Salt lake lithium ore project capacity monitoring method and device based on satellite remote sensing

Abstract

The invention discloses a salt lake lithium ore project capacity monitoring method and device based on satellite remote sensing, and relates to the technical field of mineral product remote sensing monitoring, and the method comprises the steps: obtaining satellite remote sensing data and brine parameters, carrying out the preprocessing of the data, and obtaining historical geographic information data; a K-means clustering algorithm is adopted to classify the preprocessed multispectral image to obtain a classification result, and the classification result is compared with historical geographic information data to generate a salt lake function partition map; training a random forest algorithm by using the preprocessed brine parameters and the satellite remote sensing data, establishing a lithium concentration inversion model, regularly obtaining new satellite remote sensing data, and obtaining lithium concentration inversion values of pixels in the regularly received satellite remote sensing data in combination with the salt lake function partition map and the lithium concentration inversion model; and analyzing the lithium concentration inversion value to obtain a dynamic monitoring result. According to remote sensing data and brine parameters, a model is established, and the capacity of a salt lake lithium ore project is monitored.

Description

Technical Field

[0001] The present invention relates to the technical field of mineral remote sensing monitoring, and particularly to a method and device for monitoring the production capacity of a salt lake lithium mine project based on satellite remote sensing. Background Art

[0002] Traditional monitoring of salt lake lithium mines mainly relies on ground sampling and laboratory analysis. This method has two problems. One is that it is time-consuming and laborious, and it is difficult to achieve large-scale and high-frequency real-time monitoring. The other is that the data coverage is limited and it cannot comprehensively reflect the overall situation of the mining area. For example, ground sampling is usually limited to a few fixed points, making it difficult to capture the dynamic changes of the entire mining area, resulting in inaccurate and untimely monitoring results.

[0003] With the development of satellite remote sensing technology and machine learning algorithms, these problems have been effectively solved. Satellite remote sensing can provide large-area and high-resolution multi-spectral images. Combining with brine parameters (such as lithium concentration, salinity, etc.), dynamic monitoring of salt lake lithium mine projects can be achieved. However, existing methods still face challenges in processing large-scale spatio-temporal data, especially in automatic functional zoning and lithium concentration prediction. Summary of the Invention

[0004] In view of the above existing problems, the present invention is proposed.

[0005] Therefore, the present invention provides a method for monitoring the production capacity of a salt lake lithium mine project based on satellite remote sensing to solve the problems of time-consuming and laborious traditional monitoring means and difficulty in achieving large-scale and high-frequency real-time monitoring.

[0006] To solve the above technical problems, the present invention provides the following technical solutions:

[0007] In a first aspect, the present invention provides a method for monitoring the production capacity of a salt lake lithium mine project based on satellite remote sensing, which includes obtaining satellite remote sensing data and brine parameters, respectively preprocessing them, and then obtaining historical geographical information data;

[0008] Using the K-means clustering algorithm to classify the preprocessed multi-spectral image to obtain a classification result, and generating a salt lake functional zoning map by comparing with historical geographical information data;

[0009] Training a random forest algorithm with the preprocessed brine parameters and satellite remote sensing data to establish a lithium concentration inversion model, regularly obtaining new satellite remote sensing data, and combining with the salt lake functional zoning map and the lithium concentration inversion model to obtain the lithium concentration inversion value of the pixels in the regularly received satellite remote sensing data;

[0010] Analyzing the lithium concentration inversion value to obtain a dynamic monitoring result;

[0011] Generate a production capacity monitoring report for the salt lake lithium mine project based on the dynamic monitoring results.

[0012] As a preferred embodiment of the method for monitoring the production capacity of a salt lake lithium mine project based on satellite remote sensing according to the present invention, wherein: the satellite remote sensing data includes multi-spectral images, spectral bands, longitude and latitude information, spectral reflectance of pixels, and acquisition time at multiple time points obtained by satellite remote sensing;

[0013] The brine parameters include lithium concentration values at multiple ground sampling points within the salt lake area, longitude and latitude information of the ground sampling points, and sampling time;

[0014] The historical geographical information data includes the partition ranges and geometric shapes of the historical evaporation pond area, salt pan area, and natural lake area in the salt lake functional partition;

[0015] The preprocessing includes performing radiometric correction, time series synchronization, resampling, and standardization on the satellite remote sensing data, and performing data cleaning, inverse distance weighted data interpolation, and time series synchronization on the brine parameters.

[0016] As a preferred embodiment of the method for monitoring the production capacity of a salt lake lithium mine project based on satellite remote sensing according to the present invention, wherein: the specific steps for performing radiometric correction, time series synchronization, resampling, and standardization on the satellite remote sensing data are as follows.

[0017] Using the lithium concentration data collected at different positions in the salt lake and the spectral reflectance of the pixels corresponding to the longitude and latitude information points, perform radiometric correction on the multi-spectral image to obtain the spectral reflectance of the pixels of the corrected multi-spectral image;

[0018] Taking the sampling time of the brine parameters as the reference, perform time interpolation on the multi-spectral image using spline interpolation to complete the time alignment of the multi-spectral image;

[0019] Through longitude and latitude registration, spatially align the time-aligned multi-spectral image with the brine parameters to obtain the multi-spectral image after time series synchronization;

[0020] Adopt the bilinear interpolation method to unify the spatial resolution of each band of the multi-spectral image after time series synchronization to the scale of the highest resolution band to obtain a multi-spectral image with consistent spatial resolution;

[0021] Perform Z-score standardization on the spectral reflectance of the pixels after radiometric correction and with consistent spatial resolution to obtain satellite remote sensing data in a unified format.

