Graphite purification method and system based on gradient density separation

Abstract

The invention provides a graphite purification method and system based on gradient density separation, and relates to the technical field of graphite purification, and the method comprises the following steps: carrying out physical property measurement on a cleaned graphite ore sample; a Bayesian normalization layer is introduced, and liquid gradient concentration distribution noise quantification simulation is carried out; traversing the graphite purification scene template library to identify a graphite purification strategy; performing gradient density separation and purification on the graphite ore sample, and performing matching identification on the graphite purification scene template library in combination with physical characteristic information of the graphite ore sample and a liquid gradient concentration distribution monitoring result in the purification process; and a matched graphite purification strategy template is obtained to optimize the gradient density separation and purification process. According to the method and the device, the technical problem of unstable purification effect caused by noise interference and parameter uncertainty in the graphite purification process in the prior art is solved, and the uncertainty in the graphite purification process is quantified through Bayesian, so that the stability and the reliability of the graphite separation effect are improved.

Description

Technical Field

[0001] This application relates to the field of graphite purification technology, and in particular to a graphite purification method and system based on gradient density separation. Background Technology

[0002] The purification effect of existing graphite purification processes highly depends on the precise control of key process parameters such as temperature, atmosphere, and reactant concentration. However, in actual industrial production, due to the inherent differences in raw material sources, fluctuations in the quality of auxiliary materials, and significant uncertainties in the accuracy of temperature measurement and control under high-temperature environments, it is difficult to maintain precise consistency of key parameters during the purification process. These factors collectively constitute continuous noise interference and parameter inaccuracies, making it difficult to maintain the optimal reaction window during purification. This results in drastic fluctuations in the purity of products from different batches and even within the same batch, severely affecting the stability of graphite purification results.

[0003] In summary, the existing technology suffers from unstable purification results due to noise interference and parameter uncertainties during the graphite purification process. Summary of the Invention

[0004] The purpose of this application is to provide a graphite purification method and system based on gradient density separation, in order to solve the technical problem in the prior art that the purification effect is unstable due to noise interference and parameter uncertainty in the graphite purification process.

[0005] In view of the above problems, this application provides a graphite purification method and system based on gradient density separation.

[0006] Firstly, this application provides a graphite purification method based on gradient density separation. This method is implemented using a graphite purification system based on gradient density separation. The method includes: screening a graphite sample and performing ultrasonic cleaning to remove surface impurities; measuring the physical properties of the cleaned graphite sample to obtain its physical characteristics; introducing a Bayesian normalization layer; and combining historical graphite purification data to perform noise quantification simulation of the liquid gradient concentration distribution, thus constructing a graphite purification system. A graphite purification scenario template library is established; the graphite purification scenario template library is traversed to identify graphite purification strategies, and a graphite purification strategy template library is constructed; the graphite ore sample is subjected to gradient density separation and purification, and the graphite purification scenario template library is matched and identified by combining the physical characteristic information of the graphite ore sample and the monitoring results of liquid gradient concentration distribution during the purification process to obtain matching graphite purification scenario templates; based on the matching graphite purification scenario templates, the graphite purification strategy template library is searched to obtain matching graphite purification strategy templates to optimize the gradient density separation and purification process.

[0007] Optionally, the historical graphite purification data is sorted according to physical properties, liquid gradient concentration distribution, and purification indicators to obtain a historical physical properties-liquid gradient concentration distribution-purification indicator data set; based on physical properties, the historical physical properties-liquid gradient concentration distribution-purification indicator data set is aggregated to obtain multiple aggregated historical physical properties-liquid gradient concentration distribution-purification indicator data sets; noise fluctuation scales are identified for each of the multiple aggregated historical physical properties-liquid gradient concentration distribution-purification indicator data sets to obtain multiple noise fluctuation scales; a Bayesian normalization layer is introduced to randomly perturb and expand the multiple aggregated historical physical properties-liquid gradient concentration distribution-purification indicator data sets according to the multiple noise fluctuation scales to obtain multiple expanded aggregated historical physical properties-liquid gradient concentration distribution-purification indicator data sets; by centrally sorting the data of the multiple expanded aggregated historical physical properties-liquid gradient concentration distribution-purification indicator data sets, multiple graphite purification scenario template sets are determined, and the graphite purification scenario template library is constructed.

[0008] Optionally, the liquid gradient concentration distribution fluctuation analysis is performed by traversing the multiple sets of historical polymerization physical characteristics-liquid gradient concentration distribution-purification index data to determine multiple liquid gradient concentration distribution fluctuation factors; the purification index fluctuation analysis is performed by traversing the multiple sets of historical polymerization physical characteristics-liquid gradient concentration distribution-purification index data to determine multiple purification index follow-up factors; and noise fluctuation scale identification is performed based on the multiple liquid gradient concentration distribution fluctuation factors and the multiple purification index follow-up factors to obtain multiple noise fluctuation scales.

[0009] Optionally, the liquid gradient concentration distribution and purification index data in the plurality of polymerization history physical characteristics-liquid gradient concentration distribution-purification index data sets are randomly perturbed according to the plurality of noise fluctuation scales through the Bayesian normalization layer to obtain a plurality of initial expanded polymerization history physical characteristics-liquid gradient concentration distribution-purification index data sets; the initial expanded polymerization history physical characteristics-liquid gradient concentration distribution-purification index data with an approximation greater than a preset approximation threshold in the plurality of initial expanded polymerization history physical characteristics-liquid gradient concentration distribution-purification index data sets are mean-processed to obtain the plurality of expanded polymerization history physical characteristics-liquid gradient concentration distribution-purification index data sets.

[0010] Optionally, Gaussian noise is generated based on the multiple noise fluctuation scales through the Bayesian normalization layer, and the liquid gradient concentration distribution and purification index data in the multiple sets of polymerization history physical properties-liquid gradient concentration distribution-purification index data are adjusted according to the generation results to obtain the multiple sets of initial expanded polymerization history physical properties-liquid gradient concentration distribution-purification index data.

[0011] Optionally, a graphite purification strategy recognizer is pre-constructed, wherein the graphite purification strategy recognizer includes an input layer, a convolutional layer, and an output layer; the graphite purification strategy recognizer is used to perform strategy recognition on graphite purification scene templates in the graphite purification scene template library to obtain the graphite purification strategy template library.

[0012] Optionally, using the physical characteristics of the graphite ore sample and the monitoring results of the liquid gradient concentration distribution during the purification process as indexes, cosine similarity calculations are performed on the graphite purification scene templates in the graphite purification scene template library, and the graphite purification scene template corresponding to the maximum calculated value is extracted to obtain a matching graphite purification scene template.

[0013] Optionally, the gradient density separation and purification process is continuously monitored to obtain a sequence of continuous monitoring results for the liquid gradient concentration distribution; the sequence of continuous monitoring results for the liquid gradient concentration distribution is compared with the corresponding liquid gradient concentration distribution in the matching graphite purification scenario template to obtain a deviation sequence; the deviation sequence is analyzed from two dimensions: change trend and concentrated data, to determine the deviation influence coefficient; and the matching graphite purification strategy template is optimized based on the deviation influence coefficient.

