Lithium ore surveying system and method based on data analysis

Abstract

The invention discloses a lithium ore exploration system and method based on data analysis, and relates to the technical field of lithium ore exploration. In the operation process of the system, data from different sources are collected, geological abnormal parameters, lithologic density fluctuation parameters and gravity gradient parameters are obtained, the preprocessed data are analyzed, and a lithium ore exploration result is obtained; the method comprises the following steps: extracting geological features related to mineral resources, carrying out pattern recognition by using a deep learning technology, extracting key features from geological data, obtaining lithologic density Sgeo, rock magnetic parameters Wrock and rock resistivity Rgrad, and generating a three-dimensional geological model of a mining area through an AI algorithm based on the extracted features. The AI algorithm comprises regression analysis and a convolutional neural network (CNN) to process spatial data and generate three-dimensional visualization of a geologic structure, historical mining area distribution data is combined, the system uses a prediction algorithm random forest or a support vector machine (SVM) to predict lithium ore distribution, survey feedback is generated according to a prediction result, and a survey scheme is dynamically adjusted.

Description

Technical Field

[0001] The present invention relates to the technical field of lithium ore exploration, and in particular to a lithium ore exploration system and method based on data analysis. Background Art

[0002] Lithium mineral resources are an indispensable key material in the modern energy industry and battery technology. With the sharp increase in global demand for clean energy, the importance of lithium mineral exploration has become increasingly prominent. However, traditional lithium mineral exploration methods often rely on geological field surveys and sampling, which are limited by the coverage of the survey, the complexity of the terrain, and the timeliness of data acquisition. In order to improve the efficiency of exploration and increase the accuracy of mineral prediction, lithium mineral exploration based on data analysis has gradually become an advanced and efficient solution.

[0003] At present, traditional lithium mine exploration systems have many deficiencies in data processing and result prediction, mainly manifested in data dispersion, low prediction accuracy and lack of dynamic adjustment capabilities. Existing survey systems often rely on a single data source and lack the comprehensive use of multi-source data, resulting in poor robustness of model results. In addition, the three-dimensional geological model generation in traditional systems often fails to fully reflect the complex structure of the mining area, further affecting the accuracy of lithium mine distribution prediction. The above problems make it difficult for surveyors to make timely adjustments in subsequent work, resulting in low work efficiency and increased time costs and resource waste in surveying.

[0004] These problems mainly arise from the limitations of traditional systems in large-scale data processing, insufficient analysis of spatial information, and lack of automated prediction and feedback mechanisms. Especially when faced with complex geological conditions in mining areas, the lack of flexible model calibration methods will directly lead to deviations in prediction results. In addition, traditional survey systems often ignore the necessity of real-time data updates and fail to dynamically adjust and evaluate prediction results, which further affects the accuracy of lithium resource surveys and the rationality of survey work. As a result, some high-potential mining areas are ignored, or a lot of time and money are spent in inefficient survey areas. Summary of the invention

[0005] In view of the deficiencies in the prior art, the present invention provides a lithium ore exploration system and method based on data analysis, which solves the problems mentioned in the background technology.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions: a lithium ore survey system based on data analysis, including a data acquisition module, a data preprocessing module, a geological feature analysis module, a three-dimensional geological model generation module, a lithium ore distribution prediction module and a survey feedback module;

[0007] The data acquisition module is used to collect data from different sources, including geological history data, remote sensing images of mining areas, geological structure parameters and lithology resistivity data, and obtain: geological anomaly parameters, lithology density fluctuation parameters and gravity gradient parameters;

[0008] The data preprocessing module is used to clean, normalize and denoise the collected data, including outlier removal, data unit standardization and noise removal. After the data is processed, the dimension difference and noise influence are eliminated and uploaded to the geological feature analysis module;

[0009] The geological feature analysis module is used to analyze the pre-processed data to extract geological features related to the mineral deposits, use deep learning technology for pattern recognition, extract key features from geological data, and obtain: lithology density S geo , rock magnetic parameters W rock and rock resistivity R grad ;

[0010] The 3D geological model generation module is used to generate a 3D geological model of the mining area based on the extracted features through an AI algorithm, wherein the AI ​​algorithm includes regression analysis and a convolutional neural network (CNN) to process spatial data and generate a 3D visualization of the geological structure;

[0011] The lithium ore distribution prediction module is used to predict the distribution of lithium ore based on the generated three-dimensional model and in combination with historical mining area distribution data. The system uses a prediction algorithm random forest or support vector machine (SVM) to predict the distribution of lithium ore;

[0012] The survey feedback module is used to generate survey feedback according to the prediction results and dynamically adjust the survey plan. The feedback includes reliability assessment and risk analysis of the prediction area.