[0022] As a preferred solution of the method for monitoring the production capacity of a salt lake lithium mine project based on satellite remote sensing according to the present invention, the following steps are included: classifying the preprocessed multi-spectral image by using the K-means clustering algorithm to obtain a classification result, and generating a salt lake functional zoning map by comparing with historical geographical information data. The specific steps are as follows:

[0023] Set the clustering parameters of the K-means clustering algorithm according to the number of partitions in the evaporation pond area, salt pan area, and natural lake area to obtain an initialized K-means clustering algorithm.

[0024] Use the initialized K-means clustering algorithm to perform clustering analysis on the preprocessed multi-spectral image to obtain a preliminary classification result.

[0025] Adopt a majority filtering algorithm to smooth the preliminary classification result to obtain a sketch of the functional zoning.

[0026] Perform geometric shape comparison analysis on the functional zoning sketch and historical geographical information data, identify the different parts and correct them to obtain the final salt lake functional zoning map.

[0027] As a preferred solution of the method for monitoring the production capacity of a salt lake lithium mine project based on satellite remote sensing according to the present invention, the following steps are included: training a random forest algorithm by using the preprocessed brine parameters and satellite remote sensing data to establish a lithium concentration inversion model, regularly obtaining new satellite remote sensing data, and combining the salt lake functional zoning map and the lithium concentration inversion model to obtain the lithium concentration inversion value of the pixel in the regularly received satellite remote sensing data. The specific steps are as follows:

[0028] Extract the spectral reflectance of the pixel matching the ground sampling point by using the longitude and latitude information to form a feature vector.

[0029] Take the lithium concentration value of the ground sampling point as the output target of the random forest model.

[0030] Summarize all the feature vectors and output targets to construct a data set.

[0031] Initialize the parameters of the random forest model according to the scale of the data set, and use the data set to train the random forest model to establish a lithium concentration inversion model.

[0032] Regularly obtain satellite remote sensing data covering the salt lake area, and input the spectral reflectance of each pixel in the regularly received satellite remote sensing data into the lithium concentration inversion model to obtain the lithium concentration inversion value of the pixel, denoted as

[0033]

[0034] Among them, is the lithium concentration inversion value of the pixel, T is the number of decision trees, t is the index of the decision tree, Nt is the number of leaf node samples in the t-th decision tree, i is the index of the feature vector, x is the feature vector, L t is the leaf node in the t-th decision tree, Ⅰ[x i ∈L t is the indicator function, y i is the lithium concentration value corresponding to the i-th feature vector in the leaf node.

[0035] As a preferred solution of the method for monitoring the production capacity of a salt lake lithium ore project based on satellite remote sensing according to the present invention, wherein: analyzing the inversion value of lithium concentration to obtain the dynamic monitoring result, the specific steps are as follows.

[0036] Using the lithium concentration inversion model to output the lithium concentration inversion value of the pixel, binding the lithium concentration inversion value of each pixel with the longitude and latitude to generate the lithium concentration spatial distribution map of the salt lake area;

[0037] According to the lithium concentration spatial distribution map, calculate the average lithium concentration of each functional area according to the functional area, expressed as

[0038]

[0039] where is the average lithium concentration of the m-th functional area, P m is the total number of pixels in the m-th functional area, q is the index of the pixel, is the lithium concentration inversion value of the q-th pixel in the m-th functional area, m is the index of the functional area;

[0040] Calculate the total lithium amount of each functional area, expressed as

[0041]

[0042] where R m is the total lithium amount of the m-th functional area, is the average lithium concentration of the m-th functional area, A m is the area of the m-th functional area, m is the index of the functional area;

[0043] Calculate the change range of lithium concentration, expressed as

[0044]

[0045] where ΔC m is the change range of lithium concentration of the m-th functional area, is the average lithium concentration of the m-th functional area at the current moment, is the average lithium concentration of the m-th functional area at the previous moment, m is the index of the functional area, v 2is the current moment, v 1 is the previous moment;

[0046] The average lithium concentration, the total lithium amount, and the lithium concentration change range of each functional area are used as partition statistical data, and dynamic monitoring results in the form of a lithium concentration change graph and a statistical table of lithium amount changes in functional areas are generated using the partition statistical data.

[0047] As a preferred solution of the method for monitoring the production capacity of a salt lake lithium ore project based on satellite remote sensing according to the present invention, wherein: a production capacity monitoring report of the salt lake lithium ore project is generated based on the dynamic monitoring results, and the specific steps are as follows.

[0048] Based on the lithium concentration change graph and the statistical table of lithium amount changes in functional areas in the dynamic monitoring results, the lithium concentration distribution, the functional area area, and the lithium amount change range in the partition are extracted;

[0049] Analyze the lithium concentration distribution, the functional area area, and the lithium amount change range in the partition to evaluate the lithium concentration distribution trend, the change of the functional area area, and the change of the total production capacity of the salt lake lithium ore;

[0050] Apply the matplotlib report generation tool to generate a report in PDF format.