[0014] Optionally, the mean shift algorithm is used to identify concentrated data in the deviation sequence to obtain concentrated deviation; the deviation sequence is analyzed for trend characteristics to obtain deviation trend characteristics; and the concentrated deviation is corrected based on the deviation trend characteristics to obtain the deviation influence coefficient.

[0015] Secondly, this application also provides a graphite purification system based on gradient density separation, used to execute a graphite purification method based on gradient density separation as described in the first aspect. The graphite purification system based on gradient density separation includes: a physical property measurement module for screening graphite samples and performing ultrasonic cleaning to remove surface impurities, measuring the physical properties of the cleaned graphite samples to obtain physical characteristic information of the graphite samples; and a noise quantization simulation module for introducing a Bayesian normalization layer, combining historical graphite purification data to perform noise quantization simulation of liquid gradient concentration distribution, and constructing a graphite purification scenario template library. The system includes: a purification strategy identification module for traversing the graphite purification scenario template library to identify graphite purification strategies and construct the graphite purification strategy template library; a matching identification module for performing gradient density separation purification on the graphite ore sample and, in conjunction with the physical characteristics of the graphite ore sample and the monitoring results of liquid gradient concentration distribution during the purification process, matching identification on the graphite purification scenario template library to obtain matching graphite purification scenario templates; and a strategy retrieval module for retrieving the graphite purification strategy template library based on the matching graphite purification scenario templates to obtain matching graphite purification strategy templates for optimizing the gradient density separation purification process.

[0016] One or more technical solutions provided in this application have at least the following beneficial effects: Graphite samples are screened and ultrasonically cleaned to remove surface impurities. The cleaned graphite samples are then subjected to physical property measurements to obtain their physical characteristics. A Bayesian normalization layer is introduced, and historical graphite purification data is used to simulate the noise quantification of liquid gradient concentration distribution, constructing a graphite purification scenario template library. This library is then traversed to identify graphite purification strategies, constructing a graphite purification strategy template library. The graphite samples are then purified using gradient density separation. Based on the physical characteristics of the graphite samples and the monitoring results of the liquid gradient concentration distribution during purification, the graphite purification scenario template library is matched to obtain matching graphite purification scenario templates. Finally, the matching graphite purification scenario templates are used to search the graphite purification strategy template library to obtain matching graphite purification strategy templates to optimize the gradient density separation purification process. In other words, by combining the Bayesian method with gradient density separation technology, the graphite purification process can maintain high stability and reliability in experimental environments with significant uncertainties.

[0017] The above description is merely an overview of the technical solution of this application. To better understand the technical means of this application and to facilitate its implementation according to the description, and to make the above and other objects, features, and advantages of this application more apparent, specific embodiments of this application are described below. It should be understood that the content described in this section is not intended to identify key or important features of the embodiments of this application, nor is it intended to limit the scope of this application. Other features of this application will become readily apparent through the following description. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely exemplary. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0019] Figure 1 This is a schematic diagram of the process of a graphite purification method based on gradient density separation according to this application.

[0020] Figure 2 This is a schematic diagram of the graphite purification system based on gradient density separation according to this application.

[0021] Figure labeling: 11 Physical property measurement module, 12 Noise quantization simulation module, 13 Purification strategy identification module, 14 Matching identification module, 15 Strategy retrieval module. Detailed Implementation

[0022] This application provides a graphite purification method and system based on gradient density separation, which solves the technical problem in existing technologies where unstable purification results are caused by noise interference and parameter uncertainties during the graphite purification process. By combining Bayesian methods with gradient density separation technology, the graphite purification process can maintain high stability and reliability in experimental environments with high uncertainty.

[0023] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. It should be understood that this application is not limited to the exemplary embodiments described herein. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application. It should also be noted that, for ease of description, only the parts related to this application are shown in the accompanying drawings, not all of them.

[0024] Example 1, please refer to the appendix. Figure 1This application provides a graphite purification method based on gradient density separation. The method is executed using a gradient density separation-based graphite purification system and specifically includes the following steps: Graphite ore samples were screened and ultrasonically cleaned to remove surface impurities. The cleaned graphite ore samples were then subjected to physical property measurements to obtain physical characteristic information of the graphite ore samples.

[0025] Specifically, the graphite ore sample is screened to remove impurity particles, and ultrasonic cleaning is used to remove surface impurities. Graphite ore sample screening refers to selecting suitable graphite ore samples from the raw ore for purification based on standards such as particle size, purity, and chemical composition to ensure that the selected samples meet purification requirements. The carbon content of the graphite ore sample usually needs to meet a certain standard to ensure the effectiveness of subsequent purification. For lower-quality ores, preliminary physical screening is required. The screened graphite ore sample is placed in an ultrasonic cleaning device and cleaned with deionized water. Ultrasonic vibration removes dust, grease, and other impurities from the surface of the graphite ore sample, ensuring the purity of the cleaned sample and preventing impurities from affecting subsequent physical property testing and purification processes. For example, in ultrasonic cleaning experiments, an ultrasonic vibration frequency of 40-50 kHz is used, and the cleaning time is 15-20 minutes. During this process, the impurity removal rate on the surface of the graphite ore sample can typically reach over 90%.

[0026] The cleaned graphite ore sample is then further tested to determine its physical properties, yielding information on its physical characteristics (such as density and particle size). Specific surface area and particle size distribution are typically measured using a BET surface area analyzer and a laser particle size analyzer, while density can be measured using gas adsorption or water removal methods. For example, suppose the specific surface area of ​​the cleaned graphite ore sample is measured to be 25 m² / g, and the particle size distribution is 10–50 μm.

[0027] A Bayesian normalization layer is introduced, and historical graphite purification data is combined to simulate the noise quantization of liquid gradient concentration distribution, thereby constructing a graphite purification scenario template library.

[0028] Furthermore, this application also includes the following steps: First, the historical graphite purification data is sorted according to physical properties, liquid gradient concentration distribution, and purification indicators to obtain a historical physical properties-liquid gradient concentration distribution-purification indicator data set. Second, the historical physical properties-liquid gradient concentration distribution-purification indicator data set is aggregated based on physical properties to obtain multiple aggregated historical physical properties-liquid gradient concentration distribution-purification indicator data sets. Third, noise fluctuation scales are identified for each of the multiple aggregated historical physical properties-liquid gradient concentration distribution-purification indicator data sets to obtain multiple noise fluctuation scales. Fourth, a Bayesian normalization layer is introduced to randomly perturb and expand the multiple aggregated historical physical properties-liquid gradient concentration distribution-purification indicator data sets according to the multiple noise fluctuation scales to obtain multiple expanded aggregated historical physical properties-liquid gradient concentration distribution-purification indicator data sets. Fifth, multiple graphite purification scenario template sets are determined by centrally sorting the data of the multiple expanded aggregated historical physical properties-liquid gradient concentration distribution-purification indicator data sets, and the graphite purification scenario template library is constructed.

[0029] Furthermore, this application also includes the following steps: traversing the multiple sets of polymerization historical physical characteristics-liquid gradient concentration distribution-purification index data to perform liquid gradient concentration distribution fluctuation analysis, and determining multiple liquid gradient concentration distribution fluctuation factors; traversing the multiple sets of polymerization historical physical characteristics-liquid gradient concentration distribution-purification index data to perform purification index fluctuation analysis, and determining multiple purification index follow-up factors; and identifying noise fluctuation scales based on the multiple liquid gradient concentration distribution fluctuation factors and the multiple purification index follow-up factors to obtain multiple noise fluctuation scales.