[0013] Preferably, the data acquisition module includes a geological history data acquisition unit, a remote sensing image acquisition unit and a field survey equipment acquisition unit;

[0014] The geological history data acquisition unit is used to extract historical data from the existing geological survey database, including known mining area distribution and vein morphology, to obtain: lithology density S geo ;

[0015] The remote sensing image acquisition unit is used to obtain remote sensing images of the surface of the mining area using drones and satellite equipment, and identify surface changes and potential mineral deposition areas through image analysis;

[0016] The field survey equipment acquisition unit is used to collect physical parameters of the mining area from the field through seismic waves, magnetic exploration and gravity measurement to obtain: rock magnetic parameters W rock and rock resistivity R grad , used to construct geological structure models.

[0017] Preferably, the data preprocessing module includes a data processing unit;

[0018] The data cleaning unit is used to delete invalid and abnormal data and fill in missing data, standardize data from different sources to eliminate differences in dimensions and numerical ranges, remove high-frequency noise or interference data in the collected data through signal processing algorithms, and upload them to the geological feature analysis module.

[0019] Preferably, the geological feature analysis module includes a geological feature extraction unit and a feature optimization unit;

[0020] The geological feature extraction unit is used to extract key geological features related to lithium ore from raw data, including lithology changes and structural anomalies, using a machine learning algorithm and a convolutional neural network;

[0021] The feature optimization unit optimizes the extracted features through the feature selection LASSO regression algorithm, filters out irrelevant or redundant information, uses the final geological parameters for subsequent modeling, combines multiple extracted features into comprehensive indicators, and generates global feature data of geological structures for geological model generation.

[0022] Preferably, the three-dimensional geological model generation module includes a three-dimensional structure generation unit and a model calibration unit;

[0023] The three-dimensional structure generation unit is used to generate a three-dimensional geological structure according to the geological feature data, and use computer graphics technology to draw a three-dimensional model of the mining area to display the relationship between geological layers and the distribution of the mining area;

[0024] The three-dimensional geological model M 3D Calculated by the following formula:

[0025] M 3D =α*S geo +β*W rock +γ*R grad ;

[0026] Where M 3D represents the three-dimensional geological model, S geo Indicates lithology density, W rock represents the rock magnetic parameters, R grad represents rock resistivity, α, β and γ represent lithology density S geo , rock magnetic parameters W rock and rock resistivity R grad The weight coefficient of

[0027] The model calibration unit is used to iteratively optimize the initially generated three-dimensional model through a machine learning algorithm to make the model accurate and more consistent with the known mining area distribution. Based on real-time data collected on site, the generated three-dimensional geological model is calibrated and the geological structure parameters in the model are adjusted to make them consistent with the actual data.

[0028] Preferably, the lithium ore distribution prediction module includes a prediction algorithm unit, a prediction optimization unit and a probability calculation unit;

[0029] The prediction algorithm unit is used to use a prediction algorithm based on historical data training, including a random forest and a support vector machine, to predict the lithium ore distribution area and generate the lithium ore distribution probability of each area in the mining area;

[0030] The prediction optimization unit is used to adjust the parameters, prediction accuracy and robustness of the prediction model according to the current mining area data and known prediction errors;

[0031] The probability calculation unit is used to calculate the probability value of lithium ore distribution, and obtains the following after calculation: li , and map it into a three-dimensional model to graphically display the mineral distribution area and assist decision makers in on-site surveys.

[0032] Preferably, the predicted probability P li Calculated by the following formula:

[0033]

[0034] Where P li represents the predicted probability, M 3D Represents a three-dimensional geological model, and θ0 and θ1 are obtained through machine learning training.

[0035] Preferably, the lithium ore distribution prediction module further includes an evaluation and estimation unit;

[0036] The evaluation and estimation unit is used to evaluate the exploration potential of each area according to the predicted probability of lithium ore distribution, assign risk levels according to preset thresholds, generate priorities and exploration strategy recommendations for distribution areas, and ensure that when predicting lithium ore distribution, high-probability areas can be marked as exploration priorities, while low-probability areas are treated as secondary exploration or abandoned by defining different threshold intervals;

[0037] Set three thresholds T1, T2 and T3, T1 is the threshold of the first survey priority, T2 is the threshold of the second survey priority, and T3 is the threshold of the third survey priority. The prediction result P li It will be mapped to these three thresholds and exploration priorities will be assigned according to their specific values. The specific process is as follows:

[0038] If \(PLi > T1\), obtain the first evaluation plan. The prediction result shows that the lithium ore distribution probability in this area is greater than \(T1\), indicating that the probability of lithium ore distribution is above 80%, and exploration needs to be carried out preferentially;

[0039] If \(T2 < PLi\leq T1\), obtain the second evaluation plan. The prediction result shows that the probability of lithium ore distribution is within 50%-80%, and it is recommended as the second batch of exploration areas for appropriate sampling and exploration verification;

[0040] If \(T3 < PLi\leq T2\), obtain the third evaluation plan. The prediction result shows that the probability of lithium ore distribution is within 20%-50%, with a relatively low probability but still potential, as part of the long-term exploration plan;

[0041] If \(PLi\leq T3\), obtain the fourth evaluation plan. The prediction result shows that the probability of lithium ore distribution is within 0%-20%, and it is abandoned or used as a backup option without actual exploration.

