Lithium extraction method for automated control of lithium mica concentrate roasting process

Abstract

This invention relates to the field of lithium extraction technology from lepidolite concentrate, and in particular to a method for automated control of the lepidolite concentrate roasting process. The method includes full-process control based on the initial parameter matrix of each process stage, determining the parameter matrix after control; performing process control for the target stage and collecting the actual parameter matrix; adjusting the target stage parameter matrix through a parameter deviation matrix when the difference exceeds a threshold until consistency requirements are met. A preferred scheme includes using a stage parameter relationship model, a parameter matrix iterative model, and a calibration weight table to optimize the control process, and constructing a mapping model to map the parameter fitting coefficients to the process stages. This invention can improve the automation level and control accuracy of the lepidolite concentrate roasting process, ensuring process stability and lithium extraction efficiency.

Description

Technical Field

[0001] This invention belongs to the field of metallurgy and chemical technology, specifically a method for lithium extraction by automatically controlling the roasting process of lithium mica concentrate. Background Technology

[0002] In the continuous development of lithium extraction processes from lepidolite concentrate, how to achieve efficient and low-cost lithium resource extraction through automated control has become a core issue of concern in the industry. Traditional lithium extraction processes from lepidolite concentrate typically employ high-temperature roasting (900-1100℃) and a sulfate system. While these methods can achieve a certain degree of lithium extraction, they suffer from high energy consumption, significant equipment wear and tear, and environmental impact. Therefore, developing a new method that can perform acid roasting using a sulfuric acid process at a lower temperature range (200-300℃) and combine it with automated control to improve process accuracy and efficiency is of significant practical importance.

[0003] The low-temperature, high-efficiency lithium extraction method from lepidolite lithium extraction waste (publication number CN116240400B) involves mixing the lepidolite lithium extraction waste with additives, followed by pressurized or microwave roasting. The leaching and lithium extraction processes are then completed at temperatures lower than traditional methods, ultimately producing battery-grade lithium carbonate. This technology significantly reduces roasting temperature and energy consumption while improving lithium conversion rate and overall recovery rate. However, this method is primarily designed for lepidolite lithium extraction waste and is not directly applicable to lepidolite concentrate, thus limiting its applicability. Furthermore, this approach lacks the application of automated control technology, potentially leading to inaccurate adjustment of process parameters in actual production, which could negatively impact the stability and efficiency of lithium extraction.

[0004] The above analysis shows that existing lithium extraction methods from lepidolite have made some progress in reducing roasting temperature, energy consumption, and improving lithium extraction efficiency, but there are still limitations in the application of automated control technology. Especially under complex operating conditions, the existing processes are insufficient in real-time monitoring and dynamic adjustment of key parameters (such as temperature and acidification level), which may lead to decreased process stability and affect lithium extraction efficiency. Therefore, this invention provides an automated control method for lithium extraction from lepidolite concentrate roasting, aiming to achieve precise control of the roasting process by introducing an industrial control system, optimizing the process temperature to the range of 200-300℃, and combining this with optimization of sulfuric acid process conditions, thereby improving lithium extraction efficiency, reducing energy consumption, and enhancing process stability, meeting the modern industrial demand for efficient and intelligent lithium extraction technology. Summary of the Invention

[0005] The purpose of this invention is to provide an automated method for controlling the roasting process of lithium mica concentrate to extract lithium, thereby solving the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides an automated method for lithium extraction from lithium mica concentrate roasting process, the method comprising:

[0007] At each process stage, based on the initial parameter matrix corresponding to each process stage, the roasting process of lithium mica concentrate is controlled throughout the entire process to obtain the controlled parameter matrix of the lithium mica concentrate at each process stage.

[0008] The target stage to be controlled in the roasting process of the lithium mica concentrate is determined, and the process control operation is executed based on the parameter matrix after control corresponding to the target stage. The actual parameter matrix of the lithium mica concentrate is collected through the industrial control system.

[0009] If the difference between the target stage and the actual parameter matrix is ​​greater than or equal to a preset difference threshold, a parameter deviation matrix is ​​determined based on the adjusted parameter matrix corresponding to the target stage and the original parameter matrix corresponding to the actual parameter matrix.

[0010] Based on the parameter deviation matrix, the adjusted parameter matrix corresponding to the target stage is adjusted to obtain the adjusted parameter matrix corresponding to the target stage.

