Machine learning-based lithium carbonate preparation device parameter optimization method and system

Abstract

This application discloses a method and system for optimizing equipment parameters in lithium carbonate production based on machine learning. The method includes: acquiring historical production data of lithium carbonate under different production conditions, wherein the historical production data includes equipment production parameters and corresponding production result parameters; constructing a training dataset and a validation dataset based on the historical production data; training a preset machine learning model using the training dataset to obtain the trained machine learning model; and validating and optimizing the trained machine learning model using the validation dataset to obtain a target machine learning model for optimizing lithium carbonate production parameters. This application can utilize machine learning technology to optimize equipment parameters in the lithium carbonate production process, improve the yield and quality of lithium carbonate, reduce energy consumption and costs in the lithium carbonate production process, and achieve intelligent optimization of the lithium carbonate production process.

Description

Technical Field

[0001] This application relates to the field of lithium carbonate modeling technology, specifically to a method and system for optimizing the parameters of lithium carbonate preparation equipment based on machine learning. Background Technology

[0002] With the rapid development of emerging fields such as electric vehicles and wearable devices, the demand for lithium batteries is constantly increasing, thereby driving up the market demand for lithium carbonate. Meanwhile, the purity of lithium carbonate has a significant impact on its performance, and high-purity lithium carbonate has a greater market demand in certain high-end battery fields. In the lithium carbonate preparation process, there are several equipment parameters that need optimization, such as temperature, pressure, reaction time, and raw material ratios. These parameters have a significant impact on the purity, yield, and energy consumption of lithium carbonate. Therefore, it is necessary to select appropriate process parameters to improve the preparation efficiency and product quality of lithium carbonate.

[0003] Traditional lithium carbonate preparation methods rely heavily on experience to adjust equipment parameters, making parameter decisions for lithium carbonate preparation heavily dependent on expert experience. This results in poor scientific rigor in decision-making and problems such as low yield, unstable quality, and high energy consumption. Summary of the Invention

[0004] In view of this, this application provides a method and system for optimizing the parameters of lithium carbonate preparation equipment based on machine learning. This method and system can optimize the equipment parameters in the lithium carbonate preparation process by using machine learning technology, thereby improving the yield and quality of lithium carbonate, reducing energy consumption and costs in the lithium carbonate production process, and achieving intelligent optimization of the lithium carbonate preparation process.

[0005] This application provides a method for optimizing the parameters of lithium carbonate preparation equipment based on machine learning, the method comprising:

[0006] Acquire historical production data of lithium carbonate under different production conditions, including equipment production parameters and corresponding production result parameters;

[0007] Based on the aforementioned historical production data, training and validation datasets are constructed.

[0008] The preset machine learning model is trained using the training dataset to obtain the trained machine learning model. The trained machine learning model is then validated and optimized using the validation set to obtain a target machine learning model for optimizing lithium carbonate preparation parameters.

[0009] Optionally, production data to be optimized is obtained and input into the target machine learning model to obtain production data optimization results.

[0010] Optionally, the target machine learning model includes a support vector machine, a random forest, or a neural network.

[0011] Optionally, the target machine learning model is a support vector machine, and the method for constructing the target machine learning model includes:

[0012] A genetic algorithm is used to optimize the kernel function parameters and penalty factor to obtain the optimal penalty factor and optimal kernel function parameters.

[0013] Optionally, the step of obtaining the production data to be optimized and inputting the production data to be optimized into the target machine learning model to obtain the production data optimization result includes:

[0014] Based on the historical production data, a physical model for lithium carbonate preparation was constructed.

[0015] Under the production data to be optimized, a physical model for lithium carbonate preparation was run to obtain a real-time dataset;

[0016] Input the real-time dataset into the target machine learning model, and output the production data optimization results in real time.

[0017] Optionally, the step of acquiring the production data to be optimized and inputting the production data to be optimized into the target machine learning model to obtain the production data optimization result further includes:

[0018] Based on the historical production data, a physical model for lithium carbonate preparation was constructed.