[0042] The lithium ore distribution prediction module further includes an evaluation and estimation unit;

[0043] The evaluation and estimation unit is used to evaluate the exploration potential of each area according to the predicted probability of lithium ore distribution, assign risk levels according to preset thresholds, generate the priority of the distribution area and suggestions for exploration strategies. By defining different threshold intervals, it is ensured that when predicting the lithium ore distribution, high-probability areas can be preferentially marked as the key exploration points, while low-probability areas are used as secondary exploration or abandoned;

[0044] Set three thresholds \(T1\), \(T2\) and \(T3\). \(T1\) is the threshold for the first exploration priority, \(T2\) is the threshold for the second-level exploration priority, and \(T3\) is the threshold for the third exploration priority. The prediction result \(P\) li will be mapped to these three thresholds, and the exploration priority will be assigned according to its specific value. The specific process is as follows:

[0045] If \(PLi > T1\), obtain the first evaluation plan. The prediction result shows that the lithium ore distribution probability in this area is greater than \(T1\), indicating that the probability of lithium ore distribution is above 80%, and exploration needs to be carried out preferentially;

[0046] If \(T2 < PLi\leq T1\), obtain the second evaluation plan. The prediction result shows that the probability of lithium ore distribution is within 50%-80%, and it is recommended as the second batch of exploration areas for appropriate sampling and exploration verification;

[0047] If \(T3 < PLi\leq T2\), obtain the third evaluation plan. The prediction result shows that the probability of lithium ore distribution is within 20%-50%, with a relatively low probability but still potential, as part of the long-term exploration plan;

[0048] PLi≤T3, obtain the fourth evaluation plan. The prediction results show that the probability of lithium ore distribution is within 0%-20%. Abandon it or use it as a backup option without conducting actual exploration.

[0049] Preferably, the survey feedback module includes a feedback analysis unit and a real-time adjustment unit;

[0050] The feedback analysis unit is used to analyze the reliability of the prediction results, identify high-risk or high-value areas in the survey area, and generate corresponding feedback information to provide reference for on-site survey;

[0051] The real-time adjustment unit is used to adjust the survey plan in real time according to the feedback analysis results, and dynamically modify the focus and direction of subsequent survey tasks.

[0052] A lithium ore exploration method based on data analysis comprises the following steps:

[0053] Step 1: Collect data from different sources, including geological history data, remote sensing images of mining areas, geological structure parameters and lithology resistivity data, and obtain: geological anomaly parameters, lithology density fluctuation parameters and gravity gradient parameters;

[0054] Step 2: Clean, normalize and denoise the collected data, including outlier removal, data unit standardization and noise removal. After the data is processed, dimension differences and noise effects are eliminated and uploaded to the geological feature analysis module;

[0055] Step 3: Analyze the preprocessed data to extract the geological features related to the mineral deposits, use deep learning technology for pattern recognition, extract key features from the geological data, and obtain: lithology density S geo , rock magnetic parameters W rock and rock resistivity R grad ;

[0056] Step 4: Based on the extracted features, a 3D geological model of the mining area is generated through AI algorithms, including regression analysis and convolutional neural networks (CNN) to process spatial data and generate 3D visualization of geological structures;

[0057] Step 5: Based on the generated 3D model and combined with historical mining area distribution data, the system uses the prediction algorithm random forest or support vector machine (SVM) to predict the distribution of lithium ore;

[0058] Step 6: Generate survey feedback based on the prediction results and dynamically adjust the survey plan. The feedback includes reliability assessment and risk analysis of the prediction area.

[0059] The present invention provides a lithium ore exploration method based on data analysis, which has the following beneficial effects:

[0060] (1) When the system is running, data from different sources are collected to obtain: geological anomaly parameters, lithology density fluctuation parameters and gravity gradient parameters. By analyzing the pre-processed data, geological features related to the mineral deposits are extracted. Deep learning technology is used for pattern recognition to extract key features from geological data to obtain: lithology density S geo , rock magnetic parameters W rock and rock resistivity R grad Based on the extracted features, a three-dimensional geological model of the mining area is generated through AI algorithms. The AI ​​algorithms include regression analysis and convolutional neural networks (CNN) to process spatial data and generate three-dimensional visualization of geological structures. Combined with historical mining area distribution data, the system uses prediction algorithms such as random forests or support vector machines (SVM) to predict the distribution of lithium ore, generate survey feedback based on the prediction results, and dynamically adjust the survey plan.

[0061] (2) The lithium ore exploration system based on data analysis realizes efficient lithium ore resource identification and evaluation through modular design. First, the data acquisition and integration module realizes the accurate acquisition and standardized integration of multi-source data such as geology, physics, and chemistry, avoiding the inconsistency problems caused by different data sources. The intelligent geological model generation module dynamically simulates and predicts the geological structure of the mining area based on big data analysis and machine learning technology, significantly improving the accuracy of the survey. By analyzing the changing trends of the strata and the rules of lithium ore enrichment, the system provides more accurate decision-making support for subsequent exploration and development.

