of classification models respectively, so as to determine the best classification model.

[0022] According to some embodiments of the present application, the cross-validation method adopts a K-fold cross-validation method.

[0023] According to some embodiments of the present application, the multi-dimensional inversion of the geophysical data to obtain the inversion result comprises the following steps:

[0024] At least three-dimensional inversion is performed on gravity data, seismic data, electromagnetic data in the geophysical data to obtain the resistivity distribution, velocity distribution, density distribution, and magnetic susceptibility distribution of the core as a preliminary inversion result.

[0025] The preliminary inversion result is calibrated to obtain the final inversion result.

[0026] According to the second aspect of the present application, the regional lithology interpretation system based on multi-property characteristics comprises:

[0027] A data acquisition module is configured to acquire regional logging data and geophysical data.

[0028] A preprocessing module is configured to preprocess the regional logging data to obtain a target regional logging data set, wherein the target regional logging data set has multi-property characteristics.

[0029] A model construction module is configured to construct a plurality of classification models associated with the input of the multi-property characteristics.

[0030] A model training and verification module is configured to train and verify a plurality of classification models using the target regional logging data set to determine the best classification model, wherein the best classification model is the model with the best verified classification effect among the plurality of classification models.

[0031] An inversion module is configured to perform multi-dimensional inversion on the geophysical data to obtain an inversion result.

[0032] An interpretation module is configured to input the inversion result into the best classification model for processing to obtain a lithology interpretation result.

[0033] The data-driven regional lithology interpretation system based on multi-property characteristics according to the embodiments of the present application has at least the following beneficial effects:

[0034] By establishing a plurality of classification models associated with a plurality of input of multi-property characteristics, and determining the model with the optimal verification classification effect, that is, the optimal classification model, to be applied to the lithology interpretation of the geophysical inversion result, a relatively better lithology interpretation result in accuracy, efficiency and intelligence can be obtained. Therefore, for the system of the embodiment of the present application, by using a plurality of physical property intersection characteristics as model input, the multi-solution of the lithology interpretation result is greatly reduced; and a data set is established based on the logging data of the target area, so that the intelligent classification model trained has strong adaptability to the target area and does not depend on artificial experience judgment; and a plurality of mathematical or statistical models can be selected as the intelligent classification model. Since the data characteristics of the target area are different, the classifier effect is also different, so a plurality of classification models are trained synchronously and the optimal model is selected to ensure the accuracy of the lithology interpretation result; and based on the multi-dimensional inversion result, multi-dimensional lithology intelligent interpretation is realized.

[0035] The device for regional lithology interpretation based on multi-property characteristics according to the third aspect of the embodiment of the present application comprises at least one control processor and a memory connected in communication with the at least one control processor; the memory stores instructions executable by the at least one control processor, and the instructions are executed by the at least one control processor to enable the at least one control processor to perform the method for regional lithology interpretation based on multi-property characteristics according to any one of the first aspect of the embodiment of the present application.

[0036] The computer readable storage medium according to the fourth aspect of the embodiment of the present application stores computer executable instructions for causing a computer to perform the method for data-driven regional lithology interpretation based on multi-property characteristics according to any one of the first aspect of the embodiment of the present application.

[0037] It can be understood that the beneficial effects of the above third aspect and fourth aspect compared with the related art are the same as the beneficial effects of the above first aspect compared with the related art, which can be referred to the related description in the first aspect and will not be described here.

[0038] Other features and advantages of the present application will be set forth in the following description, and in part will become apparent from the description, or can be learned by practice of the present application. BRIEF DESCRIPTION OF DRAWINGS

[0039] The above and / or additional aspects and advantages of the present application will become apparent and be readily appreciated from the following description, taken in conjunction with the accompanying drawings, in which:

[0040] Figure 1 is a flowchart of the method for regional lithology interpretation based on multi-property characteristics according to an embodiment of the present application;

[0041] Figure 2 is a schematic diagram of a regional lithology interpretation system based on multi-property features of an embodiment of the present application.