[0051] In a second aspect, the present invention provides a device for monitoring the production capacity of a salt lake lithium ore project based on satellite remote sensing, including an image acquisition unit, a salt lake zoning unit, a model acquisition unit, a dynamic monitoring unit, and a production capacity report unit;

[0052] The image acquisition unit is used to acquire satellite remote sensing data and brine parameters, perform preprocessing respectively, and then acquire historical geographical information data;

[0053] The salt lake zoning unit is used to classify the preprocessed multi-spectral image by using the K-means clustering algorithm to obtain a classification result, and generate a salt lake functional area map by comparing with the historical geographical information data;

[0054] The model acquisition unit is used to train a random forest algorithm with the preprocessed brine parameters and satellite remote sensing data, establish a lithium concentration inversion model, regularly acquire new satellite remote sensing data, and obtain the lithium concentration inversion value of the pixels in the regularly received satellite remote sensing data in combination with the salt lake functional area map and the lithium concentration inversion model;

[0055] The dynamic monitoring unit is used to analyze the lithium concentration inversion value to obtain dynamic monitoring results;

[0056] The production capacity report unit is used to generate a production capacity monitoring report of the salt lake lithium ore project based on the dynamic monitoring results.

[0057] In a third aspect, the present invention provides a computer device, including a memory and a processor, where the memory stores a computer program, and: when the computer program is executed by the processor, any step of the method for monitoring the production capacity of a salt lake lithium mine project based on satellite remote sensing as described in the first aspect of the present invention is implemented.

[0058] In a fourth aspect, the present invention provides a computer-readable storage medium, on which a computer program is stored, and: when the computer program is executed by the processor, any step of the method for monitoring the production capacity of a salt lake lithium mine project based on satellite remote sensing as described in the first aspect of the present invention is implemented.

[0059] The beneficial effects of the present invention are as follows: By acquiring and preprocessing satellite remote sensing data and brine parameters, and combining historical geographical information data, efficient integration and quality improvement of multi-source data are achieved, ensuring the accuracy and reliability of subsequent analysis and providing a comprehensive information basis for monitoring. The K-means clustering algorithm is used to classify the preprocessed multi-spectral image and compare it with historical geographical information to generate a functional zoning map of the salt lake, realizing automatic and accurate functional zoning and improving the scientificity of regional division. The preprocessed brine parameters and satellite remote sensing data are used to train the random forest algorithm to establish a lithium concentration inversion model, improving the accuracy and efficiency of lithium concentration prediction and making the dynamic monitoring results more reliable. Regularly updating satellite remote sensing data and combining it with the lithium concentration inversion model can continuously track the change of lithium concentration and timely monitor the actual production capacity of the salt lake lithium mine project. Description of the Drawings

[0060] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for description in the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention, and those of ordinary skill in the art can also obtain other drawings without creative efforts based on these drawings.

[0061] Figure 1 It is a flowchart of the method for monitoring the production capacity of a salt lake lithium mine project based on satellite remote sensing in Embodiment 1.

[0062] Figure 2 It is a flowchart of obtaining dynamic monitoring results in Embodiment 1. Detailed Embodiments

[0063] To make the above objects, features, and advantages of the present invention more obvious and understandable, the detailed embodiments of the present invention will be described in detail below with reference to the drawings in the specification.

[0064] In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention. However, the present invention may be practiced in other ways different from those described herein. Those skilled in the art can make similar extensions without departing from the connotation of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0065] Secondly, the so-called "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation manner of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it an individual or selectively mutually exclusive embodiment with other embodiments.

[0066] Embodiment 1, referring to Figure 1 and Figure 2 , is the first embodiment of the present invention. This embodiment provides a method for monitoring the production capacity of a salt lake lithium ore project based on satellite remote sensing, including the following steps:

[0067] S1. Obtain satellite remote sensing data and brine parameters, and perform preprocessing on them respectively. Then obtain historical geographical information data, including the following steps.

[0068] The satellite remote sensing data includes multi-spectral images, spectral bands, longitude and latitude information, spectral reflectance of pixels, and acquisition time at multiple time points obtained through satellite remote sensing.

[0069] The brine parameters include lithium concentration values at multiple ground sampling points in the salt lake area, longitude and latitude information of the ground sampling points, and sampling time.

[0070] The historical geographical information data includes the partition ranges and geometric shapes of the historical evaporation pond area, salt field area, and natural lake area in the salt lake functional partition.

[0071] The preprocessing includes performing radiometric correction, time series synchronization, resampling, and standardization on the satellite remote sensing data, and performing data cleaning, inverse distance weighted data interpolation, and time series synchronization on the brine parameters.