[0062] (3) The survey data preprocessing module and the ore body location identification module ensure the high quality and accuracy of the data. By automatically removing data outliers, standardizing and reducing noise, the system can effectively reduce noise interference and improve data processing efficiency. The ore body location identification module uses complex mathematical models to automatically identify potential lithium ore-rich areas, reducing the time and errors of manual analysis. In addition, the reserve estimation module provides more accurate lithium ore reserve assessment data based on advanced volume calculation and concentration distribution algorithms, greatly improving the utilization rate of exploration resources.

[0063] (4) Compared with traditional technical means, the system has significant technical improvements, especially in data processing speed, survey accuracy, reserve estimation and risk control. By introducing intelligent algorithms and adaptive adjustment mechanisms, the system reduces data errors and improves survey efficiency while also reducing costs and environmental risks in the exploration process. Overall, the system has effectively improved the accuracy and efficiency of lithium mine exploration through intelligent and automated workflows, and promoted technological innovation in the surveying industry. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1This is a block diagram of a lithium ore exploration system based on data analysis of the present invention;

[0065] Figure 2 This is a schematic diagram of the steps of a lithium ore exploration method based on data analysis of the present invention. DETAILED DESCRIPTION

[0066] 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.

[0067] Example 1

[0068] The present invention provides a lithium ore exploration system based on data analysis, please refer to Figure 1 , including data acquisition module, data preprocessing module, geological feature analysis module, 3D geological model generation module, lithium ore distribution prediction module and survey feedback module;

[0069] The data acquisition module is used to collect data from different sources, including geological history data, remote sensing images of mining areas, geological structure parameters and lithology resistivity data, and obtain: geological anomaly parameters, lithology density fluctuation parameters and gravity gradient parameters;

[0070] The data preprocessing module is used to clean, normalize and denoise the collected data, including outlier removal, data unit standardization and noise removal. After the data is processed, the dimension difference and noise influence are eliminated and uploaded to the geological feature analysis module;

[0071] The geological feature analysis module is used to analyze the pre-processed data to extract geological features related to the mineral deposits, use deep learning technology for pattern recognition, extract key features from geological data, and obtain: lithology density S geo , rock magnetic parameters W rock and rock resistivity R grad ;

[0072] The 3D geological model generation module is used to generate a 3D geological model of the mining area based on the extracted features through an AI algorithm, wherein the AI ​​algorithm includes regression analysis and a convolutional neural network (CNN) to process spatial data and generate a 3D visualization of the geological structure;

[0073] The lithium ore distribution prediction module is used to predict the distribution of lithium ore based on the generated three-dimensional model and in combination with historical mining area distribution data. The system uses a prediction algorithm random forest or support vector machine (SVM) to predict the distribution of lithium ore;

[0074] The survey feedback module is used to generate survey feedback according to the prediction results and dynamically adjust the survey plan. The feedback includes reliability assessment and risk analysis of the prediction area.

[0075] In this embodiment, data from different sources are collected to obtain: geological anomaly parameters, lithology density fluctuation parameters and gravity gradient parameters. The pre-processed data are analyzed to extract geological features related to the mineral deposits. Deep learning technology is used for pattern recognition to extract key features from geological data to obtain: lithology density S geo , rock magnetic parameters W rock and rock resistivity R grad Based on the extracted features, a three-dimensional geological model of the mining area is generated through AI algorithms. The AI ​​algorithms include regression analysis and convolutional neural networks (CNN) to process spatial data and generate three-dimensional visualization of geological structures. Combined with historical mining area distribution data, the system uses prediction algorithms such as random forests or support vector machines (SVM) to predict the distribution of lithium ore, generate survey feedback based on the prediction results, and dynamically adjust the survey plan.

[0076] Example 2

[0077] This embodiment is explained in Example 1, please refer to Figure 1 ,Specifically: the data acquisition module includes a geological history data acquisition unit, a remote sensing image acquisition unit and a field survey equipment acquisition unit;

[0078] The geological history data acquisition unit is used to extract historical data from the existing geological survey database, including known mining area distribution and vein morphology, to obtain: lithology density S geo ;

[0079] The remote sensing image acquisition unit is used to obtain remote sensing images of the surface of the mining area using drones and satellite equipment, and identify surface changes and potential mineral deposition areas through image analysis;

[0080] The field survey equipment acquisition unit is used to collect physical parameters of the mining area from the field through seismic waves, magnetic exploration and gravity measurement to obtain: rock magnetic parameters W rock and rock resistivity R grad , used to construct geological structure models.

[0081] The data preprocessing module includes a data processing unit;

[0082] The data cleaning unit is used to delete invalid and abnormal data and fill in missing data, standardize data from different sources to eliminate differences in dimensions and numerical ranges, remove high-frequency noise or interference data in the collected data through signal processing algorithms, and upload them to the geological feature analysis module.