[0042] Reference signs:

[0043] a data acquisition module 100;

[0044] a preprocessing module 200;

[0045] a model construction module 300;

[0046] a model training and verification module 400;

[0047] an inversion module 500;

[0048] an interpretation module 600. DETAILED DESCRIPTION

[0049] Embodiments of the present application will be described in detail below, examples of which are shown in the drawings, wherein the same or similar reference signs represent the same or similar elements or elements having the same or similar functions throughout. The embodiments described below by referring to the drawings are exemplary and are only used to explain the present application, and cannot be understood as a limitation of the present application.

[0050] In the description of the present application, if the first, second, etc. are described, it is only for the purpose of distinguishing technical features, and cannot be understood as indicating or implying the relative importance of the indicated technical features or implicitly indicating the number of the indicated technical features or the sequence of the indicated technical features.

[0051] In the description of the present application, it should be understood that the orientation description, such as the orientation or position relationship indicated by up, down, etc. is based on the orientation or position relationship shown in the drawings, only for the convenience of describing the present application and simplifying the description, and is not intended to indicate or imply that the device or element indicated must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation of the present application.

[0052] In the description of the present application, it should be noted that, unless otherwise explicitly limited, the words such as setting, installing, connecting, etc. should be broadly understood, and those skilled in the art can reasonably determine the specific meaning of the above words in the present application in combination with the specific content of the technical solution.

[0053] The technical solutions of the present application will be described in detail below in combination with the drawings. Obviously, the following described embodiments are part of the embodiments of the present application, not all embodiments.

[0054] Referring to Figure 1 shown, a flowchart of a regional lithology interpretation method based on multi-property features provided by an embodiment of the present application, the method comprising the following steps:

[0055] Acquire regional well logging data and geophysical data;

[0056] Preprocessing of regional logging data to obtain target region logging dataset, which has multiple physical property characteristics;

[0057] Establish multiple classification models that are associated with multiple physical property features;

[0058] Multiple classification models were trained and validated using the target area logging dataset to determine the best classification model, which was the model with the best classification performance among the multiple models.

[0059] Multidimensional inversion is performed on geophysical data to obtain inversion results;

[0060] The inversion results are input into the optimal classification model for processing to obtain lithological interpretation results.

[0061] Specifically, such as Figure 1 As shown, it is understandable that the first step is data collection and acquisition, specifically including regional well logging data and geophysical data. Regional well logging data refers to relevant data collected in a specific area using well logging technology. Well logging, also known as geophysical logging, is a method of measuring geophysical parameters by utilizing the electrochemical, electrical, acoustic, and radioactive properties of rock formations; it belongs to one of the applied geophysical methods. Geophysical data refers to data obtained through the observation and measurement of geophysical phenomena. It helps to understand the Earth's internal structure and physical properties, providing important evidence and support for Earth science research. Geophysical data mainly includes seismic data, geomagnetic data, gravity data, and electromagnetic data.

[0062] Furthermore, it can be understood that the acquired data is preprocessed to obtain a well logging dataset for the target area, which serves as preparation material for model training and testing. Then, based on the multi-physical property characteristics of the target area well logging dataset, multiple classification models associated with these characteristics are constructed. Each classification model is then trained and validated using the target area well logging dataset, thereby determining the model with the best validated classification performance as the optimal classification model. Further, multi-dimensional inversion is performed on the acquired geophysical data, and the inversion results are applied to the optimal classification model for lithological interpretation, thus achieving regional lithological interpretation based on multi-physical property characteristics.

[0063] In the embodiment, a plurality of classification models associated with the input of the multi-property characteristics are established, and the model with the optimal verification classification effect, i.e., the optimal classification model, is determined from the plurality of classification models to be applied to the lithology interpretation of the geophysical inversion result, so that the lithology interpretation result with relatively better accuracy, efficiency and intelligence can be obtained. Therefore, for the method of the embodiment, the multi-solution of the lithology interpretation result is greatly reduced by using the multi-property intersection characteristics as the model input; and the data set is established based on the logging data of the target region, so that the intelligent classification model trained has strong adaptability to the target region and does not depend on artificial experience judgment; and a plurality of mathematical or statistical models can be selected as the intelligent classification model. Since the data characteristics of the target region are different, the effects of the classifiers are also different, so a plurality of classification models are trained synchronously and the optimal model is selected to ensure the accuracy of the lithology interpretation result; and the multi-dimensional lithology intelligent interpretation is realized based on the multi-dimensional inversion result.