[0019] Under the production data to be optimized, the lithium carbonate preparation physical model and the target machine learning model are run to obtain the first real-time dataset and the second real-time dataset, respectively.

[0020] The first real-time dataset and the second real-time dataset are merged to obtain the production data optimization results.

[0021] Optionally, the physical model for lithium carbonate preparation includes a reaction kinetics model, a thermodynamic model, and a material balance model, wherein the method for constructing the material balance model includes:

[0022] Based on the chemical reaction equations in the lithium carbonate preparation process, determine the reactants and products in the lithium carbonate preparation process;

[0023] Based on the reactants and products of lithium carbonate under different production conditions, a material balance equation is constructed.

[0024] The material balance equations are converted into mathematical models for material balance analysis in the lithium carbonate preparation process.

[0025] Optionally, the physical model for lithium carbonate preparation includes a reaction kinetic model, a thermodynamic model, and a material balance model, and the method for constructing the reaction kinetic model includes:

[0026] Obtain the chemical reaction equations in the lithium carbonate preparation process, and determine the reaction rate equations by combining the principles of chemical reaction kinetics.

[0027] Based on historical production data of lithium carbonate under different production conditions, the rate constants of the reaction rate equation under different conditions were calculated.

[0028] The reaction kinetic model is constructed by combining the reaction rate equation and the material balance equation.

[0029] Optionally, the physical model for lithium carbonate preparation includes a reaction kinetics model, a thermodynamic model, and a material balance model, wherein the method for constructing the thermodynamic model includes:

[0030] Thermodynamic parameters of lithium carbonate under different production conditions are obtained, including enthalpy change, entropy change and Gibbs free energy change;

[0031] Using the chemical equilibrium constant expression and the aforementioned thermodynamic parameters, the equilibrium constant of lithium carbonate under different production conditions was calculated to construct a thermodynamic model.

[0032] Correspondingly, this application provides a machine learning-based parameter optimization system for lithium carbonate preparation equipment, comprising:

[0033] The data acquisition module is used to acquire historical production data of lithium carbonate under different production conditions. The historical production data includes equipment production parameters and corresponding production result parameters.

[0034] The data construction module is used to construct training datasets and validation datasets based on the historical production data.

[0035] The model training module is used to train a preset machine learning model using the training dataset to obtain the trained machine learning model, and to validate and optimize the trained machine learning model using the validation set to obtain a target machine learning model for optimizing lithium carbonate preparation parameters.

[0036] This application provides a method and system for optimizing the parameters of lithium carbonate preparation equipment based on machine learning. The method involves acquiring historical production data of lithium carbonate under different production conditions, including equipment production parameters and corresponding production result parameters. Based on the historical production data, a training dataset and a validation dataset are constructed. A pre-defined machine learning model is trained using the training dataset to obtain the trained machine learning model. The trained machine learning model is then validated and optimized using the validation dataset to obtain a target machine learning model for optimizing lithium carbonate preparation parameters. By utilizing machine learning technology to optimize the equipment parameters in the lithium carbonate preparation process, the yield and quality of lithium carbonate are improved, while energy consumption and costs in the lithium carbonate production process are reduced, achieving intelligent optimization of the lithium carbonate preparation process. Attached Figure Description

[0037] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0038] Figure 1 A flowchart illustrating the method for optimizing lithium carbonate preparation equipment parameters based on machine learning, provided in an embodiment of this application;

[0039] Figure 2 A schematic diagram of the structure of the lithium carbonate preparation equipment parameter optimization system based on machine learning provided in the embodiments of this application. Detailed Implementation

[0040] The embodiments of this disclosure will now be described in detail with reference to the accompanying drawings.

[0041] It should be understood that the following specific examples illustrate the implementation of this disclosure, and those skilled in the art can easily understand other advantages and effects of this disclosure from the content disclosed in this specification. Obviously, the described embodiments are only a part of the embodiments of this disclosure, and not all of them. This disclosure can also be implemented or applied through other different specific implementation methods, and the details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of this disclosure. It should be noted that, in the absence of conflict, the following embodiments and features in the embodiments can be combined with each other. Based on the embodiments in this disclosure, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this disclosure.