[0083] The geological feature analysis module includes a geological feature extraction unit and a feature optimization unit;

[0084] The geological feature extraction unit is used to extract key geological features related to lithium ore from raw data, including lithology changes and structural anomalies, using a machine learning algorithm and a convolutional neural network;

[0085] The feature optimization unit optimizes the extracted features through the feature selection LASSO regression algorithm, filters out irrelevant or redundant information, uses the final geological parameters for subsequent modeling, combines multiple extracted features into comprehensive indicators, and generates global feature data of geological structures for geological model generation.

[0086] In this embodiment, through the above data collection and preprocessing process, the system realizes the comprehensive and accurate collection and efficient processing of lithium mine exploration data, bringing significant beneficial effects. First, the multi-source data fusion of geological historical data, remote sensing images and on-site survey equipment ensures the comprehensive coverage of mining area information and greatly improves the integrity and diversity of data. At the same time, the lithology density S geo , rock magnetic parameters W rock and rock resistivity R grad The introduction of uncommon parameters such as , enriches the depth and dimension of the data, and provides a more reliable basis for subsequent analysis. In addition, through the data cleaning and standardization operations of the preprocessing module, the system effectively solves the problem of differences between different data sources, reduces noise interference and invalid information, and significantly improves the quality and consistency of the data. Ultimately, the processed data provides high-quality input for subsequent geological feature analysis and model construction, ensuring the accuracy of the analysis results.

[0087] Example 3

[0088] This embodiment is explained in Example 1, please refer to Figure 1 ,Specifically: the three-dimensional geological model generation module includes a three-dimensional structure generation unit and a model calibration unit;

[0089] The three-dimensional structure generation unit is used to generate a three-dimensional geological structure according to the geological feature data, and use computer graphics technology to draw a three-dimensional model of the mining area to display the relationship between geological layers and the distribution of the mining area;

[0090] The three-dimensional geological model M 3D Calculated by the following formula:

[0091] M 3D =α*S geo +β*W rock +γ*R grad ;

[0092] Where M 3D represents the three-dimensional geological model, S geo Indicates lithology density, W rock represents the rock magnetic parameters, R grad represents rock resistivity, α, β and γ represent lithology density S geo , rock magnetic parameters W rock and rock resistivity R grad The weight coefficient of

[0093] The model calibration unit is used to iteratively optimize the initially generated three-dimensional model through a machine learning algorithm to make the model accurate and more consistent with the known mining area distribution. Based on real-time data collected on site, the generated three-dimensional geological model is calibrated and the geological structure parameters in the model are adjusted to make them consistent with the actual data.

[0094] In this embodiment, through the three-dimensional geological model generation module, the system effectively improves the visualization ability and model accuracy in the lithium ore exploration process, bringing significant beneficial effects. First, the three-dimensional model generated by the three-dimensional structure generation unit using geological feature data clearly displays the geological hierarchy and vein distribution of the mining area, greatly enhancing the intuitive understanding of the mining area. The model can not only display the surface characteristics of the mining area, but also deeply display the relationship between geological layers, providing a powerful auxiliary tool for lithium ore exploration. In addition, based on the lithology density S geo , rock magnetic parameters W rock and rock resistivity R grad The weighted formula of parameters such as , enables the three-dimensional geological model M3D to accurately reflect the geological anomalies and gravity changes in the mining area, improving the authenticity of the model. Through the machine learning optimization and real-time data calibration of the model calibration unit, the model can be continuously adjusted and refined, and is highly consistent with the actual distribution of the mining area. The innovation of this module significantly improves the accuracy and efficiency of the survey, and provides a reliable basis for mining and planning.

[0095] Example 4

[0096] This embodiment is explained in Example 1, please refer to Figure 1 , specifically: the lithium ore distribution prediction module includes a prediction algorithm unit, a prediction optimization unit and a probability calculation unit;

[0097] The prediction algorithm unit is used to use a prediction algorithm based on historical data training, including a random forest and a support vector machine, to predict the lithium ore distribution area and generate the lithium ore distribution probability of each area in the mining area;

[0098] The prediction optimization unit is used to adjust the parameters, prediction accuracy, and model robustness of the prediction model according to the current mine area data and known prediction errors;

[0099] The probability calculation unit is used to calculate the probability value of the lithium ore distribution. After calculation, the prediction probability P li is obtained and mapped into the 3D model to graphically display the mineral distribution area to assist decision-makers in on-site exploration.

[0100] The prediction probability P li is calculated and obtained through the following formula:

[0101]

[0102] In the formula, P li represents the prediction probability, M 3D represents the 3D geological model, and θ0 and θ1 are obtained through machine learning training.