[0064] In some embodiments, the regional logging data is preprocessed to obtain a target regional logging data set, including the following steps:

[0065] The regional logging data is subjected to principal component analysis, and is subjected to outlier elimination processing and abnormal data recovery processing according to the lithology label, the principal component condition and the clustering condition, to obtain the target regional logging data set.

[0066] Specifically, it can be understood that first, the rock core physical property measurement and analysis of the drilling wells in the exploration area are carried out in the regional logging, the density, resistivity, seismic wave velocity and the like of each specimen are measured, and then the lithology label, the principal component condition and the clustering condition obtained after the measurement and analysis are used for data outlier deletion, abnormal data recovery processing, data set division and the like preprocessing work, so that the target regional logging data set is finally obtained.

[0067] In some embodiments, the outlier elimination processing is implemented by using the DBscan algorithm.

[0068] Specifically, it can be understood that for the logging data statistics and analysis, since the clustering shape of most logging data does not present a central clustering shape, in some embodiments, the DBscan algorithm is used to calculate and delete outliers, and finally the data is arranged into a training set and a test set. The DBscan (Density-Based Spatial Clustering of Applications with Noise) is a density-based clustering algorithm, which is different from the K-means and hierarchical clustering methods of other embodiments. It can find clusters of any shape and can identify noise points. The basic principle is to use the density of samples in the data space as the standard for distinguishing the clustering boundary and noise.

[0069] In some embodiments, establishing a plurality of classification models associated with the input of the multi-property feature includes the following steps:

[0070] Based on the multi-property intersection feature analysis of the regional logging data, a plurality of classification models are established, including at least a probabilistic neural network model, a support vector machine model, and a K-neighbor algorithm model. Each classification model uses at least the resistivity, velocity, density, and magnetic susceptibility of the core as model input and uses lithology as model output.

[0071] Specifically, it can be understood that due to different target region data features, different classifiers have different effects. This technology simultaneously trains multiple classifiers and selects the best one as an intelligent interpretation model. Commonly used classification models include probabilistic neural network (PNN), support vector machine (SVM), and K-neighbor (KNN) classifiers.

[0072] Specifically, the probabilistic neural network used for lithology classification adopts a four-layer structure, including an input layer, a sample layer, a summation layer, and a competitive output layer. The main function of the sample layer is to perform weighted summation operation on the input signal and send it to the next layer after passing through an activation function, and the activation function is a Gaussian function:

[0073]

[0074] Where x is the input, θ i is the output, c i is the center of the radial basis function, and σ i is the standard deviation of the Gaussian function; a larger σ will make the function wider, resulting in a smoother model; and a smaller σ will make the function narrower, making the model pay more attention to the details of the data.

[0075] Further, support vector machine (SVM) is a machine learning algorithm based on geometric boundaries. Its main purpose is to find a hyperplane to maximize the interval between the two classes. For the lithology classification problem, a Gaussian radial basis function (RBF) is used as the kernel function of the support vector machine.

[0076] Further, K-neighbor is a classification algorithm based on the proximity of input samples in feature space to make decisions. For the lithology classification problem of multiple properties, since KNN is a distance-based algorithm, normalization of different properties is required first. The best k value is obtained through a search algorithm to achieve the best classification effect.

[0077] In some embodiments, the target region logging data set is used to train and verify the plurality of classification models to determine the best classification model, including the following steps:

[0078] The target region logging data set is divided into a training set and a test set;

[0079] Based on the cross-validation method, the plurality of classification models are trained respectively by using the training set, and the classification effects of the plurality of trained classification models are verified respectively by using the test set, so as to determine the best classification model.