[0072] Furthermore, using the lithium concentration data collected at different positions in the salt lake and the spectral reflectance of the pixels corresponding to the longitude and latitude information, the multi-spectral image is radiometrically corrected to obtain the spectral reflectance of the pixels of the corrected multi-spectral image. Based on the sampling time of the brine parameters, spline interpolation is used to perform time interpolation on the multi-spectral image to complete the time alignment of the multi-spectral image. Through longitude and latitude registration, the time-aligned multi-spectral image is spatially aligned with the brine parameters to obtain a multi-spectral image after time series synchronization. Using the bilinear interpolation method, the spatial resolution of each band of the multi-spectral image after time series synchronization is unified to the scale of the highest resolution band to obtain a multi-spectral image with consistent spatial resolution. The spectral reflectance of the pixels that are radiometrically corrected and have consistent spatial resolution is Z-score standardized to obtain satellite remote sensing data in a unified format.

[0073] It should be noted that the multi-spectral image can capture the changes in the surface reflectance in different spectral bands and reflect the optical characteristics of the surface materials in the salt lake area. Through multi-time point data, the spatio-temporal changes in the salt lake area can be analyzed. The spectral reflectance is an important indicator reflecting the composition of surface materials and can be used to extract the mineral information covered by the salt lake surface. Lithium resources are usually related to the distribution of surface minerals and brine characteristics in the salt lake. Therefore, the spectral reflectance is the key data. The longitude and latitude information is used to spatially register the remote sensing data with the brine parameters of the ground sampling points, and the time information is used for time series analysis and time alignment of multi-data sources. The lithium concentration is the core indicator for the study of salt lake resources. The ground truth lithium concentration is obtained through sample collection and is used to verify and correct the inversion results of remote sensing images. There are usually different functional zones in the salt lake area (such as evaporation pond area, salt field area, natural lake area), and the historical changes in these zones directly affect the resource distribution and environmental characteristics of the salt lake. By analyzing the historical geographical information, the spatio-temporal change law of lithium resources can be understood, and geographical background support can be provided for the inversion of lithium concentration.

[0074] S2. Use the K-means clustering algorithm to classify the preprocessed multi-spectral image to obtain the classification result, and generate a salt lake functional zoning map by comparing with the historical geographical information data, including the following steps.

[0075] According to the number of partitions in the evaporation pond area, salt field area, and natural lake area, set the clustering parameters of the K-means clustering algorithm to obtain the initialized K-means clustering algorithm. Specifically, according to the requirements of the salt lake functional zoning (evaporation pond area, salt field area, natural lake area), set the number of clustering categories K = 3. Use the random initialization method or the prior knowledge of the multi-spectral image (such as having known classifications for the spectral characteristics of certain regions) to initialize the three initial clustering centers of the K-means clustering algorithm. The clustering centers can be the spectral reflectance values (in vector form) of some pixels in the multi-spectral image, and each clustering center corresponds to a functional zone.

[0076] Use the initialized K-means clustering algorithm to perform clustering analysis on the preprocessed multi-spectral image to obtain a preliminary classification result. Specifically, flatten the preprocessed multi-spectral image into a two-dimensional feature matrix, where the rows represent pixels and the columns represent the spectral reflectance values of each spectral band. For each pixel, calculate the Euclidean distance between its spectral features and all cluster centers. According to the Euclidean distance, assign each pixel to the category corresponding to the nearest cluster center. For each category, calculate the mean of the spectral features of all pixels belonging to that category as the new cluster center. Mark the category of each pixel as its clustering result to generate a preliminary classification result (classification map).

[0077] Use the majority filtering algorithm to smooth the preliminary classification result to obtain a sketch of the functional partition. Specifically, the sliding window size is usually selected as a 3×3 or 5×5 window, and the specific size depends on the resolution of the classification map and the detail requirements of the classification result. For each pixel, extract the pixel categories within the surrounding sliding window. Count the number of each category within the window. Replace the category of the central pixel with the category that appears the most times. If the number of appearances is the same, the current category is preferentially retained. For the boundary area of the classification map, expand the window by mirror filling or zero filling to ensure that all pixels can be processed. After majority filtering, generate a smooth sketch of the functional partition, removing isolated pixels and classification noise.

[0078] Perform a geometric shape comparison analysis between the sketch of the functional partition and the historical geographic information data, identify the different parts and make corrections to obtain the final salt lake functional partition map. Specifically, vectorize the historical data to generate the polygon outline of each functional partition. Perform a spatial overlay analysis between the sketch of the functional partition and the historical geographic information data to calculate the overlapping area and inconsistent areas (such as newly added areas and missing areas) between the two. Identify the pixels in the classification result that exceed the historical partition range (which may be misclassified). Check the pixels within the historical partition range that are not classified as the corresponding functional area (which may be missed). Mark the inconsistent areas graphically to assist subsequent corrections. Correct the different areas through neighborhood analysis. Generate the final salt lake functional partition map from the corrected sketch of the functional partition to ensure that the partition result is consistent with the historical data and reflects new changes at the same time.

[0079] It should be noted that majority filtering is a window-based image smoothing algorithm mainly used to eliminate isolated pixels or noise in the classification result. In the classification map, with each pixel as the center, select a sliding window of a fixed size (such as 3×3 or 5×5), count the number of each category within the window, and replace the category of the central pixel with the category that appears the most times within the window.