[0011] The adjusted parameter matrix corresponding to the target stage is used as the new adjusted parameter matrix corresponding to the target stage, and the step of performing process control operation based on the adjusted parameter matrix corresponding to the target stage is returned until the difference between the target stage and the actual parameter matrix is ​​less than the preset difference threshold.

[0012] Preferably, the step of controlling the entire roasting process of lepidolite concentrate based on the initial parameter matrix corresponding to each of the process stages to obtain the controlled parameter matrix of the lepidolite concentrate under each of the process stages includes:

[0013] For any of the aforementioned process stages, based on the initial parameter matrix corresponding to the process stage, process control operations for the roasting process of the lithium mica concentrate are performed, and process data of the lithium mica concentrate are acquired using an industrial control system, from which the actual parameter matrix is ​​extracted.

[0014] Based on the actual parameter matrix, determine the process consistency variance;

[0015] When the process consistency variance is greater than or equal to a preset variance threshold, the actual parameter matrix corresponding to the actual parameter matrix is ​​determined based on a pre-constructed stage parameter relationship model; the stage parameter relationship model characterizes the correspondence between the process stage and the parameters.

[0016] The actual parameter matrix is ​​iterated using a parameter matrix iterative model to obtain the iterated parameter matrix corresponding to the actual parameter matrix;

[0017] The iterative parameter matrix is ​​used as the initial parameter matrix, and the process control operation of the lithium mica concentrate roasting process is performed based on the initial parameter matrix corresponding to the process stage until the process consistency variance is less than the preset variance threshold.

[0018] The iterative parameter matrix is ​​used as the adjusted parameter matrix.

[0019] Preferably, the step of iterating the actual parameter matrix using a parameter matrix iteration model to obtain the iterated parameter matrix corresponding to the actual parameter matrix includes:

[0020] Based on the actual parameter matrix and the initial parameter matrix, determine the deviation matrix for the current iteration round;

[0021] Obtain the calibration weight matrix corresponding to the current iteration round, and use the calibration weight matrix corresponding to the current iteration round to adjust the deviation matrix of the current iteration round, thereby obtaining the adjustment information of the current iteration round for the initial parameter matrix;

[0022] The initial parameter matrix is ​​adjusted using the adjustment information for the current iteration to obtain the parameter matrix after iteration.

[0023] Preferably, the calibration weight matrix includes calibration weights for each process node in the roasting process of the lepidolite concentrate; obtaining the calibration weight matrix corresponding to the current iteration round includes:

[0024] Obtain the deviation change direction information, deviation change magnitude information, and deviation change trend information for each process node in the current iteration round;

[0025] Based on the deviation change direction information, deviation change magnitude information, and deviation change trend information corresponding to each process node, an iterative state vector corresponding to each process node is generated.

[0026] Using the iterative state vector corresponding to each process node as an index, the calibration weight that matches each process node is queried in the pre-built calibration weight table.

[0027] Based on the calibration weights that match each of the process nodes, a calibration weight matrix corresponding to the current iteration round is generated.

[0028] Preferably, after the step of controlling the entire roasting process of lepidolite concentrate based on the initial parameter matrix corresponding to each of the process stages to obtain the controlled parameter matrix of the lepidolite concentrate under each of the process stages, the method further includes:

[0029] Based on the initial parameter matrix and the adjusted parameter matrix corresponding to each process stage, determine the parameter fitting coefficient matrix corresponding to each process stage.

[0030] A mapping model is constructed based on each process stage and the parameter fitting coefficient matrix corresponding to each process stage; the mapping model represents the mapping relationship between the process stage and the parameter fitting coefficient matrix.

[0031] Preferably, the step of adjusting the post-regulation parameter matrix corresponding to the target stage based on the parameter deviation matrix to obtain the adjusted parameter matrix corresponding to the target stage includes:

[0032] Based on the mapping model, determine the parameter fitting coefficient matrix for the target stage;

[0033] The parameter deviation matrix is ​​adjusted using the parameter fitting coefficient matrix to obtain the adjustment information of the post-control parameter matrix corresponding to the target stage;

[0034] The adjusted parameter matrix corresponding to the target stage is adjusted using the adjustment information to obtain the adjusted parameter matrix corresponding to the target stage.