[0103] The lithium ore distribution prediction module further includes an evaluation and estimation unit;

[0104] The evaluation and estimation unit is used to evaluate the exploration potential of each area according to the prediction probability of the lithium ore distribution, assign risk levels according to preset thresholds, generate the priority of the distribution area and suggestions for exploration strategies. By defining different threshold intervals, it is ensured that when predicting the lithium ore distribution, high-probability areas can be preferentially marked as key exploration areas, while low-probability areas are secondary exploration or abandoned;

[0105] Three thresholds T1, T2, and T3 are set. T1 is the threshold for the first exploration priority, T2 is the threshold for the second-level exploration priority, and T3 is the threshold for the third exploration priority. The prediction result P li will be mapped to these three thresholds, and the exploration priority will be assigned according to its specific value. The specific process is as follows:

[0106] PLi > T1, obtain the first evaluation plan. The prediction result shows that the probability of lithium ore distribution in this area is greater than T1, indicating that the probability of lithium ore distribution is above 80%, and exploration needs to be carried out preferentially;

[0107] T2 < PLi ≤ T1, obtain the second evaluation plan. The prediction result shows that the probability of lithium ore distribution is within 50%-80%, and it is recommended as the second batch of exploration areas for appropriate sampling and exploration verification;

[0108] T3 < PLi ≤ T2, obtain the third evaluation plan. The prediction result shows that the probability of lithium ore distribution is within 20%-50%, with a relatively low probability but still potential, as part of the long-term exploration plan;

[0109] PLi≤T3, obtain the fourth evaluation plan. The prediction results show that the probability of lithium ore distribution is within 0%-20%. Abandon it or use it as a backup option without conducting actual exploration.

[0110] The survey feedback module includes a feedback analysis unit and a real-time adjustment unit;

[0111] The feedback analysis unit is used to analyze the reliability of the prediction results, identify high-risk or high-value areas in the survey area, and generate corresponding feedback information to provide reference for on-site survey;

[0112] The real-time adjustment unit is used to adjust the survey plan in real time according to the feedback analysis results, and dynamically modify the focus and direction of subsequent survey tasks.

[0113] In this embodiment, through the lithium ore distribution prediction module, the system greatly improves the accuracy and decision-making efficiency of mining exploration, bringing multiple beneficial effects. First, the prediction algorithm unit accurately predicts the lithium ore distribution probability Pli in each area of ​​the mining area based on historical data and three-dimensional geological model M3D through advanced algorithms such as random forests and support vector machines, making the exploration work more targeted and feasible. The prediction optimization unit further improves the prediction accuracy and robustness of the model to ensure that the prediction results of lithium ore distribution are more reliable. In addition, the probability calculation unit visualizes the prediction results and intuitively displays the mineral distribution area to help decision makers quickly identify potential high-value exploration areas. Through the threshold division mechanism of the evaluation and estimation unit, the module can allocate priorities and survey strategies according to the probability of lithium ore distribution in different regions to ensure the reasonable allocation of exploration resources. High-probability areas are prioritized as exploration priorities to ensure that the input-output ratio of exploration work is maximized; medium and low-probability areas are reasonably planned as secondary or long-term exploration targets to avoid waste of resources. Finally, combined with the real-time adjustment function of the survey feedback module, the system can dynamically adjust the exploration plan according to actual feedback to ensure the flexibility and efficiency of subsequent tasks. This feedback closed-loop mechanism significantly improves the adaptability and effectiveness of the entire exploration process, reduces exploration risks, and optimizes the development and utilization of lithium mineral resources.

[0114] Example 5

[0115] A lithium ore exploration method based on data analysis, please refer to Figure 2 , specifically: including the following steps:

[0116] Step 1: Collect data from different sources, including geological history data, remote sensing images of mining areas, geological structure parameters and lithology resistivity data, and obtain: geological anomaly parameters, lithology density fluctuation parameters and gravity gradient parameters;

[0117] Step 2: Clean, normalize and denoise the collected data, including outlier removal, data unit standardization and noise removal. After the data is processed, dimension differences and noise effects are eliminated and uploaded to the geological feature analysis module;

[0118] Step 3: Analyze the preprocessed data to extract the geological features related to the mineral deposits, use deep learning technology for pattern recognition, extract key features from the geological data, and obtain: lithology density S geo , rock magnetic parameters W rock and rock resistivity R grad ;

[0119] Step 4: Based on the extracted features, a 3D geological model of the mining area is generated through AI algorithms, including regression analysis and convolutional neural networks (CNN) to process spatial data and generate 3D visualization of geological structures;

[0120] Step 5: Based on the generated 3D model and combined with historical mining area distribution data, the system uses the prediction algorithm random forest or support vector machine (SVM) to predict the distribution of lithium ore;

[0121] Step 6: Generate survey feedback based on the prediction results and dynamically adjust the survey plan. The feedback includes reliability assessment and risk analysis of the prediction area.