[0080] S3. Use the preprocessed brine parameters and satellite remote sensing data to train the random forest algorithm, establish a lithium concentration inversion model, regularly obtain new satellite remote sensing data, and combine the salt lake functional zoning map and the lithium concentration inversion model to obtain the lithium concentration inversion values of the pixels in the regularly received satellite remote sensing data, including the following steps:

[0081] Extract the spectral reflectance of the pixels matching the ground sampling points using the longitude and latitude information to form a feature vector. Specifically, traverse the longitude and latitude information of all ground sampling points, find the multispectral image pixels closest to the geographical location of the sampling points (pixels within the spatial resolution range), and use the nearest neighbor interpolation method for matching. For each sampling point, extract the reflectance values of all spectral bands corresponding to the matching pixels to generate a feature vector.

[0082] Use the lithium concentration values of the ground sampling points as the output targets of the random forest model. Specifically, extract the lithium concentration values of each sampling point from the ground sampling data as the output targets of the model. Match the extracted lithium concentration values with the feature vectors to ensure a one-to-one correspondence between the feature vectors and the output targets.

[0083] Summarize all the feature vectors and output targets to construct a dataset. Specifically, integrate the feature vectors of all sampling points and the corresponding lithium concentration values into a dataset. Prepare the training data and labels, use the feature vectors as the input data of the model, and the lithium concentration values as the output targets of the model. Randomly divide the dataset according to a certain ratio (such as 80% training set, 20% test set) for model training and verification.

[0084] Initialize the parameters of the random forest model according to the dataset size, and use the dataset to train the random forest model to establish a lithium concentration inversion model. Specifically, set the key parameters of the random forest model according to the dataset size, n_estimators (the number of decision trees), set the number of decision trees, and the recommended range is 100 - 500 (less can be set when the dataset is small). max_depth (the maximum depth of the tree), set the depth of each tree to avoid overfitting. Select according to the data complexity (such as 5 - 15). min_samples_split (the minimum number of samples for splitting), set the minimum number of samples when the decision tree splits nodes, and the recommended value is 2 or 5. random_state (random seed), to ensure the reproducibility of the results, set the random seed.

[0085] Regularly obtain satellite remote sensing data covering the salt lake area, and input the spectral reflectance of each pixel in the regularly received satellite remote sensing data into the lithium concentration inversion model to obtain the lithium concentration inversion value of the pixel, expressed as:

[0086]

[0087] Wherein, is the inversion value of lithium concentration of the pixel, T is the number of decision trees, t is the index of the decision tree, and N t is the number of leaf node samples in the t-th decision tree, i is the index of the feature vector, x is the feature vector, and L t is the leaf node in the t-th decision tree, Ⅰ[x i ∈L t is the indicator function, and y i is the lithium concentration value corresponding to the i-th feature vector in the leaf node.

[0088] It should be noted that the nearest neighbor interpolation method is simple and direct, and can quickly find the pixel closest to the sampling point. This method has good results in the case of fewer sampling points and higher spatial resolution, avoiding the increase in computational complexity caused by complex interpolation. Random forest is an ensemble learning method based on decision trees, which has the advantages of strong adaptability to high-dimensional data, high robustness to noise data, and strong interpretability.

[0089] S4. Analyze the inversion value of lithium concentration to obtain the dynamic monitoring results, including the following steps:

[0090] Use the lithium concentration inversion model to output the inversion value of lithium concentration of the pixel, bind the inversion value of lithium concentration of each pixel with the longitude and latitude, and generate the spatial distribution map of lithium concentration in the salt lake area. Specifically, extract the spectral reflectance feature vector of each pixel. Input the feature vectors of all pixels into the lithium concentration inversion model. Use the bound longitude and latitude and the inversion value of lithium concentration to generate the spatial distribution map of lithium concentration in the salt lake area through spatial interpolation or direct visualization, and display the spatial distribution characteristics of lithium concentration.

[0091] According to the spatial distribution map of lithium concentration, calculate the average value of lithium concentration of each functional area by functional area, expressed as

[0092]

[0093] where is the average value of lithium concentration of the m-th functional area, and P m is the total number of pixels in the m-th functional area, q is the index of the pixel, is the inversion value of lithium concentration of the q-th pixel in the m-th functional area, and m is the index of the functional area.

[0094] Calculate the total lithium amount of each functional area, expressed as

[0095]

[0096] where R m is the total lithium amount of the m-th functional area, is the average value of lithium concentration of the m-th functional area, and Am is the area of the m-th functional partition, where m is the index of the functional partition.

[0097] Calculate the change range of lithium concentration, denoted as

[0098]

[0099] where ΔC m is the change range of lithium concentration in the m-th functional partition, is the average lithium concentration in the m-th functional partition at the current moment, is the average lithium concentration in the m-th functional partition at the previous moment, where m is the index of the functional partition, and v 2 is the current moment, and v 1 is the previous moment.

[0100] The average lithium concentration, the total lithium amount, and the change range of lithium concentration in each functional partition are used as partition statistical data, and dynamic monitoring results in the form of a lithium concentration change map and a statistical table of lithium amount change in functional partitions are generated. Specifically, the change range of lithium concentration is visualized by functional partition to generate a lithium concentration change map, intuitively showing the changes in each partition. The average lithium concentration, total lithium amount, and change range of lithium concentration in each functional partition are summarized to generate a table.