[0035] Preferably, the step of extracting the actual parameter matrix from the process data includes normalizing the process data, mapping the attribute values ​​of each process node to the 0-1 range, and arranging them according to the node topology order to form a two-dimensional matrix structure.

[0036] Preferably, the method for determining the process consistency variance includes: calculating the difference between the data value of each process node and the data mean, squaring the difference, and taking the arithmetic mean of the squared values ​​of all nodes as the process consistency variance.

[0037] Preferably, the pre-constructed calibration weight table is formed through the following steps: collecting deviation change samples of each process node in historical process control, classifying and labeling the direction, amplitude and trend information in the samples, calculating the optimal calibration weight corresponding to each type of state vector, and establishing a mapping relationship table between state vectors and calibration weights.

[0038] Preferably, the step of constructing the mapping model includes: using the process stage as the independent variable and the corresponding parameter fitting coefficient matrix as the dependent variable, and using a multiple linear regression method to fit the functional relationship between the independent variable and the dependent variable to form the mapping model.

[0039] The automated control method for lithium extraction from lepidolite concentrate roasting has several beneficial effects, as detailed below:

[0040] 1. Improve automation and control precision: By controlling the entire process based on the initial parameter matrix of each process stage, combined with the real-time acquisition of the actual parameter matrix by the industrial control system, and using parameter deviation matrix, iterative model and other methods for dynamic adjustment, the automated and precise control of the roasting process of lithium mica concentrate is realized, reducing manual intervention and improving the control precision of key parameters (such as temperature, acidification degree, etc.).

[0041] 2. Ensure process stability: The difference between actual and target parameters is determined by calculating the process consistency variance. When the difference exceeds a threshold, iterative adjustments are made until the consistency requirements are met. Simultaneously, tools such as stage parameter relationship models and calibration weight tables are used to optimize the control process, effectively reducing fluctuations in process parameters and enhancing the stability of the entire roasting process.

[0042] 3. Improve lithium extraction efficiency: Optimize the roasting temperature to a lower range of 200-300℃, and combine it with the optimization of sulfuric acid process conditions. While reducing energy consumption, precise parameter control ensures that lithium in lepidolite concentrate is fully extracted, thereby improving the lithium conversion rate and total recovery rate, and thus improving lithium extraction efficiency.

[0043] 4. Enhanced process adaptability and scalability: The constructed mapping model realizes the mapping between parameter fitting coefficients and process stages. The calibration weight table is constructed based on historical data and can be updated, enabling the method to adapt to different process stages and complex operating conditions. It has good scalability and flexibility and can meet the needs of modern industry for efficient and intelligent lithium extraction technology. Attached Figure Description

[0044] Figure 1 This is a schematic diagram of the overall process of the lithium extraction method for the automated control of lithium mica concentrate roasting process provided in an embodiment of the present invention.

[0045] Figure 2 This is a flowchart illustrating the specific process of iteratively adjusting the parameter matrix based on the actual parameter matrix and the initial parameter matrix in an embodiment of the present invention.

[0046] Figure 3 This is a logic block diagram of the calibration weight matrix generation process in an embodiment of the present invention.

[0047] Figure 4 This is a flowchart illustrating the mapping model construction method in an embodiment of the present invention.

[0048] Figure 5 This is a flowchart of the process consistency variance calculation method in an embodiment of the present invention. Detailed Implementation

[0049] This invention provides an automated method for lithium extraction from lithium mica concentrate during roasting. This method achieves precise control of the entire process through parameter matrix adjustment across multiple process stages. (See attached diagram.) Figure 1 To be continued Figure 5 The specific embodiments of the present invention will be described in detail below.

[0050] During implementation, the entire roasting process of lepidolite concentrate needs to be controlled based on the initial parameter matrix. The overall operation of this process is shown in the attached figure. Figure 1 As shown, the initial parameter matrix contains the basic parameter values ​​for each process stage, which are determined based on historical data and experimental results. The initial parameter matrix is ​​input into the lepidolite concentrate roasting equipment through the industrial control system for process control operations. In actual operation, the industrial control system collects the actual process data of the lepidolite concentrate and extracts the actual parameter matrix from it. The extraction process of the actual parameter matrix includes normalizing the process data, mapping the attribute values ​​of each process node to the 0-1 range, and arranging them according to the node topology to form a two-dimensional matrix structure.