[0122] In this embodiment, through the systematic process of the above six steps, the entire lithium ore exploration process has been significantly optimized, bringing many beneficial effects. First, the data acquisition module ensures the comprehensiveness and accuracy of the exploration data through the comprehensive application of multi-source data, such as geological historical data, remote sensing images and physical parameters, especially the acquisition of geological anomaly parameters, lithology density fluctuation parameters and gravity gradient parameters, laying a solid foundation for subsequent analysis. Secondly, the data preprocessing module effectively eliminates data noise and dimensional differences through cleaning, normalization and denoising, making the data more unified and accurate, and providing high-quality input for geological feature analysis. The subsequent geological feature extraction and three-dimensional model generation process further improves the accuracy and efficiency of exploration. Through deep learning and AI algorithms, the system not only extracts key geological features related to the mineral deposits, but also generates a three-dimensional geological model of the mining area, realizing the visualization of the geological structure. Combined with prediction algorithms such as random forests and support vector machines, the system can accurately predict the distribution area of ​​lithium mines and dynamically adjust the survey plan through the feedback module. This closed-loop system ensures the accuracy and reliability of prediction results, reduces exploration risks, improves survey efficiency, and optimizes the allocation and utilization of survey resources.

[0123] Although embodiments of the present invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the invention, and the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A lithium ore exploration system based on data analysis, characterized in that: It includes data acquisition module, data preprocessing module, geological feature analysis module, 3D geological model generation module, lithium ore distribution prediction module and survey feedback module; The data acquisition module is used to collect data from different sources, including geological history data, remote sensing images of mining areas, geological structure parameters and lithology resistivity data, and obtain: geological anomaly parameters, lithology density fluctuation parameters and gravity gradient parameters; The data preprocessing module is used to clean, normalize and denoise the collected data, including outlier removal, data unit standardization and noise removal. After the data is processed, the dimension difference and noise influence are eliminated and uploaded to the geological feature analysis module; The geological feature analysis module is used to analyze the pre-processed data to extract geological features related to the mineral deposits, use deep learning technology for pattern recognition, extract key features from geological data, and obtain: lithology density S geo , rock magnetic parameters W rock and rock resistivity R grad ; The three-dimensional geological model generation module is used to generate a three-dimensional geological model of the mining area based on the extracted features through an AI algorithm, wherein the AI ​​algorithm includes regression analysis and convolutional neural network to process spatial data and generate a three-dimensional visualization of the geological structure; The lithium ore distribution prediction module is used to predict the distribution of lithium ore based on the generated three-dimensional model and the historical mining area distribution data. The system uses the prediction algorithm random forest or support vector machine to predict the distribution of lithium ore; lithology density, rock magnetic parameters, rock resistivity The survey feedback module is used to generate survey feedback according to the prediction results and dynamically adjust the survey plan. The feedback includes reliability assessment and risk analysis of the prediction area.

2. A lithium ore exploration system based on data analysis according to claim 1, characterized in that: The data acquisition module includes a geological history data acquisition unit, a remote sensing image acquisition unit and a field survey equipment acquisition unit; The geological history data acquisition unit is used to extract historical data from the existing geological survey database, including known mining area distribution and vein morphology, to obtain: lithology density S geo ; The remote sensing image acquisition unit is used to obtain remote sensing images of the surface of the mining area using drones and satellite equipment, and identify surface changes and potential mineral deposition areas through image analysis; The field survey equipment acquisition unit is used to collect physical parameters of the mining area from the field through seismic waves, magnetic exploration and gravity measurement to obtain: rock magnetic parameters W rock and rock resistivity R grad , used to construct geological structure models.

3. A lithium ore exploration system based on data analysis according to claim 1, characterized in that: The data preprocessing module includes a data processing unit; The data cleaning unit is used to delete invalid and abnormal data and fill in missing data, standardize data from different sources to eliminate differences in dimensions and numerical ranges, remove high-frequency noise or interference data in the collected data through signal processing algorithms, and upload them to the geological feature analysis module.

4. A lithium ore exploration system based on data analysis according to claim 1, characterized in that: The geological feature analysis module includes a geological feature extraction unit and a feature optimization unit; The geological feature extraction unit is used to extract key geological features related to lithium ore from raw data, including lithology changes and structural anomalies, using a machine learning algorithm and a convolutional neural network; The feature optimization unit optimizes the extracted features through the feature selection LASSO regression algorithm, filters out irrelevant or redundant information, uses the final geological parameters for subsequent modeling, combines multiple extracted features into comprehensive indicators, and generates global feature data of geological structures for geological model generation.

5. A lithium ore exploration system based on data analysis according to claim 2, characterized in that: The three-dimensional geological model generation module includes a three-dimensional structure generation unit and a model calibration unit; The three-dimensional structure generation unit is used to generate a three-dimensional geological structure according to the geological feature data, and use computer graphics technology to draw a three-dimensional model of the mining area to display the relationship between geological layers and the distribution of the mining area; The three-dimensional geological model M 3D Calculated by the following formula: M 3D =α*S geo +β*W rock +γ*R grad ; Where M 3D represents the three-dimensional geological model, S geo Indicates lithology density, W rock represents the rock magnetic parameters, R grad represents rock resistivity, α, β and γ represent lithology density S geo , rock magnetic parameters W rock and rock resistivity R grad The weight coefficient of The model calibration unit is used to iteratively optimize the preliminarily generated 3D model through machine learning algorithms, making the model accurate and more matching with the known mining area distribution. Based on the real-time data collected on-site, the generated 3D geological model is calibrated, and the geological structure parameters in the model are adjusted to make it consistent with the actual data.