[0101] It should be noted that the lithium concentration change map can provide intuitive spatial distribution information, facilitating the quick identification of areas with large changes in lithium concentration and supporting the dynamic monitoring and decision-making of resource management departments. The statistical table of lithium amount change in functional partitions digitally presents the changes in lithium concentration and total lithium amount, facilitating subsequent analysis and report writing, and supporting the long-term tracking and trend analysis of lithium resource reserves.

[0102] S5. Generate a production capacity monitoring report for the salt lake lithium mine project based on the dynamic monitoring results, including the following steps:

[0103] Based on the lithium concentration change map and the statistical table of lithium amount change in functional partitions in the dynamic monitoring results, extract the lithium concentration distribution, the area of functional partitions, and the change range of lithium amount in partitions. Specifically, from the lithium concentration change map, extract the inversion value of pixel lithium concentration in each functional partition and the average lithium concentration of the partition. Classify the extracted partition lithium concentrations by interval, and count the number of pixels and the distribution range in each concentration interval. Read the total area of each functional partition from the functional partition map. If there are changes in the functional partition (such as the addition of an evaporation pond or the adjustment of the salt field area), calculate the area change amount.

[0104] Analyze the lithium concentration distribution, the area of functional partitions, and the variation range of lithium quantity in each partition to evaluate the lithium concentration distribution trend, the change of functional partition area, and the change of total production capacity of the salt lake lithium mine. Specifically, analyze the lithium concentration distribution in the entire salt lake area, observe the change from high to low concentration, and judge the main distribution area of lithium resource reserves. Compare the average lithium concentration of different functional partitions to identify the areas where high-concentration lithium resources are concentrated (such as evaporation ponds, salt pans, etc.). For example, whether the evaporation ponds are gradually concentrating high-lithium-concentration resources; whether the lithium concentration in the natural lake area shows a downward trend. Compare the area changes of functional partitions to judge whether areas such as salt pans and evaporation ponds have an expanding or shrinking trend. If the area changes significantly, combined with the lithium concentration distribution, evaluate whether the lithium concentration in the new area meets expectations and judge the development effect. Analyze the variation range of lithium quantity in each partition, and focus on the change trend of lithium resource production capacity. If the lithium quantity in some partitions increases, it may indicate that good results have been achieved in the new development area; if the lithium quantity decreases, combined with the concentration change and area change, judge whether it is related to over-exploitation or natural changes.

[0105] Use the matplotlib report generation tool to generate a report in PDF format. Specifically, take the lithium concentration change graph as the core content of the report, mark the high, medium, and low concentration areas, represent them with color gradients, and highlight the key partitions. Use bar charts to show the area changes of each functional partition. Use line charts or bar charts to show the variation range of lithium quantity in each partition, and at the same time mark the change trend of the total lithium quantity in the salt lake. Insert a statistical table of the change of lithium quantity in functional partitions in the report to summarize the average lithium concentration, lithium quantity, and variation range of each partition. Generate a report in PDF format through programming using the matplotlib report generation tool.

[0106] It should be noted that the PDF report provides a complete analysis result of lithium resources, which is convenient for archiving, sharing, and decision-making. The charts and tables intuitively show the lithium concentration distribution, the change of functional partition area, and the change of total production capacity, providing support for government management departments and enterprises to formulate development plans.

[0107] This embodiment also provides a production capacity monitoring device for salt lake lithium mines based on satellite remote sensing, including: an image acquisition unit, a salt lake zoning unit, a model acquisition unit, a dynamic monitoring unit, and a production capacity report unit;

[0108] The image acquisition unit is used to acquire satellite remote sensing data and brine parameters, preprocess them respectively, and then acquire historical geographical information data;

[0109] The salt lake zoning unit is used to classify the preprocessed multi-spectral image by using the K-means clustering algorithm to obtain a classification result, and generate a salt lake functional zoning map by comparing with historical geographical information data;

[0110] A model acquisition unit is configured to train a random forest algorithm using the preprocessed brine parameters and satellite remote sensing data, establish a lithium concentration inversion model, regularly obtain new satellite remote sensing data, and combine the salt lake functional zoning map and the lithium concentration inversion model to obtain the lithium concentration inversion values of the pixels in the regularly received satellite remote sensing data.

[0111] A dynamic monitoring unit is configured to analyze the lithium concentration inversion values to obtain dynamic monitoring results.

[0112] A production capacity reporting unit is configured to generate a production capacity monitoring report for the salt lake lithium ore project based on the dynamic monitoring results.

[0113] This embodiment also provides a computer device applicable to the case of the production capacity monitoring method for the salt lake lithium ore project based on satellite remote sensing, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the production capacity monitoring method for the salt lake lithium ore project based on satellite remote sensing as proposed in the above embodiment.

[0114] The computer device may be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen, and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, a carrier network, NFC (Near Field Communication), or other technologies. The display screen of the computer device may be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device may be a touch layer covered on the display screen, or a button, a trackball, or a touchpad provided on the outer shell of the computer device, or an external keyboard, a touchpad, or a mouse, etc.