[0030] Furthermore, this application also includes the following steps: The liquid gradient concentration distribution and purification index data in the plurality of polymerization history physical characteristics-liquid gradient concentration distribution-purification index data sets are randomly perturbed according to the plurality of noise fluctuation scales through the Bayesian normalization layer to obtain a plurality of initial expanded polymerization history physical characteristics-liquid gradient concentration distribution-purification index data sets; The initial expanded polymerization history physical characteristics-liquid gradient concentration distribution-purification index data with an approximation greater than a preset approximation threshold in the plurality of initial expanded polymerization history physical characteristics-liquid gradient concentration distribution-purification index data sets are averaged to obtain the plurality of expanded polymerization history physical characteristics-liquid gradient concentration distribution-purification index data sets.

[0031] Furthermore, this application also includes the following steps: generating Gaussian noise based on the multiple noise fluctuation scales through the Bayesian normalization layer, and adjusting the liquid gradient concentration distribution and purification index data in the multiple sets of polymerization history physical properties-liquid gradient concentration distribution-purification index data respectively according to the generation results, to obtain the multiple sets of initial expanded polymerization history physical properties-liquid gradient concentration distribution-purification index data.

[0032] Specifically, historical graphite purification experimental data is collected, including the physical properties of graphite samples (such as density, particle size, and morphology) and experimental parameters (such as liquid gradient concentration, centrifugal force, and temperature) under different experimental conditions. This historical data includes experimental results (such as graphite purity, purification efficiency, and impurity removal rate) under different purification methods, as well as specific values ​​for various parameters. The collected historical graphite purification data is then cleaned to remove missing or inconsistent values. Finally, the cleaned historical graphite purification data is standardized to ensure comparability under different experimental conditions and to provide a unified input format for subsequent Bayesian modeling.

[0033] Physical properties are the basic physical parameters of graphite ore samples, such as particle size, specific surface area, density, and electrical conductivity. During the graphite purification process, the concentration of the liquid may have a gradient distribution in different regions, affecting the separation efficiency of graphite. The liquid gradient concentration distribution is the spatial distribution of liquid concentration, which directly affects the reaction of the graphite ore sample and the purification effect during the purification process. Purification indicators are indicators used to measure the purity or purification efficiency of graphite during the purification process, such as graphite purity, impurity removal rate, and purification efficiency.

[0034] Historical graphite purification data was systematically organized based on physical properties, liquid gradient concentration distribution, and purification indicators. Different types of data were grouped into a unified dataset, forming a historical physical properties-liquid gradient concentration distribution-purification indicator dataset. This dataset is a composite dataset encompassing physical properties, liquid gradient concentration distribution, and purification indicators. Each dataset consists of three parts: physical properties, liquid gradient concentration distribution, and purification indicators. For example, a complete historical dataset includes physical properties: particle size range (10-50 μm), specific surface area (25 m² / g), density (2.3 g / cm³); liquid gradient concentration distribution increasing from 10% solvent concentration to 50%; and purification indicators: graphite purity increasing from 55% to 85%, and impurity removal rate reaching 45%.

[0035] Based on the physical properties of graphite ore samples, historical physical properties-liquid gradient concentration distribution-purification index data sets were aggregated into similar datasets. Data sets with similar physical properties were then grouped together to form multiple aggregated historical physical properties-liquid gradient concentration distribution-purification index data sets. The purpose of this similar aggregation was to identify data groups with similar physical properties.

[0036] During graphite purification, fluctuations in process parameters and external interference can lead to uncertainties or noise fluctuations. Noise fluctuation scales were identified and the amplitude and range of noise fluctuations were quantified for multiple datasets of historical polymerization physical properties, liquid gradient concentration distribution, and purification indicators.

[0037] Specifically, this study analyzes the fluctuations in liquid gradient concentration distribution across multiple datasets of historical physical properties, liquid gradient concentration distribution, and purification indicators during the purification process of different graphite ore samples. This analysis identifies multiple fluctuation factors in liquid gradient concentration distribution. In graphite purification, the change in liquid concentration at different locations and times typically exhibits a gradient distribution. Liquid gradient concentration distribution fluctuation analysis involves statistically analyzing these concentration gradient changes to identify the patterns of fluctuation in liquid concentration distribution and understand the impact of concentration fluctuations on purification efficiency. Fluctuation factors in liquid gradient concentration distribution refer to the specific characteristics or causes of liquid concentration changes. For example, factors such as temperature changes, liquid flow rate, or solubility can all lead to fluctuations in liquid concentration. Identifying these fluctuation factors helps to understand which factors influence liquid concentration fluctuations.

[0038] Similarly, by traversing multiple datasets of historical physical properties, liquid gradient concentration distribution, and purification index data, a fluctuation analysis of purification indexes is conducted. Fluctuations in purification indexes are typically related to various factors during the purification process, particularly liquid concentration, graphite sample physical properties, and temperature. This process identifies multiple influencing factors for each purification index. Analyzing the fluctuations of various indices (such as graphite purity and impurity removal rate) during the purification process helps identify key factors affecting the purification effect. Purification indexes fluctuate with purification parameters. Purification index fluctuation analysis refers to the statistical analysis of fluctuations in relevant indices (such as graphite purity and impurity removal rate) during the purification process to identify which factors or conditions lead to fluctuations in the purification effect. Influencing factors for purification indexes refer to dynamic variables related to fluctuation factors during the purification process, such as temperature, liquid concentration, and particle size. These influencing factors reflect the dynamic conditions affecting index fluctuations during the purification process.

[0039] Based on the liquid gradient concentration distribution fluctuation factor and the purification index follow-up factor, noise fluctuation scale identification is performed to quantify the amplitude and range of fluctuations, thereby determining the degree of impact of these fluctuations on the purification process. Noise fluctuation scale identification determines the amplitude and range of influence by analyzing the fluctuations in liquid concentration and purification indexes. By identifying the noise fluctuation scale, errors caused by unstable factors in the purification process can be predicted and reduced, thus improving the stability of the purification process. Appropriate noise intensity and fluctuation scale are determined by analyzing the error range in historical data. Different noise scales can simulate different types of disturbances in experiments, such as changes in liquid concentration, temperature fluctuations, and fluctuations in centrifugal force.