[0042] It should be noted that various aspects of embodiments within the scope of the appended claims are described below. It will be apparent that the aspects described herein can be embodied in a wide variety of forms, and any particular structure and / or function described herein is merely illustrative. Based on this disclosure, those skilled in the art will understand that one aspect described herein can be implemented independently of any other aspect, and two or more of these aspects can be combined in various ways. For example, any number of aspects set forth herein can be used to implement the device and / or practice the method. Additionally, this device and / or method can be implemented using structures and / or functionalities other than one or more of the aspects set forth herein.

[0043] It should also be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of this disclosure. The drawings only show the components related to this disclosure and are not drawn according to the number, shape and size of the components in actual implementation. In actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0044] Furthermore, specific details are provided in the following description to facilitate a thorough understanding of the examples. However, those skilled in the art will understand that the described aspects can be practiced without these specific details.

[0045] Example 1

[0046] Figure 1 This is a flowchart illustrating a method for optimizing parameters of a lithium carbonate preparation equipment based on machine learning, which includes steps S1 to S3.

[0047] S1. Obtain historical production data of lithium carbonate under different production conditions, wherein the historical production data includes equipment production parameters and corresponding production result parameters;

[0048] In this disclosure, the purpose of this step is to collect historical production data of lithium carbonate under different production conditions. Before using the historical production data for model training, it is necessary to select features from the historical production data that have a significant impact on the yield and quality of lithium carbonate. Among these, equipment production parameters are used to characterize the equipment parameters in the lithium carbonate preparation process. These equipment production parameters include pressure, temperature, material flow rate, reaction time, and reactor design. The corresponding production result parameters include the yield, quality, and energy consumption of lithium carbonate produced. The above data will serve as the basis for training the machine learning model. Preferably, the historical production data is cleaned to remove outliers and noise, and the data is standardized to facilitate analysis.

[0049] S2. Based on the historical production data, construct a training dataset and a validation dataset.

[0050] In this disclosure, after acquiring historical production data, the next step is to divide the historical production data into two parts: a training dataset and a validation dataset. The training dataset is used to train the machine learning model, while the validation dataset is used to evaluate and optimize the model's performance. This step is crucial in machine learning, ensuring the model's effectiveness and generalization ability.

[0051] S3. The preset machine learning model is trained using the training dataset to obtain the trained machine learning model. The trained machine learning model is then validated and optimized using the validation set to obtain a target machine learning model for optimizing lithium carbonate preparation parameters.

[0052] In this disclosure, a preset machine learning model is selected based on the characteristics of historical production data under different production conditions, and a training dataset is used to train it. The purpose of the training is to enable the machine learning model to learn the mapping relationship from equipment production parameters to production result parameters. In this way, by using machine learning technology, the equipment parameters in the lithium carbonate preparation process are optimized, the yield and quality of lithium carbonate are improved, the energy consumption and cost in the lithium carbonate production process are reduced, and intelligent optimization of the lithium carbonate preparation process is achieved.

[0053] Specifically, target machine learning models include Support Vector Machines (SVMs), Random Forests, or Neural Networks. SVMs are suitable for small datasets, especially linearly separable problems or problems that can be linearly classified by mapping to a high-dimensional space using kernel functions. Random Forests are suitable for high-dimensional data, requiring no dimensionality reduction or feature selection, and are robust to imbalanced datasets and missing / outlier features. Neural Networks are suitable for large-scale datasets, especially those requiring the capture of complex nonlinear relationships. They have wide applications in image processing, speech recognition, and natural language processing.

[0054] In this embodiment, the machine learning model used is a Support Vector Machine (SVM), a supervised learning model primarily used for classification and regression analysis. It distinguishes different categories of data points by finding a hyperplane, with the core idea being to maximize the boundary between two categories. SVM can effectively handle complex data patterns in high-dimensional spaces. In the lithium carbonate production process, SVM can be used to classify and predict multiple equipment production parameters that affect the production results.