[0115] This embodiment also provides a storage medium, on which a computer program is stored. When the program is executed by a processor, it implements the method for monitoring the production capacity of a salt lake lithium mine project based on satellite remote sensing as proposed in the above embodiment; the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (Static Random Access Memory, abbreviated as SRAM), electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory, abbreviated as EEPROM), erasable programmable read-only memory (Erasable Programmable Read Only Memory, abbreviated as EPROM), programmable read-only memory (Programmable Red-Only Memory, abbreviated as PROM), read-only memory (Read-Only Memory, abbreviated as ROM), magnetic memory, flash memory, magnetic disk or optical disc.

[0116] In summary, the present invention obtains and preprocesses satellite remote sensing data and brine parameters, combines historical geographical information data, realizes the efficient integration and quality improvement of multi-source data, ensures the accuracy and reliability of subsequent analysis, and provides a comprehensive information basis for monitoring. The K-means clustering algorithm is used to classify the preprocessed multi-spectral image, and it is compared with historical geographical information to generate a functional zoning map of the salt lake, realizing automatic and accurate functional zoning and improving the scientificity of regional division. The preprocessed brine parameters and satellite remote sensing data are used to train the random forest algorithm to establish a lithium concentration inversion model, improving the accuracy and efficiency of lithium concentration prediction and making the dynamic monitoring results more reliable. Regularly updating the satellite remote sensing data and combining it with the lithium concentration inversion model can continuously track the change of lithium concentration and timely monitor the actual production capacity of the salt lake lithium mine project.

[0117] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered within the scope of the claims of the present invention.

Claims

1. A method for monitoring the production capacity of a salt lake lithium mine project based on satellite remote sensing, characterized in that: include, Obtain satellite remote sensing data and brine parameters, pre-process them respectively, and then obtain historical geographic information data; The K-means clustering algorithm is used to classify the preprocessed multispectral images to obtain the classification results, and the functional zoning map of the salt lake is generated by comparing it with the historical geographic information data; The pre-processed brine parameters and satellite remote sensing data are used to train the random forest algorithm, establish a lithium concentration inversion model, regularly obtain new satellite remote sensing data, and combine the salt lake functional zoning map and the lithium concentration inversion model to obtain the lithium concentration inversion value of the pixel in the regularly received satellite remote sensing data; Analyze the lithium concentration inversion value to obtain dynamic monitoring results; Based on the dynamic monitoring results, a capacity monitoring report for the salt lake lithium mine project is generated.

2. The method for monitoring the production capacity of a salt lake lithium mine project based on satellite remote sensing according to claim 1, characterized in that: The satellite remote sensing data includes multispectral images, spectral bands, longitude and latitude information, spectral reflectance of pixels and acquisition time at multiple time points acquired through satellite remote sensing; The brine parameters include lithium concentration values ​​at multiple ground sampling points within the salt lake area, longitude and latitude information of the ground sampling points, and sampling time; The historical geographic information data include the zoning range and geometric shape of the historical evaporation pond area, salt pan area and natural lake area of ​​the salt lake functional zoning; The preprocessing includes radiation correction, time series synchronization, resampling and standardization of satellite remote sensing data, data cleaning, inverse distance weighted data interpolation and time series synchronization of brine parameters.

3. The method for monitoring the production capacity of a salt lake lithium mine project based on satellite remote sensing according to claim 2, characterized in that: The specific steps of performing radiation correction, time series synchronization, resampling and standardization on satellite remote sensing data are as follows: Using the lithium concentration data collected at different locations of the salt lake and the spectral reflectance of the pixels at the points corresponding to the longitude and latitude information, the multispectral image is radiometrically corrected to obtain the spectral reflectance of the pixels of the corrected multispectral image; Taking the sampling time of brine parameters as the benchmark, spline interpolation is used to perform time interpolation on multispectral images to complete the time alignment of multispectral images. Through longitude and latitude registration, the time-aligned multispectral images are spatially aligned with the brine parameters to obtain the multispectral images after time series synchronization; Bilinear interpolation method is used to unify the spatial resolution of each band of the multispectral image after time series synchronization to the scale of the highest resolution band, so as to obtain a multispectral image with consistent spatial resolution. The spectral reflectance of pixels with consistent spatial resolution after radiation correction is normalized by Z-score to obtain satellite remote sensing data in a unified format.

4. The method for monitoring the production capacity of a salt lake lithium mine project based on satellite remote sensing as claimed in claim 3, characterized in that: The K-means clustering algorithm is used to classify the preprocessed multispectral images to obtain the classification results, and the functional zoning map of the salt lake is generated by comparing it with the historical geographic information data. The specific steps are as follows: According to the number of partitions of the evaporation pond area, salt pan area and natural lake area, the clustering parameters of the K-means clustering algorithm are set to obtain the initialized K-means clustering algorithm; Use the initialized K-means clustering algorithm to perform cluster analysis on the preprocessed multispectral images to obtain preliminary classification results; The majority filtering algorithm is used to smooth the preliminary classification results to obtain a functional zoning sketch; The geometric shapes of the functional zoning sketch and the historical geographic information data were compared and analyzed, the differences were identified and corrected, and the final salt lake functional zoning map was obtained.