[0040] The liquid gradient concentration distribution and purification index data from multiple historical physical properties of polymerization—liquid gradient concentration distribution—purification index datasets are processed using a Bayesian normalization layer. This layer introduces randomness into the liquid gradient concentration distribution and purification index data to simulate real-world uncertainty. Based on the fluctuation characteristics of the historical data, particularly the noise fluctuation scale (e.g., the amplitude of liquid concentration fluctuations or the variation of purification indexes), Gaussian-distributed noise is generated. The noise generation process is based on prior analysis of data fluctuations. The mean and standard deviation of the Gaussian noise match the fluctuations in the historical data. Once the Gaussian noise is generated, it is used to adjust the liquid gradient concentration distribution and purification index data from multiple historical physical properties of polymerization—liquid gradient concentration distribution—purification index datasets. The adjusted data will contain a certain degree of randomness to simulate potential real-world uncertainties. The Bayesian normalization layer calculates the mean and variance of the input data, considers the data uncertainty, and performs normalization processing to ensure consistency between the adjusted data and the original data, avoiding excessive data perturbation. Normalization ensures that the liquid gradient concentration distribution data and purification index data vary within a suitable range, while maintaining the relative relationships and operability of the data. After normalization in the Bayesian normalization layer, the liquid gradient concentration distribution and purification index data are adjusted to an expanded initial dataset, simulating more purification scenarios and providing rich test data for subsequent optimization. All expanded datasets generated through random perturbation are integrated with the original historical dataset to form a complete training set containing diverse data, i.e., multiple initial expanded aggregated historical physical characteristics-liquid gradient concentration distribution-purification index datasets. The expanded dataset can simulate purification effects under different fluctuation and uncertainty conditions. For example, in some scenarios, the liquid concentration fluctuates significantly, and the purity may fluctuate greatly; in other scenarios, the fluctuation is smaller, and the purification effect is more stable. By expanding the dataset, the model can learn purification strategies under different conditions, thereby improving its adaptability. For example, suppose there are three graphite ore sample datasets from different sources, including the physical properties of the graphite ore samples (such as particle size, specific surface area, etc.), liquid concentration (fluctuating between 20% and 50%), and corresponding purification indicators (graphite purity and impurity removal rate). Historical datasets: Experiment 1, liquid concentration 30%, purity 80%, impurity removal rate 50%; Experiment 2, liquid concentration 40%, purity 82%, impurity removal rate 52%; Experiment 3, liquid concentration 50%, purity 85%, impurity removal rate 55%. Fluctuation analysis shows that the liquid concentration fluctuates by ±5%, and the purification indicator fluctuates by ±2% purity. Random fluctuations in liquid concentration and purification indicators are simulated by generating Gaussian-distributed noise.Bayesian normalization was used to adjust for Gaussian noise, ensuring the data varied within a reasonable range. For example, with a liquid concentration of 30%, the generated purity data might be adjusted from 80% to between 78% and 82%. The expanded dataset is as follows: Experiment 4: Liquid concentration 30%, purity 79%, impurity removal rate 51%; Experiment 5: Liquid concentration 40%, purity 83%, impurity removal rate 53%; Experiment 6: Liquid concentration 50%, purity 84%, impurity removal rate 56%.

[0041] Bayesian normalization is a normalization method based on Bayesian inference that processes data by considering its uncertainty. In data augmentation and noise analysis, Bayesian normalization layers can be reasonably adjusted and processed based on the noise characteristics of historical data, thereby improving the robustness and predictive ability of the model. By applying random perturbations to multiple datasets of historical physical properties of polymerization—liquid gradient concentration distribution, and purification indicators—using Bayesian normalization layers, a certain degree of randomness is introduced to augment the dataset, simulating uncertainties in the real world. Random perturbations help simulate noise and uncertainty in the data, enabling the model to more accurately handle complex real-world situations.

[0042] Multiple initial expanded polymerization historical physical properties, liquid gradient concentration distribution, and purification index datasets are subjected to approximation analysis to measure the similarity between different datasets. When the approximation of some datasets exceeds a preset threshold, it indicates that they are highly similar in some aspects. Approximation refers to the degree of similarity between data points, which can be measured by calculating the distance or similarity between data points. A preset approximation threshold, such as 0.8, is set. The initial expanded polymerization historical physical properties, liquid gradient concentration distribution, and purification index datasets with approximation exceeding the preset threshold are then averaged. By calculating the average of these similar data points, the influence of noise and outliers is reduced, resulting in a more stable and reliable dataset. Averaging involves averaging the values ​​of multiple datasets to reduce data volatility and extract more stable trends. This is used to adjust the expanded data with approximation exceeding the threshold, ensuring the stability of these datasets and reducing unnecessary fluctuations. For example, assuming that the approximation of the augmented data of graphite sample 1 and graphite sample 2 is greater than a preset threshold (such as 0.85), the two datasets are averaged: the average values ​​of Experiment 1 and Experiment 2 are: liquid concentration 32.5%, purity 79%, and impurity removal rate 51%.

[0043] After averaging, an expanded dataset of historical physical properties, liquid gradient concentration distribution, and purification indicators was obtained. This dataset includes historical physical properties, liquid gradient concentration distribution, and purification indicator data, reflecting the impact of variations in different process parameters (such as liquid concentration and temperature) on the purification effect. It also helps improve the accuracy and robustness of the purification process. By using random perturbation and averaging, the dataset was expanded to simulate the purification effect under different fluctuation conditions. Averaging reduced data fluctuations and enhanced the stability and reliability of the data.

[0044] By centrally analyzing multiple datasets of historical physical properties, liquid gradient concentration distribution, and purification indicators from extended polymerization processes, effective data was selected, and redundant or irrelevant data were removed. For example, if certain data appeared repeatedly in multiple experiments, they could be summarized into representative values ​​to avoid duplicate calculations. All data was ensured to be organized according to a unified standard format, and all data items (such as liquid concentration and graphite purity) were converted to a unified unit of measurement to avoid errors caused by inconsistent units. From the centrally analyzed data, key features affecting graphite purification were extracted, including the physical properties of the graphite sample (such as particle size, specific surface area, and density), the liquid gradient concentration distribution (such as solvent concentration changes), and indicators during the purification process (such as purity and impurity removal rate). Each dataset was compiled into an independent graphite purification scenario template, including different liquid concentrations, sample characteristics, and purification indicators, reflecting the graphite purification effects under different process conditions.

[0045] Through the above process, multiple graphite purification scenario templates were obtained, covering different purification environments, process parameters, and purification effects. All purification scenario templates were then aggregated to form a complete graphite purification scenario template library. This library is a database comprised of multiple purification scenario template sets, representing a complete set of templates under different purification conditions. It helps in selecting and optimizing purification strategies during actual production. By sorting and aggregating historical data, efficient data utilization was ensured, while Bayesian normalization reduced the impact of uncertainty on purification effects. By generating the expanded dataset and the graphite purification scenario template library, purification processes under different conditions can be simulated, aiding in the selection of the most suitable purification strategy.

[0046] The graphite purification scenario template library is traversed to identify graphite purification strategies, and a graphite purification strategy template library is constructed.

[0047] Furthermore, this application also includes the following steps: pre-constructing a graphite purification strategy recognizer, wherein the graphite purification strategy recognizer includes an input layer, a convolutional layer, and an output layer; using the graphite purification strategy recognizer to perform strategy recognition on graphite purification scene templates in the graphite purification scene template library to obtain the graphite purification strategy template library.