[0055] Optionally, the target machine learning model may use a support vector machine (SVM). For example, an SVM may be used to predict optimal stress parameters to optimize product quality. The method for constructing the target machine learning model includes:

[0056] S301. The choice between using Support Vector Machines (SVC) for classification (SVC) or Regression (SVR) depends on whether the target variable (i.e., the prediction target) is categorical or numerical. When the target variable is categorical, the classification variant of Support Vector Machines, namely Support Vector Classification (SVC), is typically used. SVC is designed to handle classification problems and attempts to find a hyperplane to separate different categories. In this context, the target variable usually represents discrete category labels, such as "yes / no," "high / medium / low," etc. Conversely, when the target variable is numerical, the regression variant of Support Vector Machines, namely Support Vector Regression (SVR), should be used. SVR is designed to handle regression problems and attempts to predict a continuous numerical output. In this context, the target variable is usually a real number, such as a price, score, or any other continuous measure.

[0057] S302. Select a suitable kernel function. The Support Vector Machine (SVM) maps the data to a high-dimensional space through the kernel function in order to find an optimal separating hyperplane in this space. Kernel functions include linear kernels, polynomial kernels, and radial basis function (RBF) kernels. In this embodiment, a suitable kernel function is selected based on the characteristics of the lithium carbonate preparation data. For example, if the data is non-linearly separable, a radial basis function (RBF) kernel can be considered. The parameters of the SVM are then initialized, such as the regularization parameter C and the kernel function parameters (e.g., the degree of the polynomial kernel, the γ of the RBF kernel).

[0058] S303. Initially train the Support Vector Machine using the training dataset. This step typically involves running the algorithm on the training data to find the optimal decision boundary. Adjusting model parameters, such as changing the value of C to control the penalty for misclassification, and adjusting the kernel function parameters to change how the data is mapped to a higher-dimensional space, directly affects the model's generalization ability and its fit to the training data. Specifically, a genetic algorithm is used to optimize the kernel function parameters and penalty factor to obtain the optimal penalty factor and optimal kernel function parameters.

[0059] S304. The support vector machine is trained iteratively multiple times. In each iteration, the parameters are adjusted based on the model's performance (such as classification accuracy and regression error) until the optimal model parameters are found. The model will be used to optimize the parameters of the lithium carbonate production equipment to improve yield and quality and reduce energy consumption.

[0060] S305. Evaluate model performance using validation datasets or cross-validation. Assess performance metrics such as accuracy, recall, and F1 score to ensure the model maintains stable predictive ability across different data subsets. Monitor changes in the crystal structure and purity of adjusted lithium carbonate to evaluate the adjustment effect, ensuring the model performs well not only on the training dataset but also maintains stable performance on unseen data. Cross-validation is a statistical method for evaluating model performance. It reduces the risk of overfitting and underfitting by training and testing the model multiple times, providing a more accurate estimate of the model's generalization ability. For example, in k-fold cross-validation, the original dataset is divided into k equal-sized subsets (or "folds"). The model is then trained k times, each time using k-1 subsets as the training set and the remaining subset as the test set.

[0061] In some embodiments, a neural network is used to optimize temperature control. Temperature is a key factor affecting the yield and quality of lithium carbonate during its production. By analyzing historical data using a neural network model, the optimal temperature parameters can be predicted, enabling real-time adjustments. Furthermore, a random forest model can be used to optimize material ratio parameters. Different raw material ratios directly affect the purity and yield of lithium carbonate. The random forest model analyzes the impact of different ratios on yield and purity to find the optimal material ratio parameters.

[0062] In this embodiment, the production data to be optimized is obtained and input into the target machine learning model to obtain the production data optimization result.