5. The method for monitoring the production capacity of a salt lake lithium mine project based on satellite remote sensing according to claim 4, characterized in that: The pre-processed brine parameters and satellite remote sensing data are used to train the random forest algorithm, establish a lithium concentration inversion model, regularly obtain new satellite remote sensing data, and combine the salt lake functional zoning map and the lithium concentration inversion model to obtain the lithium concentration inversion value of the pixel in the regularly received satellite remote sensing data. The specific steps are as follows: The spectral reflectance of the pixels matching the ground sampling points is extracted using the longitude and latitude information to form a feature vector; The lithium concentration values ​​of the ground sampling points were used as the output targets of the random forest model; Summarize all feature vectors and output targets to build a dataset; Initialize the parameters of the random forest model according to the size of the data set, and use the data set to train the random forest model to establish a lithium concentration inversion model; Satellite remote sensing data covering the salt lake area is obtained regularly, and the spectral reflectance of each pixel in the regularly received satellite remote sensing data is input into the lithium concentration inversion model to obtain the lithium concentration inversion value of the pixel, which is expressed as, in, is the lithium concentration inversion value of the pixel, T is the number of decision trees, t is the index of the decision tree, N t is the number of leaf node samples in the t-th decision tree, i is the index of the feature vector, x is the feature vector, L t is the leaf node in the t-th decision tree, I[x i ∈L t ] is the indicator function, y i is the lithium concentration value corresponding to the i-th eigenvector in the leaf node.

6. The method for monitoring the production capacity of a salt lake lithium mine project based on satellite remote sensing according to claim 5, characterized in that: Analyze the lithium concentration inversion value and obtain dynamic monitoring results. The specific steps are as follows: The lithium concentration inversion model is used to output the lithium concentration inversion value of each pixel, and the lithium concentration inversion value of each pixel is bound to the longitude and latitude to generate a spatial distribution map of lithium concentration in the salt lake area; According to the spatial distribution map of lithium concentration, the mean lithium concentration of each functional zone is calculated according to the functional zone, which is expressed as: in, is the mean lithium concentration of the mth functional partition, P m is the total number of pixels in the mth functional partition, q is the index of the pixel, is the lithium concentration inversion value of the qth pixel in the mth functional partition, where m is the index of the functional partition; The total lithium content of each functional partition is calculated and expressed as, Among them, R m is the total lithium content of the mth functional partition, is the mean lithium concentration of the mth functional partition, A m is the area of ​​the mth functional partition, and m is the index of the functional partition; The range of lithium concentration change is calculated and expressed as, Where, ΔC m is the variation range of lithium concentration in the mth functional partition, is the mean lithium concentration of the mth functional partition at the current moment, is the mean lithium concentration of the mth functional partition at the previous moment, m is the index of the functional partition, v2 is the current moment, and v1 is the previous moment; The mean lithium concentration of each functional zone, the total lithium content of each functional zone and the range of lithium concentration change are used as zone statistical data. The zone statistical data are used to generate dynamic monitoring results in the form of lithium concentration change graphs and functional zone lithium content change statistical tables.

7. The method for monitoring the production capacity of a salt lake lithium mine project based on satellite remote sensing according to claim 6, characterized in that: According to the dynamic monitoring results, the capacity monitoring report of the salt lake lithium mine project is generated. The specific steps are as follows: Based on the lithium concentration change graph and the functional zone lithium content change statistical table in the dynamic monitoring results, the lithium concentration distribution, functional zone area and zone lithium content change range are extracted; Analyze the distribution of lithium concentration, functional zoning area and the change range of lithium content in each zoning area, and evaluate the lithium concentration distribution trend, functional zoning area change and total production capacity change of salt lake lithium mines; Use the matplotlib report generation tool to generate reports in PDF format.

8. A device for monitoring the production capacity of a salt lake lithium mine project based on satellite remote sensing, based on the method for monitoring the production capacity of a salt lake lithium mine project based on satellite remote sensing according to any one of claims 1 to 7, characterized in that: Including, image acquisition unit, salt lake zoning unit, model acquisition unit, dynamic monitoring unit and capacity reporting unit; The image acquisition unit is used to acquire satellite remote sensing data and brine parameters, and pre-process them respectively, and then acquire historical geographic information data; The salt lake zoning unit is used to classify the preprocessed multispectral image using a K-means clustering algorithm to obtain a classification result, and generate a salt lake functional zoning map by comparing it with historical geographic information data; The model acquisition unit is used to train a random forest algorithm using pre-processed brine parameters and satellite remote sensing data, establish a lithium concentration inversion model, regularly acquire new satellite remote sensing data, and combine the salt lake functional zoning map and the lithium concentration inversion model to obtain the lithium concentration inversion value of the pixel in the regularly received satellite remote sensing data; The dynamic monitoring unit is used to analyze the lithium concentration inversion value to obtain the dynamic monitoring result; The production capacity reporting unit is used to generate a production capacity monitoring report for the salt lake lithium mine project based on the dynamic monitoring results.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method for monitoring the production capacity of a salt lake lithium mine project based on satellite remote sensing as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for monitoring the production capacity of a salt lake lithium mine project based on satellite remote sensing as described in any one of claims 1 to 7 are implemented.

Images