[0051] To ensure process consistency, after extracting the actual parameter matrix, it is necessary to calculate the process consistency variance. The method for calculating the process consistency variance is attached. Figure 5 As shown. The specific steps are as follows: First, calculate the difference between the data value of each process node and the data mean; then, square the difference; finally, take the arithmetic mean of the squared values ​​of all nodes as the process consistency variance. If the process consistency variance is greater than or equal to the preset variance threshold, it indicates that the parameters of the current process stage need further adjustment. At this time, the pre-constructed stage parameter relationship model will be used to determine the actual parameter matrix corresponding to the actual parameter matrix. The stage parameter relationship model represents the correspondence between process stages and parameters, and its construction process is based on historical data and experimental analysis.

[0052] After determining the actual parameter matrix, an iterative parameter matrix model is used to iterate it to obtain the iterated parameter matrix. (Appendix) Figure 2 This iterative process is described in detail. First, the deviation matrix for the current iteration is determined based on the actual parameter matrix and the initial parameter matrix. The deviation matrix reflects the difference between the actual parameters and the target parameters. Then, the calibration weight matrix corresponding to the current iteration is obtained, and this calibration weight matrix is ​​used to adjust the deviation matrix, yielding adjustment information for the initial parameter matrix. The process of generating the calibration weight matrix is ​​attached. Figure 3As shown, the process first obtains the deviation change direction, deviation change magnitude, and deviation change trend information for each process node in the current iteration. Then, based on this information, iterative state vectors corresponding to each process node are generated. These iterative state vectors are used as indices to query the pre-built calibration weight table for calibration weights that match each process node. The calibration weight table is formed by collecting and classifying deviation change samples from historical process control, where the optimal calibration weight for each type of state vector is statistically analyzed and a mapping relationship is established. Finally, a calibration weight matrix corresponding to the current iteration is generated based on the calibration weights that match each process node.

[0053] After calibrating the deviation matrix, the initial parameter matrix is ​​adjusted using the adjustment information to obtain the iterated parameter matrix. This iterated parameter matrix serves as the new initial parameter matrix, and the process control operation for the lithium mica concentrate roasting process is returned to the initial parameter matrix until the process consistency variance is less than a preset variance threshold. When this condition is met, the iterated parameter matrix is ​​considered the controlled parameter matrix.

[0054] After completing the overall process control, it is necessary to determine the target stage for the roasting process of lepidolite concentrate. Based on the post-control parameter matrix corresponding to the target stage, process control operations are executed, and the actual parameter matrix of the lepidolite concentrate is collected through the industrial control system. If the difference between the target stage and the actual parameter matrix is ​​greater than or equal to a preset difference threshold, a parameter deviation matrix needs to be determined based on the post-control parameter matrix corresponding to the target stage and the original parameter matrix corresponding to the actual parameter matrix. The generation process of the parameter deviation matrix is ​​consistent with the calculation method of the deviation matrix described above.

[0055] Based on the parameter deviation matrix, the post-regulation parameter matrix corresponding to the target stage needs to be adjusted to obtain the adjusted parameter matrix corresponding to the target stage. The adjustment process is attached. Figure 4 As shown, firstly, the parameter fitting coefficient matrix 0 for the target stage is determined based on the mapping model. The mapping model characterizes the mapping relationship between the process stage and the parameter fitting coefficient matrix 0. Its construction process includes using the process stage as the independent variable and the corresponding parameter fitting coefficient matrix 0 as the dependent variable, and using a multiple linear regression method to fit the functional relationship between the independent and dependent variables. Subsequently, the parameter deviation matrix is ​​adjusted using the parameter fitting coefficient matrix 0 to obtain the adjustment information of the adjusted parameter matrix corresponding to the target stage. Finally, the adjusted information is used to adjust the adjusted parameter matrix corresponding to the target stage to obtain the adjusted parameter matrix corresponding to the target stage.

[0056] The adjusted parameter matrix corresponding to the target stage is used as the new adjusted parameter matrix corresponding to the target stage. The process of executing process control operations based on the adjusted parameter matrix corresponding to the target stage is then repeated until the difference between the target stage and the actual parameter matrix is ​​less than a preset difference threshold. Through this closed-loop control process, automated and precise control of the lithium mica concentrate roasting process can be achieved.