[0048] Specifically, a graphite purification strategy recognizer is a machine learning-based method designed to identify and select suitable purification strategies. It predicts or selects the optimal purification strategy based on input data (such as the physical properties of the graphite sample, liquid gradient concentration, and purification indices). The graphite purification strategy recognizer learns key features in the purification process through training data and outputs the optimal purification strategy through its output layer. A graphite purification strategy recognizer is constructed based on a convolutional neural network (CNN) architecture. The graphite purification strategy recognizer includes an input layer, convolutional layers, and an output layer. The input layer receives features from historical graphite purification data, including the physical properties of the graphite sample (such as particle size, specific surface area, and density), liquid gradient concentration distribution, and purification indices (such as graphite purity and impurity removal rate). The convolutional layers extract features from the input data. The convolutional layers scan the input data using filters to identify patterns and trends affecting the purification effect. For example, the convolutional layers identify the impact of liquid concentration fluctuations and changes in the physical properties of the ore sample on the purification effect. The output layer is used to output the final purification strategy based on the network's calculation results, including the setting of parameters such as liquid concentration, temperature control, and reaction time.

[0049] A large amount of graphite purification sample data was collected and organized, including the physical properties of graphite ore samples, liquid gradient concentration distribution, and purification indices. The graphite purification sample data was divided into an 80% training set and a 20% validation set. The graphite purification strategy recognizer was trained using the training set. During training, the graphite purification strategy recognizer attempted to match input data with corresponding purification strategies. The parameters of the graphite purification strategy recognizer were continuously adjusted through an optimization algorithm to minimize the loss function. During training, the performance of the graphite purification strategy recognizer was periodically evaluated using the validation set to ensure that it did not overfit and could generalize to new, unseen data. Through training, the graphite purification strategy recognizer learned the complex mapping relationship from graphite purification scenarios to purification strategies.

[0050] Using a pre-trained graphite purification strategy recognizer, strategy identification is performed on each template in the graphite purification scenario template library. By inputting data such as ore sample characteristics, liquid gradient concentration distribution, and purification indicators for each purification scenario template, the graphite purification strategy recognizer assigns a suitable purification strategy to each scenario template based on existing learning patterns. For example, assuming a template contains graphite ore samples with liquid concentrations ranging from 30% to 50%, the graphite purification strategy recognizer will analyze the data of this template (such as particle size, liquid concentration fluctuation range, etc.) and select the most suitable purification strategy, including increasing liquid concentration and adjusting temperature.

[0051] By identifying each graphite purification scenario template, a corresponding purification strategy template is generated, forming a new graphite purification strategy template library. This library summarizes optimized purification strategies selected based on different ore sample characteristics and process conditions, encompassing strategies for various purification processes to adapt to different graphite ore samples and production conditions. A pre-built graphite purification strategy recognizer automatically identifies and extracts effective purification strategies from the graphite purification scenario template library, improving the efficiency of purification strategy formulation and reducing the cost of manual analysis and experimentation.

[0052] The graphite ore sample is subjected to gradient density separation and purification. Based on the physical characteristics of the graphite ore sample and the monitoring results of the liquid gradient concentration distribution during the purification process, the graphite purification scenario template library is matched and identified to obtain a matching graphite purification scenario template.

[0053] Furthermore, this application also includes the following steps: using the physical characteristic information of the graphite ore sample and the monitoring results of the liquid gradient concentration distribution during the purification process as indexes, performing cosine similarity calculations on the graphite purification scene templates in the graphite purification scene template library, extracting the graphite purification scene template corresponding to the maximum calculated value, and obtaining a matching graphite purification scene template.

[0054] Specifically, gradient density separation is used to purify graphite ore samples. Gradient density separation is a common physical method for mineral purification that utilizes the density differences of different substances in a liquid to separate minerals through a liquid medium with gradually changing densities. In the graphite purification process, the graphite ore sample is added to a liquid with a gradient density. The graphite is separated by the density difference, and impurities are removed. When the graphite ore sample is placed in a liquid medium with a gradient density, the liquid concentration has a gradient between different levels, and the density gradually increases. This density difference causes different mineral particles to settle or float in the liquid according to their density, ultimately separating the graphite from the impurities. The graphite ore sample stratifies in the liquid according to density differences; graphite usually remains in a specific density region, while impurities are eliminated.

[0055] The physical characteristics of graphite ore samples are their fundamental physical properties, such as particle size, specific surface area, density, and hardness. During graphite purification, the distribution and variation of liquid concentration affect the purification effect. The liquid gradient concentration distribution monitoring results show the changes in liquid concentration under different purification conditions, used to analyze concentration fluctuations during the purification process and their impact on the ore sample purification effect. Using the physical characteristics of the graphite ore samples and the liquid gradient concentration distribution monitoring results during the purification process as indexes, these data are used as input features for comparison with each template in the graphite purification scenario template library.

[0056] By calculating cosine similarity, the physical characteristics of the graphite ore sample and the monitoring results of the liquid gradient concentration distribution during the purification process are compared with graphite purification scene templates in the graphite purification scene template library. The resulting similarity value represents the similarity between the current graphite ore sample and each graphite purification scene template in the library. A higher similarity value indicates a greater similarity between the current graphite ore sample and that template, making it more suitable for the purification strategy using that template. The graphite purification scene template corresponding to the highest calculated cosine similarity value is extracted as the best matching template for the current ore sample. For example, in one example, graphite ore sample A has a particle size of 20-50 μm, a liquid concentration of 30%-50%, a graphite purity of 80%, and an impurity removal rate of 50%. Templates in the graphite purification scenario template library: Template 1: Particle size 20-50μm, liquid concentration 30%-50%, graphite purity 80%, impurity removal rate 50%; Template 2: Particle size 10-40μm, liquid concentration 40%-60%, graphite purity 85%, impurity removal rate 55%; Template 3: Particle size 50-100μm, liquid concentration 20%-40%, graphite purity 75%, impurity removal rate 45%. Mineral sample A has a similarity of 1.0 with Template 1 (complete match); a similarity of 0.85 with Template 2 (similar but not a complete match); and a similarity of 0.6 with Template 3 (low similarity). Template 1 has the highest cosine similarity, therefore Template 1 is extracted.

[0057] By matching the sample with all graphite purification scenario templates in the graphite purification scenario template library, the most suitable purification strategy for the current ore sample can be selected, optimizing the purification effect and reducing experimental errors and fluctuations. Cosine similarity calculation helps identify the purification strategy template most similar to the current ore sample, thereby adjusting production conditions based on the process parameters (such as liquid concentration, reaction time, etc.) in the template, optimizing process parameters, and improving graphite purity and purification efficiency.

[0058] Based on the matching graphite purification scenario template, the graphite purification strategy template library is searched to obtain matching graphite purification strategy templates to optimize the gradient density separation and purification process.

[0059] Furthermore, this application also includes the following steps: continuously monitoring the gradient density separation and purification process to obtain a sequence of continuous monitoring results for the liquid gradient concentration distribution; performing a deviation analysis between the sequence of continuous monitoring results for the liquid gradient concentration distribution and the corresponding liquid gradient concentration distribution in the matching graphite purification scenario template to obtain a deviation sequence; analyzing the deviation sequence from two dimensions—change trend and concentrated data—to determine the deviation influence coefficient; and optimizing the matching graphite purification strategy template based on the deviation influence coefficient.

[0060] Furthermore, this application also includes the following steps: using the mean shift algorithm to perform centralized data identification on the deviation sequence to obtain centralized deviation; performing trend feature analysis on the deviation sequence to obtain deviation trend features; and correcting the centralized deviation based on the deviation trend features to obtain the deviation influence coefficient.