[0063] In some embodiments, production data to be optimized is obtained, and the production data to be optimized is input into a target machine learning model to obtain the production data optimization result. The method for obtaining the production data optimization result includes:

[0064] Based on historical production data, a physical model for lithium carbonate production is constructed. Under the production data to be optimized, the physical model is run to obtain a real-time dataset. This real-time dataset is then input into the target machine learning model, which outputs the production data optimization results in real time. Using the output of the physical model as input to a Support Vector Machine (SVM) model is an effective method to improve prediction accuracy. This leads to the construction of a highly efficient digital twin system capable of accurately simulating and predicting the behavior of complex systems, thereby improving the accuracy and efficiency of decision-making. This system can then be applied to equipment parameter optimization.

[0065] Specifically, the steps for constructing a physical model for lithium carbonate preparation include: analyzing and determining the physical characteristics and operating mechanisms of the target system, such as mechanical and thermodynamic properties; establishing a mathematical model based on physical laws and relevant data to describe the physical behavior of the target system; and using computational software (such as MATLAB or ANSYS) to perform model simulation to verify the accuracy of the model.

[0066] For example, firstly, in the model fusion stage, the output of the lithium carbonate preparation physical model (such as the theoretical yield and energy consumption of lithium carbonate preparation) is used as one of the input features of the support vector machine. For characteristics that cannot be directly obtained from the physical model (such as the long-term durability of the equipment and maintenance requirements), the support vector machine is used for prediction and analysis. Furthermore, the time series characteristics of the data are considered during model fusion to ensure that the outputs of the lithium carbonate preparation physical model and the support vector machine model are aligned in time. Secondly, in the model optimization stage, various strategies are used to weight and optimize the output of the fused model. For example, the weights are adjusted based on the performance evaluation of historical data, and optimization algorithms (such as genetic algorithms and simulated annealing) are applied to find the best combination of model parameters to achieve the best prediction effect. Sensitivity analysis of the model is performed to understand the degree of influence of different input parameters on the final prediction result, so as to adjust the model more effectively.

[0067] In another embodiment, the process includes acquiring production data to be optimized and inputting it into a target machine learning model to obtain the production data optimization result. Specifically, this also includes:

[0068] Based on historical production data, a physical model for lithium carbonate production is constructed. Under the production data to be optimized, the physical model and the target machine learning model are run to obtain a first real-time dataset and a second real-time dataset, respectively. The first and second real-time datasets are then fused to obtain the production data optimization result. This involves combining the outputs of the physical model and the support vector machine to generate the final prediction result. This process can be achieved through simple averaging, weighted averaging, voting, or other more complex ensemble methods.

[0069] Optionally, the physical model for lithium carbonate preparation includes a reaction kinetics model, a thermodynamic model, and a material balance model. The methods for constructing the material balance model include:

[0070] Based on the chemical reaction equations in the lithium carbonate preparation process, determine the reactants and products in the lithium carbonate preparation process. Specifically, identify and list all reactants and products involved in the lithium carbonate preparation process. For example, if lithium carbonate is prepared by the reaction of carbonate and lithium source, then the two raw materials mentioned above, as well as the generated lithium carbonate and by-products, need to be considered, and the effects of pressure parameters on the solubility, volatility, and reaction rate of the reactants should be taken into account.

[0071] Based on the reactant and product conditions of lithium carbonate under different production conditions, material balance equations are constructed. Specifically, material balance equations are written for each step in lithium carbonate preparation, taking into account the consumption of reactants and the formation of products, including material losses during the reaction process, such as possible volatiles or byproducts. Furthermore, the impact of pressure parameter changes on the overall system's material balance is analyzed, particularly the solubility and volatility of reactants. The material balance equations are then adjusted to reflect the influence of pressure parameter changes on the reaction process and product separation, i.e., calculating the theoretical and actual yields of lithium carbonate under different pressure conditions, optimizing raw material usage and product recovery processes to improve the yield and purity of lithium carbonate.

[0072] The material balance equations are converted into mathematical models for material balance analysis in the lithium carbonate production process. The mathematical models generated in the above process can be numerically solved using specialized mathematical software or programming languages ​​(such as MATLAB and Python).