[0080] Specifically, it can be understood that cross-validation is a statistical method for estimating the performance of a machine learning model, which is a method for evaluating how the results of statistical analysis are generalized to independent data sets. The basic idea of cross-validation is to group the original data in some sense, one part as the training set and the other part as the validation set (test set), first train the classifier with the training set, and then test the trained model with the validation set, which is used as the performance index of the classifier. The purpose of this is to reduce the problem of model overfitting caused by unreasonable division of a single data set. In model selection, cross-validation is a means to avoid overfitting and a method to solve overfitting, so the best generalization performance model can be selected from multiple candidate models.

[0081] In some embodiments, the cross-validation method adopts a K-fold cross-validation method.

[0082] Specifically, it can be understood that K-fold cross-validation is a method for evaluating the performance of a machine learning model. The core idea is to divide the original data set into K non-overlapping subsets (or called "folds"), and then repeatedly use these subsets for model training and validation. By using this method, the performance of the model can be evaluated multiple times, and a better estimate of the stability of the model can be obtained.

[0083] In some other embodiments, the cross-validation method can also adopt a Hold Out cross-validation method, a Leave One Out cross-validation method, etc.

[0084] In some embodiments, the geophysical data is subjected to multi-dimensional inversion to obtain an inversion result, including the following steps:

[0085] At least the gravity data, seismic data, electromagnetic data in the geophysical data are subjected to three-dimensional inversion to obtain the resistivity distribution, velocity distribution, density distribution, and magnetic susceptibility distribution of the core as a preliminary inversion result.

[0086] The preliminary inversion result is calibrated to obtain a final inversion result.

[0087] Specifically, it can be understood that geophysical inversion is a method of inferring the distribution of underground physical properties from geophysical observation data. Its main idea is that the geophysical response generated by the distribution of underground physical properties (such as resistivity, density, wave velocity, etc.) can be measured at the surface or somewhere underground, and then the inverse process is used to solve these physical properties. The electromagnetic, gravity, and seismic inversion results correspond to the underground resistivity, density, and wave velocity distribution, respectively.

[0088] Further, using the best classification model trained by the embodiment of the present application, the resistivity, density, and wave velocity of the well logging core can be used as input to obtain the classification result of the lithology, and then the three-dimensional inversion results of the electromagnetic, gravity, and seismic can be used as input to obtain the three-dimensional lithology interpretation.

[0089] Further, it can be understood that due to the difference in regularization and initial model selection of the inversion process, the preliminary inversion result may have errors compared with the actual situation of the stratum, so the same depth logging data is used to calibrate the inversion result, and a fixed coefficient is used to correct the overall inversion result, so that the numerical value of the inversion area intersecting the well logging is the same as the well logging value. Specifically, the related mathematical model is as follows:

[0090] M * =M·P,

[0091] wherein M * and M represent the inversion results after and before calibration respectively, and the parameter P is defined as follows:

[0092]

[0093] wherein N well and N M represent the well logging value and the inversion value at the intersection of the well logging and the inversion data, respectively.

[0094] In addition, as Figure 2As shown, the embodiment of the present application also provides a regional lithology interpretation system based on multi-property characteristics, comprising: a data acquisition module 100, a preprocessing module 200, a model construction module 300, a model training and verification module 400, an inversion module 500, and an interpretation module 600. The data acquisition module 100 is configured to acquire regional logging data and geophysical data; the preprocessing module 200 is configured to preprocess the regional logging data to obtain a target regional logging data set, the target regional logging data set having multi-property characteristics; the model construction module 300 is configured to establish a plurality of classification models associated with multi-property characteristic inputs; the model training and verification module 400 is configured to train and verify the plurality of classification models using the target regional logging data set to determine an optimal classification model, the optimal classification model being a model with the best verified classification effect among the plurality of classification models; the inversion module 500 is configured to perform multi-dimensional inversion on the geophysical data to obtain an inversion result; and the interpretation module 600 is configured to input the inversion result into the optimal classification model for processing to obtain a lithology interpretation result.