[0061] Specifically, based on the matching graphite purification scenario template, a search is performed in the graphite purification strategy template library to obtain the most suitable matching graphite purification strategy template, which includes the corresponding purification process parameters, such as liquid concentration, temperature, and reaction time. In other words, the most suitable purification strategy template for the current scenario is selected from the graphite purification strategy template library.

[0062] In the gradient density separation and purification process, sensors monitor the gradient concentration distribution in the liquid in real time, reflecting the concentration changes in different regions of the liquid. Continuously recording the gradient concentration distribution data forms a continuous monitoring result sequence, reflecting the changes in liquid concentration during the purification process. During gradient density separation and purification, the liquid concentration changes with time and depth. By continuously monitoring the gradient concentration distribution and recording the concentration change data sequence, a continuous monitoring result sequence of the liquid gradient concentration distribution is obtained.

[0063] By analyzing the discrepancy between the continuous monitoring sequence of liquid gradient concentration distribution and the corresponding liquid gradient concentration distribution in a matching graphite purification scenario template, the difference between the actual measurement result and the ideal result is analyzed. The discrepancy sequence reflects the difference between the actual liquid concentration and the corresponding liquid gradient concentration distribution in the graphite purification scenario template. In other words, for each moment of liquid gradient concentration distribution monitoring, it is compared with the corresponding liquid gradient concentration distribution, and the discrepancy is calculated. The discrepancy is typically the difference between two data points, expressed as a numerical value. For example, if the liquid gradient concentration in the graphite purification scenario template is 40%, while the actual monitored value is 38%, the discrepancy is 2%. If the monitored value is 42%, the discrepancy is also 2%.

[0064] By comparing the monitoring data and template data at all times, the deviation at each time point is calculated, and these deviations are arranged in chronological order to generate a deviation sequence. For example, the deviation at time point 1 is 0.5%, the deviation at time point 2 is 1.2%, the deviation at time point 3 is 0.8%, and the deviation at time point 4 is 1.0%.

[0065] By analyzing the deviation sequence, the stability of liquid concentration fluctuations during the purification process can be assessed. Trend analysis can reveal whether the deviation gradually increases, gradually decreases, or remains stable. If the deviation is small, it indicates that the liquid concentration varies within the ideal range, and the purification process is relatively stable; if the deviation is large, it indicates that the liquid concentration fluctuates significantly during the purification process, which may lead to unstable purification results and requires optimization.

[0066] The mean-shift algorithm is used to identify concentrated areas of deviation in the deviation sequence, representing stable phases of liquid concentration deviation during the purification process. Given the deviation sequence, for each data point, the mean-shift algorithm moves towards areas of higher density based on the density of surrounding data points until it reaches the maximum density point, thus identifying concentrated areas of deviation in the sequence. By analyzing the data density of the deviation sequence, the mean-shift algorithm identifies stable phases of deviation (i.e., areas with smaller fluctuations). Time series analysis is then performed on the deviation sequence to analyze its trend over time, identifying patterns of deviation change, whether the deviation gradually increases, gradually decreases, or remains stable, and extracting specific trend characteristics. For example, deviation may increase at certain times due to external factors (such as temperature changes), or remain stable due to good liquid concentration control.

[0067] The concentration deviation is corrected based on the changing trend characteristics of the deviation, thus obtaining the deviation influence coefficient. The concentration deviation is corrected according to the trend analysis results. For example, if the deviation gradually increases in a certain stage, the concentration deviation can be adjusted by increasing its weight to reflect the gradual impact of the deviation on the purification effect. The deviation influence coefficient is calculated based on the corrected concentration deviation. The deviation influence coefficient is usually determined by the magnitude of the corrected concentration deviation and the strength of the changing trend. A larger deviation influence coefficient indicates a stronger impact of the deviation on the purification effect; conversely, a smaller deviation influence coefficient indicates a weaker impact of the deviation on the purification effect. For example, suppose there is a graphite purification experiment with a liquid concentration range of 30%-50%, and the data on the change in liquid concentration during the purification process is acquired through monitoring equipment, generating a deviation sequence [0.5%, 1.2%, 0.8%, 1.0%, 0.7%], representing the deviation at different time points. The mean-shift algorithm was used to analyze the deviation sequence and identify concentrated regions. Regions with deviation fluctuating between 0.5% and 1.0% were considered concentrated regions, with a calculated concentrated deviation of 0.8%. Time series analysis revealed that the deviation gradually increased in the first half of the experiment (0.5% to 1.2%), while remaining relatively stable in the second half (1.0% to 0.7%). The trend of deviation change was initially increasing and then stabilizing. Based on the trend analysis, the concentrated deviation was corrected to 0.9% to reflect the impact of deviation stabilizing in the later stages of the experiment. Based on the corrected concentrated deviation and its trend, the deviation influence coefficient was calculated to be 0.6, indicating that the deviation has a moderate impact on the purification effect.

[0068] Based on the deviation influence coefficient, the matching graphite purification strategy template was optimized by adjusting the liquid concentration range, temperature control, and reaction time to obtain the optimized matching graphite purification strategy template, which set process parameters such as liquid concentration, reaction time, and temperature. Fluctuations in liquid concentration were monitored in real time, and the liquid concentration distribution during the purification process was adjusted based on the deviation analysis results to ensure the stability of the purification process. By analyzing the trend of deviation changes, key factors affecting the purification effect could be identified, and the matching graphite purification strategy template could be optimized by adjusting the process parameters within it. For example, assuming the liquid concentration of graphite ore sample C is 40%-50%, the purification result is 90% graphite purity and 60% impurity removal rate; continuous monitoring of the gradient density separation purification process of graphite ore sample C yielded monitoring data of [40%, 45%, 48%, 50%, 49%]; the matched purification scenario template had a liquid concentration range of 40%-50%. The ideal concentration range for matching the template is 40%-50%, and the concentration fluctuations in the monitoring data are [40%, 45%, 48%, 50%, 49%], resulting in a deviation sequence of [0%, 0%, 2%, 0%, 1%]. The low deviation indicates relatively precise liquid concentration control. The small deviation fluctuation, with a deviation influence coefficient of 0.1, indicates that the purification process is very stable in terms of liquid concentration control. By matching the most suitable graphite purification scenario template and performing deviation analysis, the most suitable purification strategy for the current ore sample can be selected, ensuring a stable and efficient purification process.

[0069] In summary, the graphite purification method based on gradient density separation provided in this application has the following beneficial effects: Graphite samples are screened and ultrasonically cleaned to remove surface impurities. The cleaned graphite samples are then subjected to physical property measurements to obtain their physical characteristics. A Bayesian normalization layer is introduced, and historical graphite purification data is used to simulate the noise quantification of liquid gradient concentration distribution, constructing a graphite purification scenario template library. This library is then traversed to identify graphite purification strategies, constructing a graphite purification strategy template library. The graphite samples are then purified using gradient density separation. Based on the physical characteristics of the graphite samples and the monitoring results of the liquid gradient concentration distribution during purification, the graphite purification scenario template library is matched to obtain matching graphite purification scenario templates. Finally, the matching graphite purification scenario templates are used to search the graphite purification strategy template library to obtain matching graphite purification strategy templates to optimize the gradient density separation purification process. In other words, by combining the Bayesian method with gradient density separation technology, the graphite purification process can maintain high stability and reliability in experimental environments with significant uncertainties.