[0073] Optionally, the physical model for lithium carbonate preparation includes a reaction kinetic model, a thermodynamic model, and a material balance model. The methods for constructing the reaction kinetic model include:

[0074] Obtain the chemical reaction equations for the lithium carbonate preparation process and, using chemical reaction kinetics principles, determine the reaction rate equations. This process involves laboratory tests to determine the reaction order and rate constants. If pressure parameters significantly affect the reaction rate, a pressure term should be included in the rate equation. For example, consider the effect of pressure parameters on reactant concentration and activation energy.

[0075] Specifically, the main chemical reactions in the preparation of lithium carbonate are studied and determined. For example, lithium carbonate can be prepared by the reaction of carbonic acid and lithium salt. The corresponding chemical reaction equations are written, and the reaction rate equations are determined based on the principles of chemical reaction kinetics.

[0076] Based on historical production data of lithium carbonate under different production conditions, the rate constants of the reaction rate equation under different conditions were calculated. Specifically, the Arrhenius equation was used to calculate the reaction rate constants at different temperatures and pressures, and regression analysis was performed on the historical production data to accurately determine the parameters in the Arrhenius equation, such as the activation energy and pre-exponential factor.

[0077] A reaction kinetic model is constructed by combining the reaction rate equation and the material balance equation. This model includes all key reactants and products, as well as the relationships between these key reactants and products and time and pressure parameters.

[0078] Optionally, the physical model for lithium carbonate preparation includes a reaction kinetics model, a thermodynamic model, and a material balance model. The thermodynamic model can be constructed using the following methods:

[0079] Thermodynamic parameters of lithium carbonate under different production conditions are obtained, including enthalpy change, entropy change, and Gibbs free energy change. Specifically, it is necessary to identify the main thermodynamic processes in the lithium carbonate preparation process. For example, the exothermic or endothermic characteristics of the reaction, evaporation and condensation processes, etc., should be considered.

[0080] Using the chemical equilibrium constant expression and the aforementioned thermodynamic parameters, the equilibrium constant of lithium carbonate under different production conditions is calculated to construct a thermodynamic model. This thermodynamic model is then applied to the lithium carbonate production process to guide the optimization of production conditions, such as adjusting pressure parameters to improve yield and product quality.

[0081] Specifically, firstly, based on the chemical reaction equation and thermodynamic data tables, the standard enthalpy change and entropy change of the reaction are calculated. Considering the influence of pressure parameters on the enthalpy change, relevant thermodynamic formulas (such as the van Hoff equation) are used to calculate the enthalpy change under different pressures. Secondly, Le Chatelier's principle is applied to analyze how pressure changes affect the reaction equilibrium. Using the chemical equilibrium constant expression, combined with enthalpy change and entropy change data, the equilibrium constant under different pressures is calculated. Finally, combining the thermodynamic equations and data, numerical methods are used to solve the model and predict the reaction behavior under different pressure parameters.

[0082] Optionally, the numerical solution and optimization methods for the physical model of lithium carbonate preparation include: selecting an appropriate numerical solution method; for complex nonlinear equations, numerical optimization algorithms such as gradient descent, Newton's method, advanced algorithms, or genetic algorithms may be required. The established kinetic, thermodynamic, and material balance equations are then transformed into a numerical model. In the above process, specialized mathematical software or programming languages ​​(such as MATLAB and Python) can be used to numerically solve the equations.

[0083] Example 2

[0084] Figure 2 This is a schematic diagram of a machine learning-based parameter optimization system for lithium carbonate preparation equipment, provided in an embodiment of this disclosure. The machine learning-based lithium carbonate preparation equipment parameter optimization system 200 includes:

[0085] Data acquisition module 201 is used to acquire historical production data of lithium carbonate under different production conditions. The historical production data includes equipment production parameters and corresponding production result parameters.

[0086] Data construction module 202 is used to build training and validation datasets based on historical production data;

[0087] The model training module 203 is used to train a preset machine learning model using a training dataset to obtain the trained machine learning model, and to validate and optimize the trained machine learning model using a validation set to obtain a target machine learning model for optimizing lithium carbonate preparation parameters.