[0057] Throughout the implementation process, a close working relationship exists between the initial parameter matrix, actual parameter matrix, adjusted parameter matrix, parameter deviation matrix, calibration weight matrix, mapping model, process consistency variance, iterated parameter matrix, calibration weight table, and parameter fitting coefficient matrix. The initial parameter matrix provides the basic parameter values; the actual parameter matrix is ​​collected in real-time by the industrial control system and reflects the actual process state; and the adjusted parameter matrix contains optimized parameter values ​​obtained after multiple iterations. The parameter deviation matrix quantifies the difference between the target value and the actual value, and the calibration weight matrix achieves dynamic parameter calibration by adjusting the deviation matrix. The mapping model and parameter fitting coefficient matrix work together to adjust parameters in the target stage, ensuring optimal control at each process stage. The process consistency variance and iterated parameter matrix are used to evaluate process consistency and optimize the parameter matrix, respectively. The calibration weight table provides data support for the generation of the calibration weight matrix, thereby ensuring the stability and reliability of the entire system.

[0058] To enable those skilled in the art to fully understand and implement this invention, the following supplementary explanation of the specific implementation principle of this invention is provided in conjunction with a specific application scenario.

[0059] In the actual production of lithium extraction from lepidolite concentrate by roasting, the process flow first needs to be initialized based on an initial parameter matrix. The parameter values ​​in the initial parameter matrix are derived from historical experimental data and industrial experience, covering key variables such as temperature, acidification degree, and reaction time, and are input into the roasting equipment through the industrial control system. For example, in a certain process stage, the initial parameter matrix sets the temperature to 250℃, the acidifier concentration to 3 mol / L, and the reaction time to 60 minutes. These parameters are controlled in real time by the industrial control system, and actual operating data is collected through sensors. The collected data, after normalization, forms the actual parameter matrix, whose structure is consistent with the initial parameter matrix but reflects the process parameters under actual operating conditions.

[0060] To ensure process consistency, the process consistency variance needs to be calculated, and the calculation process is shown in the attached figure. Figure 5As shown, taking a specific process node as an example, assuming the temperature data for this node are 248℃, 250℃, and 252℃, then its mean is 250℃. The differences between each data value and the mean are -2℃, 0℃, and 2℃, respectively, which, when squared, yield 4, 0, and 4. Taking the arithmetic mean of these squared values, the final process consistency variance is 2.67. If this variance is greater than a preset threshold (e.g., 1.5), it indicates that the current process parameters have issues.

[0061] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0062] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for lithium extraction through automated control of the roasting process of lepidolite concentrate, characterized in that, The method includes: At each process stage, based on the initial parameter matrix corresponding to each process stage, the roasting process of lepidolite concentrate is controlled throughout the entire process to obtain the controlled parameter matrix of the lepidolite concentrate at each process stage, including: For any of the aforementioned process stages, based on the initial parameter matrix corresponding to the process stage, process control operations for the roasting process of the lithium mica concentrate are performed, and process data of the lithium mica concentrate are acquired using an industrial control system, from which the actual parameter matrix is ​​extracted. Based on the actual parameter matrix, determine the process consistency variance; If the process consistency variance is greater than or equal to a preset variance threshold, a process parameter relationship matrix associated with the actual parameter matrix is ​​determined based on a pre-constructed stage parameter relationship model; the stage parameter relationship model characterizes the correspondence between the process stage and the parameter. The actual parameter matrix is ​​iterated using a parameter matrix iteration model to obtain the iterated parameter matrix corresponding to the actual parameter matrix, including: Based on the actual parameter matrix and the initial parameter matrix, determine the deviation matrix for the current iteration round; Obtaining the calibration weight matrix corresponding to the current iteration round includes: Obtain the deviation change direction information, deviation change magnitude information, and deviation change trend information for each process node in the current iteration round; Based on the deviation change direction information, deviation change magnitude information, and deviation change trend information corresponding to each process node, an iterative state vector corresponding to each process node is generated. Using the iterative state vector corresponding to each process node as an index, the calibration weight that matches each process node is queried in the pre-built calibration weight table. Based on the calibration weights that match each of the process nodes, a calibration weight matrix corresponding to the current iteration round is generated; The deviation matrix of the current iteration round is adjusted using the calibration weight matrix corresponding to the current iteration round to obtain the adjustment information of the current iteration round for the initial parameter matrix; Using the adjustment information for the initial parameter matrix in the current iteration, the initial parameter matrix is ​​adjusted to obtain the iterated parameter matrix. The iterative parameter matrix is ​​used as the initial parameter matrix, and the process control operation of the lithium mica concentrate roasting process based on the initial parameter matrix corresponding to the process stage is returned until the process consistency variance is less than the preset variance threshold. The parameter matrix after iteration is used as the parameter matrix after adjustment. The target stage to be controlled in the roasting process of the lithium mica concentrate is determined, and the process control operation is executed based on the parameter matrix after control corresponding to the target stage. The actual parameter matrix of the lithium mica concentrate is collected through the industrial control system. If the difference between the target stage and the actual parameter matrix is ​​greater than or equal to a preset difference threshold, a parameter deviation matrix is ​​determined based on the adjusted parameter matrix corresponding to the target stage and the original parameter matrix corresponding to the actual parameter matrix. Based on the parameter deviation matrix, the adjusted parameter matrix corresponding to the target stage is adjusted to obtain the adjusted parameter matrix corresponding to the target stage. The adjusted parameter matrix corresponding to the target stage is used as the new adjusted parameter matrix corresponding to the target stage, and the step of performing process control operation based on the adjusted parameter matrix corresponding to the target stage is returned until the difference between the target stage and the actual parameter matrix is ​​less than the preset difference threshold.