[0070] Example 2: Based on the same inventive concept as the graphite purification method based on gradient density separation in Example 1, this application also provides a graphite purification system based on gradient density separation. Please refer to the appendix. Figure 2 The graphite purification system based on gradient density separation includes: The system comprises the following modules: a physical property measurement module 11, which screens graphite samples and performs ultrasonic cleaning to remove surface impurities, then measures the physical properties of the cleaned graphite samples to obtain their physical characteristics; a noise quantization simulation module 12, which introduces a Bayesian normalization layer and combines historical graphite purification data to perform noise quantization simulation of liquid gradient concentration distribution, thereby constructing a graphite purification scenario template library; a purification strategy identification module 13, which traverses the graphite purification scenario template library to identify graphite purification strategies, thereby constructing a graphite purification strategy template library; a matching identification module 14, which performs gradient density separation purification on the graphite samples and, based on the physical characteristics of the graphite samples and the monitoring results of liquid gradient concentration distribution during the purification process, matches and identifies the graphite purification scenario template library to obtain matching graphite purification scenario templates; and a strategy retrieval module 15, which retrieves the graphite purification strategy template library based on the matching graphite purification scenario templates to obtain matching graphite purification strategy templates for optimizing the gradient density separation purification process.

[0071] Furthermore, the noise quantization simulation module 12 in the graphite purification system based on gradient density separation is also used to: sort the historical graphite purification data according to physical characteristics, liquid gradient concentration distribution, and purification indicators respectively, to obtain a historical physical characteristics-liquid gradient concentration distribution-purification indicator data set; aggregate the historical physical characteristics-liquid gradient concentration distribution-purification indicator data set according to physical characteristics to obtain multiple aggregated historical physical characteristics-liquid gradient concentration distribution-purification indicator data sets; identify noise fluctuation scales for each of the multiple aggregated historical physical characteristics-liquid gradient concentration distribution-purification indicator data sets to obtain multiple noise fluctuation scales; introduce a Bayesian normalization layer to randomly perturb and expand the multiple aggregated historical physical characteristics-liquid gradient concentration distribution-purification indicator data sets according to the multiple noise fluctuation scales to obtain multiple expanded aggregated historical physical characteristics-liquid gradient concentration distribution-purification indicator data sets; and determine multiple graphite purification scenario template sets by centrally sorting the multiple expanded aggregated historical physical characteristics-liquid gradient concentration distribution-purification indicator data sets, and summarize and construct the graphite purification scenario template library.

[0072] Furthermore, the noise quantization simulation module 12 in the graphite purification system based on gradient density separation is also used to: traverse the multiple sets of polymerization historical physical characteristics-liquid gradient concentration distribution-purification index data to perform liquid gradient concentration distribution fluctuation analysis and determine multiple liquid gradient concentration distribution fluctuation factors; traverse the multiple sets of polymerization historical physical characteristics-liquid gradient concentration distribution-purification index data to perform purification index fluctuation analysis and determine multiple purification index follow-up factors; and identify noise fluctuation scales based on the multiple liquid gradient concentration distribution fluctuation factors and the multiple purification index follow-up factors to obtain multiple noise fluctuation scales.

[0073] Furthermore, the noise quantization simulation module 12 in the graphite purification system based on gradient density separation is also used to: randomly perturb the liquid gradient concentration distribution and purification index data in the multiple sets of polymerization history physical characteristics-liquid gradient concentration distribution-purification index data according to the multiple noise fluctuation scales through the Bayesian normalization layer to obtain multiple initial expanded sets of polymerization history physical characteristics-liquid gradient concentration distribution-purification index data; and perform mean processing on the initial expanded sets of polymerization history physical characteristics-liquid gradient concentration distribution-purification index data whose approximation is greater than a preset approximation threshold to obtain the multiple expanded sets of polymerization history physical characteristics-liquid gradient concentration distribution-purification index data.

[0074] Furthermore, the noise quantization simulation module 12 in the graphite purification system based on gradient density separation is also used to: generate Gaussian noise based on the multiple noise fluctuation scales through the Bayesian normalization layer, and adjust the liquid gradient concentration distribution and purification index data in the multiple sets of polymerization history physical characteristics-liquid gradient concentration distribution-purification index data respectively according to the generation results, so as to obtain the multiple sets of initial expanded polymerization history physical characteristics-liquid gradient concentration distribution-purification index data.

[0075] Furthermore, the purification strategy recognition module 13 in the graphite purification system based on gradient density separation is also used for: pre-constructing a graphite purification strategy recognizer, wherein the graphite purification strategy recognizer includes an input layer, a convolutional layer, and an output layer; and using the graphite purification strategy recognizer to perform strategy recognition on graphite purification scene templates in the graphite purification scene template library to obtain the graphite purification strategy template library.

[0076] Furthermore, the matching and identification module 14 in the graphite purification system based on gradient density separation is also used to: use the physical characteristic information of the graphite ore sample and the monitoring results of the liquid gradient concentration distribution during the purification process as indexes to perform cosine similarity calculation on the graphite purification scene templates in the graphite purification scene template library, extract the graphite purification scene template corresponding to the maximum value, and obtain the matching graphite purification scene template.

[0077] Furthermore, the strategy retrieval module 15 in the graphite purification system based on gradient density separation is also used for: continuously monitoring the gradient density separation purification process to obtain a sequence of continuous monitoring results of liquid gradient concentration distribution; performing a deviation analysis between the sequence of continuous monitoring results of liquid gradient concentration distribution and the corresponding liquid gradient concentration distribution in the matching graphite purification scenario template to obtain a deviation sequence; analyzing the deviation sequence from two dimensions, namely, the trend of change and the concentrated data, to determine the deviation influence coefficient; and optimizing the matching graphite purification strategy template based on the deviation influence coefficient.

[0078] Furthermore, the strategy retrieval module 15 in the graphite purification system based on gradient density separation is also used to: use the mean drift algorithm to perform centralized data identification on the deviation sequence to obtain centralized deviation; perform trend feature analysis on the deviation sequence to obtain deviation trend features; and correct the centralized deviation based on the deviation trend features to obtain the deviation influence coefficient.

[0079] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Figure 1 The graphite purification method and specific example based on gradient density separation in Example 1 are also applicable to the graphite purification system based on gradient density separation in this embodiment. Through the foregoing detailed description of the graphite purification method based on gradient density separation, those skilled in the art can clearly understand the graphite purification system based on gradient density separation in this embodiment. Therefore, for the sake of brevity, it will not be described in detail here.

[0080] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0081] Obviously, those skilled in the art can make several improvements and modifications to this application without departing from the principles of this application, and these improvements and modifications also fall within the protection scope of this application.

Claims

1. A graphite purification method based on gradient density separation, characterized in that, include: Graphite ore samples were screened and ultrasonically cleaned to remove surface impurities. The physical properties of the cleaned graphite ore samples were then measured to obtain physical characteristic information of the graphite ore samples. A Bayesian normalization layer is introduced, and historical graphite purification data is combined to simulate the noise quantization of liquid gradient concentration distribution, thereby constructing a graphite purification scenario template library. The graphite purification scenario template library is traversed to identify graphite purification strategies, and a graphite purification strategy template library is constructed. The graphite ore sample is subjected to gradient density separation and purification. Based on the physical characteristics of the graphite ore sample and the monitoring results of the liquid gradient concentration distribution during the purification process, the graphite purification scenario template library is matched and identified to obtain a matching graphite purification scenario template. Based on the matching graphite purification scenario template, the graphite purification strategy template library is searched to obtain matching graphite purification strategy templates to optimize the gradient density separation and purification process.