[0088] In some embodiments, the method of using the model output by the model training module 203 includes:

[0089] The parameter optimization module is used to acquire the production data to be optimized and input the production data to be optimized into the target machine learning model to obtain the production data optimization result.

[0090] In some embodiments, the model training module 203 includes:

[0091] The model optimization module is used to select a support vector machine for the target machine learning model. The target machine learning model construction method includes: using a genetic algorithm to optimize the kernel function parameters and penalty factor to obtain the optimal penalty factor and optimal kernel function parameters.

[0092] In some embodiments, the model training module 203 includes:

[0093] The model building module is used to build a physical model for lithium carbonate production based on historical production data;

[0094] The model output module is used to run the physical model of lithium carbonate preparation under the production data to be optimized to obtain a real-time dataset;

[0095] The model stacking module is used to input real-time datasets into the target machine learning model and output production data optimization results in real time.

[0096] In some embodiments, the model training module 203 further includes:

[0097] The model building module is used to build a physical model for lithium carbonate production based on historical production data;

[0098] The model output module is used to run the lithium carbonate preparation physical model and the target machine learning model under the production data to be optimized, and obtain the first real-time dataset and the second real-time dataset respectively.

[0099] The model integration module is used to fuse the first real-time dataset and the second real-time dataset to obtain the optimized results of the production data.

[0100] In some embodiments, the model training module 203 includes:

[0101] The material identification module is used to determine the reactants and products in the lithium carbonate preparation process based on the chemical reaction equations in the lithium carbonate preparation process.

[0102] The equation building module is used to construct material balance equations based on the reactants and products of lithium carbonate under different production conditions.

[0103] The model conversion module is used to convert material balance equations into mathematical models for material balance analysis in the lithium carbonate preparation process.

[0104] In some embodiments, the model training module 203 includes:

[0105] The equation determination module is used to obtain the chemical reaction equations in the lithium carbonate preparation process and, in conjunction with the principles of chemical reaction kinetics, determine the reaction rate equations.

[0106] The data calculation module is used to calculate the rate constant of the reaction rate equation under different conditions based on historical production data of lithium carbonate under different production conditions.

[0107] The model building module is used to construct a reaction kinetic model by combining the reaction rate equation and the material balance equation.

[0108] In some embodiments, the model training module 203 includes:

[0109] The parameter acquisition module is used to acquire the thermodynamic parameters of lithium carbonate under different production conditions. The thermodynamic parameters include enthalpy change, entropy change and Gibbs free energy change.

[0110] The model building module is used to calculate the equilibrium constant of lithium carbonate under different production conditions by using the chemical equilibrium constant expression and combining thermodynamic parameters, so as to build a thermodynamic model.

[0111] The system of this disclosure embodiment can execute the method provided in this disclosure embodiment, and the implementation principle is similar. The actions performed by each module in the system of each disclosure embodiment correspond to the steps in the method of each disclosure embodiment. For detailed functional descriptions of each module of the system, please refer to the descriptions in the corresponding methods shown above, which will not be repeated here.

[0112] The above description is only an optional implementation method for some implementation scenarios of this disclosure. It should be noted that for those skilled in the art, other similar implementation methods based on the technical concept of this disclosure without departing from the technical concept of this disclosure also fall within the protection scope of the embodiments of this disclosure.

Claims

1. A method for optimizing parameters of lithium carbonate preparation equipment based on machine learning, characterized in that, The method includes: Acquire historical production data of lithium carbonate under different production conditions, including equipment production parameters and corresponding production result parameters; Based on the aforementioned historical production data, training and validation datasets are constructed. The training dataset is used to train a preset machine learning model to obtain the trained machine learning model. The training dataset is then used to validate and optimize the trained machine learning model to obtain a target machine learning model for optimizing lithium carbonate preparation parameters. Obtain the production data to be optimized, and input the production data to be optimized into the target machine learning model to obtain the production data optimization result; The step of acquiring the production data to be optimized and inputting the production data to be optimized into the target machine learning model to obtain the production data optimization result includes: Based on the historical production data, a physical model for lithium carbonate preparation was constructed. Under the production data to be optimized, a physical model for lithium carbonate preparation was run to obtain a real-time dataset; Input the real-time dataset into the target machine learning model, and output the production data optimization results in real time; The physical model for lithium carbonate preparation includes a reaction kinetics model, a thermodynamic model, and a material balance model.