[0095] Specifically, with reference to Figure 2 It can be understood that the regional lithology interpretation system based on multi-property characteristics of the embodiment of the present application is used to implement the regional lithology interpretation method based on multi-property characteristics, and the regional lithology interpretation system based on multi-property characteristics of the embodiment of the present application corresponds to the aforementioned regional lithology interpretation method based on multi-property characteristics. For details of the processing process, please refer to the aforementioned regional lithology interpretation method based on multi-property characteristics, which will not be described here.

[0096] In the embodiment, a plurality of classification models associated with multi-property characteristic inputs are established, and a model with the best verified classification effect, i.e., an optimal classification model, is determined therefrom, to be applied to lithology interpretation of geophysical inversion results, so that a lithology interpretation result with relatively better accuracy, efficiency and intelligence can be obtained. Therefore, for the system of the embodiment of the present application, a plurality of property intersection characteristics are used as model inputs, so that the multi-solution property of the lithology interpretation result is greatly reduced; a data set is established based on target regional logging data, so that the intelligent classification model trained has strong target regional adaptability and does not rely on artificial experience judgment; and a plurality of mathematical or statistical models can be selected as the intelligent classification model. Since the target regional data characteristics are different, the classifier effect is also different, so a plurality of classification models are simultaneously trained and the best model is selected to ensure the accuracy of the lithology interpretation result; and multi-dimensional lithology intelligent interpretation is realized based on multi-dimensional inversion results.

[0097] In addition, the embodiment of the present application also provides a regional lithology interpretation device based on multi-property characteristics, comprising: at least one control processor and a memory in communication connection with the at least one control processor.

[0098] The memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer-executable programs. In addition, the memory can include high-speed random access memory and can also include non-transitory memory, such as at least one magnetic disk storage device, a flash memory device, or other non-transitory solid-state memory device. In some embodiments, the memory can optionally include remote memory that is disposed remotely from the processor, and the remote memory can be connected to the processor through a network. Examples of the network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof.

[0099] The non-transitory software programs and instructions required to implement the above-described embodiments of the method for regional lithology interpretation based on multi-property features are stored in the memory, and when executed by the processor, perform the method for regional lithology interpretation based on multi-property features described in the above embodiments, for example, perform the method described in the above Figure 1 .

[0100] The system embodiments described above are only schematic, and units described as separate units can or can not be physically separate, i.e., can be located in one place or distributed over multiple network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the present embodiment.

[0101] In addition, the present embodiment also provides a computer-readable storage medium, which stores computer-executable instructions, and the computer-executable instructions are executed by one or more control processors, so that the one or more control processors execute the method for regional lithology interpretation based on multi-property features described in the above method embodiments, for example, execute the method described in the above Figure 1 .

[0102] As will be appreciated by one of ordinary skill in the art, all or some of the steps, systems, and techniques disclosed herein can be embodied in software, firmware, hardware, or any suitable combination thereof. Some or all of the physical components can be implemented as software executed by a processor, such as a central processing unit, a digital signal processor, or a microprocessor, or as hardware, or as an integrated circuit, such as an application- specific integrated circuit. Such software can be distributed on computer readable media, which can comprise computer storage media (or non-transitory media), and communication media (or transitory media). As will be appreciated by one of ordinary skill in the art, the term computer storage media includes all physical and tangible computer storage media, such as a volatile and non-volatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data. Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium which can be used to store the desired information and which can be accessed by a computer. Further, as will be appreciated by one skilled in the art, communication media typically embodies computer readable instructions, data structures, program modules or other data in a modulated data signal such as a carrier wave or other transport mechanism and includes any information delivery media.

[0103] The embodiments of the present application disclosed above are only used to explain the principle of the present application, and the present application is not limited to the above embodiments. Various changes can be made by those skilled in the art without departing from the spirit of the present application.