2. The graphite purification method based on gradient density separation as described in claim 1, characterized in that, A Bayesian normalization layer is introduced, and historical graphite purification data is used to simulate the noise quantization of liquid gradient concentration distribution. A graphite purification scenario template library is constructed, including: The historical graphite purification data were sorted out according to physical properties, liquid gradient concentration distribution and purification index to obtain a historical physical properties-liquid gradient concentration distribution-purification index data set. Based on physical characteristics, the historical physical characteristics-liquid gradient concentration distribution-purification index data set is aggregated in the same category to obtain multiple aggregated historical physical characteristics-liquid gradient concentration distribution-purification index data sets. Based on multiple datasets of historical physical properties of polymerization, liquid gradient concentration distribution, and purification indicators, noise fluctuation scales are identified to obtain multiple noise fluctuation scales. A Bayesian normalization layer is introduced to randomly perturb and expand the multiple sets of polymerization history physical properties-liquid gradient concentration distribution-purification index data according to the multiple noise fluctuation scales, thereby obtaining multiple expanded sets of polymerization history physical properties-liquid gradient concentration distribution-purification index data. By centrally sorting out the data of the multiple extended polymerization historical physical properties, liquid gradient concentration distribution and purification index data sets, multiple graphite purification scenario template sets were determined, and the graphite purification scenario template library was constructed.

3. The graphite purification method based on gradient density separation as described in claim 2, characterized in that, Noise fluctuation scales were identified based on multiple datasets of historical polymerization physical properties, liquid gradient concentration distribution, and purification indicators, resulting in multiple noise fluctuation scales, including: By traversing the multiple sets of historical physical properties of polymerization, liquid gradient concentration distribution, and purification index data, liquid gradient concentration distribution fluctuation analysis is performed to determine multiple liquid gradient concentration distribution fluctuation factors. By traversing multiple sets of historical physical properties, liquid gradient concentration distribution, and purification index data, a fluctuation analysis of purification indexes is performed to determine multiple purification index follow-up factors. Based on the multiple liquid gradient concentration distribution fluctuation factors and the multiple purification index follow-up factors, noise fluctuation scales are identified to obtain multiple noise fluctuation scales.

4. The graphite purification method based on gradient density separation as described in claim 2, characterized in that, A Bayesian normalization layer is introduced to randomly perturb and expand the multiple sets of polymerization history physical properties-liquid gradient concentration distribution-purification index data according to the multiple noise fluctuation scales, resulting in multiple expanded sets of polymerization history physical properties-liquid gradient concentration distribution-purification index data, including: The Bayesian normalization layer randomly perturbs the liquid gradient concentration distribution and purification index data in the multiple sets of polymerization history physical properties-liquid gradient concentration distribution-purification index data according to the multiple noise fluctuation scales to obtain multiple initial expanded sets of polymerization history physical properties-liquid gradient concentration distribution-purification index data. The average value of the initial expansion polymerization historical physical properties-liquid gradient concentration distribution-purification index data with an approximation greater than a preset approximation threshold in the multiple initial expansion polymerization historical physical properties-liquid gradient concentration distribution-purification index data sets is processed to obtain the multiple expansion polymerization historical physical properties-liquid gradient concentration distribution-purification index data sets.

5. The graphite purification method based on gradient density separation as described in claim 4, characterized in that, Gaussian noise is generated based on the multiple noise fluctuation scales through the Bayesian normalization layer, and the liquid gradient concentration distribution and purification index data in the multiple polymerization history physical properties-liquid gradient concentration distribution-purification index data sets are adjusted according to the generation results to obtain the multiple initial expanded polymerization history physical properties-liquid gradient concentration distribution-purification index data sets.

6. The graphite purification method based on gradient density separation as described in claim 1, characterized in that, The graphite purification scenario template library is traversed to identify graphite purification strategies, and a graphite purification strategy template library is constructed, including: A pre-built graphite purification strategy recognizer is provided, wherein the graphite purification strategy recognizer includes an input layer, a convolutional layer, and an output layer; The graphite purification strategy recognizer is used to perform strategy recognition on the graphite purification scenario templates in the graphite purification scenario template library to obtain the graphite purification strategy template library.

7. The graphite purification method based on gradient density separation as described in claim 1, characterized in that, Using the physical characteristics of the graphite ore sample and the monitoring results of the liquid gradient concentration distribution during the purification process as indexes, cosine similarity calculations are performed on the graphite purification scene templates in the graphite purification scene template library. The graphite purification scene template corresponding to the maximum calculated value is extracted to obtain the matching graphite purification scene template.

8. The graphite purification method based on gradient density separation as described in claim 1, characterized in that, include: The gradient density separation and purification process is continuously monitored to obtain a sequence of continuous monitoring results of liquid gradient concentration distribution. The deviation sequence is obtained by performing a deviation analysis between the continuous monitoring result sequence of the liquid gradient concentration distribution and the corresponding liquid gradient concentration distribution in the matching graphite purification scenario template. The deviation sequence is analyzed from two dimensions: the trend of change and the central tendency of data, to determine the deviation influence coefficient; The matching graphite purification strategy template is optimized based on the deviation influence coefficient.

9. The graphite purification method based on gradient density separation as described in claim 8, characterized in that, The deviation sequence is analyzed from two dimensions: trend of change and central tendency, to determine the deviation influence coefficient, including: The mean shift algorithm is used to perform clustered data identification on the deviation sequence to obtain the clustered deviation. The deviation sequence is analyzed to obtain the deviation change trend characteristics; The concentration deviation is corrected based on the deviation change trend characteristics to obtain the deviation influence coefficient.

10. A graphite purification system based on gradient density separation, characterized in that, The step of implementing the graphite purification method based on gradient density separation according to any one of claims 1 to 9, wherein the graphite purification system based on gradient density separation comprises: The physical property measurement module is used to screen graphite ore samples and perform ultrasonic cleaning to remove surface impurities. The cleaned graphite ore samples are then subjected to physical property measurements to obtain physical characteristic information of the graphite ore samples. The noise quantization simulation module is used to introduce a Bayesian normalization layer, combine historical graphite purification data to perform noise quantization simulation of liquid gradient concentration distribution, and build a graphite purification scenario template library. The purification strategy identification module is used to traverse the graphite purification scenario template library to identify graphite purification strategies and construct the graphite purification strategy template library. The matching and identification module is used to perform gradient density separation and purification on the graphite ore sample, and combine the physical characteristic information of the graphite ore sample and the monitoring results of liquid gradient concentration distribution during the purification process to match and identify the graphite purification scene template library to obtain a matching graphite purification scene template. The strategy retrieval module is used to search the graphite purification strategy template library based on the matching graphite purification scenario template, and obtain matching graphite purification strategy templates to optimize the gradient density separation and purification process.

Images