2. The method for lithium extraction in the automated controlled roasting process of lithium mica concentrate according to claim 1, characterized in that, After the step of controlling the entire roasting process of lepidolite concentrate based on the initial parameter matrix corresponding to each of the process stages to obtain the controlled parameter matrix of the lepidolite concentrate under each of the process stages, the method further includes: Based on the initial parameter matrix and the adjusted parameter matrix corresponding to each process stage, determine the parameter fitting coefficient matrix corresponding to each process stage. A mapping model is constructed based on each process stage and the parameter fitting coefficient matrix corresponding to each process stage; the mapping model represents the mapping relationship between the process stage and the parameter fitting coefficient matrix.

3. The method for lithium extraction in the automated controlled roasting process of lithium mica concentrate according to claim 2, characterized in that, The step of adjusting the post-regulation parameter matrix corresponding to the target stage based on the parameter deviation matrix to obtain the adjusted parameter matrix corresponding to the target stage includes: Based on the mapping model, determine the parameter fitting coefficient matrix for the target stage; The parameter deviation matrix is ​​adjusted using the parameter fitting coefficient matrix to obtain the adjustment information of the post-control parameter matrix corresponding to the target stage; The adjusted parameter matrix corresponding to the target stage is adjusted using the adjustment information to obtain the adjusted parameter matrix corresponding to the target stage.

4. The method for lithium extraction in the automated controlled roasting process of lithium mica concentrate according to claim 1, characterized in that, The step of extracting the actual parameter matrix from the process data includes normalizing the process data, mapping the attribute values ​​of each process node to the 0-1 range, and arranging them according to the node topology order to form a two-dimensional matrix structure.

5. The method for lithium extraction in the automated controlled roasting process of lithium mica concentrate according to claim 1, characterized in that, The method for determining the process consistency variance includes: calculating the difference between the data value of each process node and the data mean, squaring the difference, and taking the arithmetic mean of the squared values ​​of all nodes as the process consistency variance.

6. The method for lithium extraction in the automated controlled roasting process of lithium mica concentrate according to claim 1, characterized in that, The pre-constructed calibration weight table is formed through the following steps: collecting deviation change samples of each process node in historical process control, classifying and labeling the direction, amplitude and trend information in the samples, calculating the optimal calibration weight corresponding to each type of state vector, and establishing a mapping relationship table between state vectors and calibration weights.

7. The method for lithium extraction in the automated controlled roasting process of lithium mica concentrate according to claim 2, characterized in that, The steps for constructing the mapping model include: using the process stage as the independent variable and the corresponding parameter fitting coefficient matrix as the dependent variable, and using the multiple linear regression method to fit the functional relationship between the independent and dependent variables to form the mapping model.

Images