2. The method for optimizing lithium carbonate preparation equipment parameters based on machine learning according to claim 1, characterized in that, The target machine learning model includes support vector machines, random forests, or neural networks.

3. The method for optimizing lithium carbonate preparation equipment parameters based on machine learning according to claim 2, characterized in that, The target machine learning model is a support vector machine, and the method for constructing the target machine learning model includes: A genetic algorithm is used to optimize the kernel function parameters and penalty factor to obtain the optimal penalty factor and optimal kernel function parameters.

4. The method for optimizing lithium carbonate preparation equipment parameters based on machine learning according to claim 1, characterized in that, The step of acquiring the production data to be optimized and inputting the production data to be optimized into the target machine learning model to obtain the production data optimization result further includes: Based on the historical production data, a physical model for lithium carbonate preparation was constructed. Under the production data to be optimized, the lithium carbonate preparation physical model and the target machine learning model are run to obtain the first real-time dataset and the second real-time dataset, respectively. The first real-time dataset and the second real-time dataset are merged to obtain the production data optimization results.

5. The method for optimizing lithium carbonate preparation equipment parameters based on machine learning according to claim 1 or 4, characterized in that, The method for constructing the material balance model includes: Based on the chemical reaction equations in the lithium carbonate preparation process, determine the reactants and products in the lithium carbonate preparation process; Based on the reactants and products of lithium carbonate under different production conditions, a material balance equation is constructed. The material balance equations are converted into mathematical models for material balance analysis in the lithium carbonate preparation process.

6. The method for optimizing lithium carbonate preparation equipment parameters based on machine learning according to claim 5, characterized in that, The method for constructing the reaction kinetics model includes: Obtain the chemical reaction equations in the lithium carbonate preparation process, and determine the reaction rate equations by combining the principles of chemical reaction kinetics. Based on historical production data of lithium carbonate under different production conditions, the rate constants of the reaction rate equation under different conditions were calculated. The reaction kinetic model is constructed by combining the reaction rate equation and the material balance equation.

7. The method for optimizing lithium carbonate preparation equipment parameters based on machine learning according to claim 5, characterized in that, The thermodynamic model construction method includes: Thermodynamic parameters of lithium carbonate under different production conditions are obtained, including enthalpy change, entropy change and Gibbs free energy change; Using the chemical equilibrium constant expression and the aforementioned thermodynamic parameters, the equilibrium constant of lithium carbonate under different production conditions was calculated to construct a thermodynamic model.

8. A parameter optimization system for lithium carbonate preparation equipment based on machine learning, characterized in that, include: The data acquisition module is used to acquire historical production data of lithium carbonate under different production conditions. The historical production data includes equipment production parameters and corresponding production result parameters. The data construction module is used to construct training datasets and validation datasets based on the historical production data. The model training module is used to train a preset machine learning model using the training dataset to obtain the trained machine learning model, and to validate and optimize the trained machine learning model using the validation dataset to obtain a target machine learning model for optimizing lithium carbonate preparation parameters. Obtain the production data to be optimized, and input the production data to be optimized into the target machine learning model to obtain the production data optimization result; The step of acquiring the production data to be optimized and inputting the production data to be optimized into the target machine learning model to obtain the production data optimization result includes: Based on the historical production data, a physical model for lithium carbonate preparation was constructed. Under the production data to be optimized, a physical model for lithium carbonate preparation was run to obtain a real-time dataset; Input the real-time dataset into the target machine learning model, and output the production data optimization results in real time; The physical model for lithium carbonate preparation includes a reaction kinetics model, a thermodynamic model, and a material balance model.

Images