Claims

1. A method for regional lithology interpretation based on multi-property features, characterized in that, The method comprises the following steps: obtaining regional logging data and geophysical data; preprocessing the regional logging data to obtain a target regional logging data set, the target regional logging data set having a plurality of physical characteristics; establishing a plurality of classification models associated with the input of the plurality of physical characteristics; training and verifying the plurality of classification models using the target regional logging data set to determine an optimal classification model, the optimal classification model being the model with the optimal classification effect among the plurality of classification models; performing multidimensional inversion on the geophysical data to obtain an inversion result; inputting the inversion result into the optimal classification model for processing to obtain a lithology interpretation result; the preprocessing of the regional logging data to obtain the target regional logging data set comprises the following steps: performing principal component analysis on the regional logging data, and performing outlier elimination processing and abnormal data recovery processing according to lithology labels, principal component conditions and clustering conditions to obtain the target regional logging data set; the establishment of the plurality of classification models associated with the input of the plurality of physical characteristics comprises the following steps: based on the analysis of the intersection characteristics of the plurality of physical properties of the regional logging data, a plurality of classification models are established, the plurality of classification models at least including a probabilistic neural network model, a support vector machine model and a K-nearest neighbor algorithm model, each of the classification models at least taking the resistivity, velocity, density and magnetic susceptibility of the core as the model input and taking the lithology as the model output.

2. The method for regional lithology interpretation based on multi-physical property features according to claim 1, characterized in that, The outlier elimination processing is implemented by using a DBscan algorithm.

3. The method for regional lithology interpretation based on multi-physical property features according to claim 2, characterized in that, the training and verification of the plurality of classification models using the target regional logging data set to determine the optimal classification model comprises the following steps: dividing the target regional logging data set into a training set and a test set; based on a cross-validation method, the training set is used to train the plurality of classification models, and the test set is used to verify the classification effect of the trained plurality of classification models to determine the optimal classification model.

4. The method for regional lithology interpretation based on multi-property features according to claim 3, characterized in that, The cross-validation method adopts a K-fold cross-validation method.

5. The method for regional lithology interpretation based on multi-physical property features according to claim 4, characterized in that, the multidimensional inversion of the geophysical data to obtain an inversion result comprises the following steps: performing three-dimensional inversion on at least gravity data, seismic data and electromagnetic data in the geophysical data to obtain the resistivity distribution, velocity distribution, density distribution and magnetic susceptibility distribution of the core as a preliminary inversion result; calibrating the preliminary inversion result to obtain the final inversion result.

6. A multi-property feature based regional lithology interpretation system, characterized in that, The method comprises the following steps: a data acquisition module for acquiring regional logging data and geophysical data; a preprocessing module for preprocessing the regional logging data to obtain a target regional logging data set, the target regional logging data set having a plurality of physical characteristics; a model construction module for establishing a plurality of classification models associated with the input of the plurality of physical characteristics; a model training and verification module for training and verifying the plurality of classification models using the target regional logging data set to determine an optimal classification model, the optimal classification model being the model with the optimal classification effect among the plurality of classification models; An inversion module is configured to perform multi-dimensional inversion on the geophysical data to obtain an inversion result; An interpretation module is configured to input the inversion result into the optimal classification model to obtain a lithology interpretation result; The pre-processing of the regional logging data to obtain a target regional logging data set comprises the following steps: The principal component analysis of the regional logging data is performed, and the outlier elimination processing and the abnormal data recovery processing are performed according to the lithology label, the principal component condition and the clustering condition, so as to obtain the target regional logging data set; The establishment of the multiple classification models associated with the multi-property feature input comprises the following steps: Based on the multi-property intersection feature analysis of the regional logging data, multiple classification models are established, and the multiple classification models at least include a probabilistic neural network model, a support vector machine model and a K-nearest neighbor algorithm model. Each of the classification models at least takes the resistivity, the velocity, the density and the magnetic susceptibility of the core as the model input and takes the lithology as the model output.

7. A multi-property feature based regional lithology interpretation device, characterized by, The computer readable storage medium stores computer executable instructions for causing a computer to perform the regional lithology interpretation method based on the multi-property feature as claimed in any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that, The computer readable storage medium stores computer executable instructions for causing a computer to perform the regional lithology interpretation method based on the multi-property feature as claimed in any one of claims 1 to 5.

Images