6. A lithium ore exploration system based on data analysis according to claim 1, characterized in that: The lithium ore distribution prediction module includes a prediction algorithm unit, a prediction optimization unit, and a probability calculation unit; The prediction algorithm unit is used to use prediction algorithms trained based on historical data, including random forest and support vector machine, to predict the lithium ore distribution area and generate the lithium ore distribution probability of each area in the mining area; The prediction optimization unit is used to adjust the parameters, prediction accuracy, and robustness of the prediction model according to the current mining area data and known prediction errors; The probability calculation unit is used to calculate the probability value of lithium ore distribution, and obtains the following after calculation: li , and map it into a three-dimensional model to graphically display the mineral distribution area and assist decision makers in on-site surveys.

7. A lithium ore exploration system based on data analysis according to claim 6, characterized in that: The predicted probability P li Calculated by the following formula: Where P li represents the predicted probability, M 3D Represents a three-dimensional geological model, and θ0 and θ1 are obtained through machine learning training.

8. A lithium ore exploration system based on data analysis according to claim 6, characterized in that: The lithium ore distribution prediction module further includes an evaluation and estimation unit; The evaluation and estimation unit is used to evaluate the exploration potential of each area according to the predicted probability of lithium ore distribution, assign risk levels according to preset thresholds, generate the priority of the distribution area and suggestions for exploration strategies. By defining different threshold intervals, it is ensured that when predicting the lithium ore distribution, high-probability areas can be preferentially marked as key exploration areas, while low-probability areas are used as secondary exploration or abandoned; Set three thresholds T1, T2 and T3, T1 is the threshold of the first survey priority, T2 is the threshold of the second survey priority, and T3 is the threshold of the third survey priority. The prediction result P li It will be mapped to these three thresholds and exploration priorities will be assigned according to their specific values. The specific process is as follows: PLi>T1, obtain the first evaluation plan. The prediction result shows that the lithium ore distribution probability in this area is greater than T1, indicating that the lithium ore distribution probability is above 80%, and exploration needs to be carried out preferentially; T2<PLi≤T1, obtain the second evaluation plan. The prediction result shows that the lithium ore distribution probability is within 50%-80%, and it is recommended to be the second batch of exploration areas for moderate sampling and exploration verification; T3<PLi≤T2, obtain the third evaluation plan. The prediction result shows that the lithium ore distribution probability is within 20%-50%, with a relatively low probability but still potential, and it is part of the long-term exploration plan; PLi≤T3, obtain the fourth evaluation plan. The prediction result shows that the lithium ore distribution probability is within 0%-20%, and it is abandoned or used as an alternative option without actual exploration.

9. The lithium ore exploration system based on data analysis according to claim 1, characterized in that: The exploration feedback module includes a feedback analysis unit and a real-time adjustment unit; The feedback analysis unit is used to analyze the reliability of the prediction results, identify high-risk or high-value areas within the exploration area, and generate corresponding feedback information for reference in on-site exploration; The real-time adjustment unit is used to adjust the exploration plan in real time according to the feedback analysis results and dynamically modify the focus and direction of subsequent exploration tasks.

10. A lithium ore exploration method based on data analysis, applied to a lithium ore exploration system based on data analysis according to any one of claims 1 to 9, characterized in that: It includes the following steps: Step 1: Collect data from different sources, including geological history data, remote sensing images of the mining area, geological structure parameters, and lithologic resistivity data, and obtain geological anomaly parameters, lithologic density fluctuation parameters, and gravity gradient parameters; Step 2: Clean, normalize, and denoise the collected data, including outlier removal, data unit standardization, and noise removal. After the data is processed, the dimension difference and noise influence are eliminated and uploaded to the geological feature analysis module; Step 3: Analyze the preprocessed data to extract geological features related to the mineral deposits, use deep learning technology for pattern recognition, extract key features from geological data, and obtain: lithology density S geo , rock magnetic parameters W rock and rock resistivity R grad ; Step 4: Based on the extracted features, a 3D geological model of the mining area is generated through an AI algorithm, which includes regression analysis and convolutional neural networks to process spatial data and generate a 3D visualization of the geological structure. Step 5: Based on the generated 3D model and combined with historical mining area distribution data, the system uses the prediction algorithm random forest or support vector machine to predict the distribution of lithium ore; Step 6: Generate survey feedback based on the prediction results and dynamically adjust the survey plan. The feedback includes reliability assessment and risk analysis of the prediction area.

Images