Lepidolite roasting process optimization method and device, electronic equipment and storage medium

Abstract

The invention provides a lepidolite roasting process optimization method and device, electronic equipment and a storage medium, and the method comprises the steps: obtaining a plurality of pieces of sample data, each piece of sample data including a target parameter value corresponding to a decision parameter variable in a lepidolite roasting process and index data corresponding to a preset index, the preset indexes comprise at least one of the following indexes: lepidolite transformation rate, energy consumption intensity and carbon emission intensity; constructing a mapping function between the decision parameter variable and a preset index according to the multiple pieces of sample data; and by taking an index target corresponding to the preset index as an optimization target, optimizing the decision parameter variable based on a preset optimization algorithm and the mapping function to obtain a target parameter set corresponding to the decision parameter variable. According to the embodiment provided by the scheme, the technological parameters in the roasting process of the lepidolite rotary kiln can be optimized, the reliability of the technological parameters in the roasting process of the lepidolite rotary kiln is improved, and then the roasting stability of the lepidolite is improved.

Description

Technical Field

[0001] The present application relates to the technical field of lithium mica roasting, and in particular to a lithium mica roasting process optimization method, device, electronic equipment and storage medium. Background Art

[0002] Lithium and lithium compounds are widely used in new energy vehicles, aerospace, nuclear industry, medicine, petrochemicals and other fields. They are known as "new energy metals". Lithium is an important element to promote the development of the 21st century. Lithium mica is one of the mineral raw materials for extracting lithium. In related technologies, the lithium extraction process is usually a rotary kiln roasting of lithium mica assisted by sulfuric acid mixed salt. During the roasting process, a certain amount of potassium, sodium and calcium mixed salt additives are generally added. However, the distribution ratio of each component is diverse, and the mechanism of the roasting process is not clear. Affected by the multivariable strong coupling, nonlinear and large time-delay reaction process of "pyrolysis-exchange-phase change" of complex components such as lithium, potassium, sodium and calcium in the mixture, the roasting process still has problems such as high energy consumption, large carbon emissions and large fluctuations in clinker transformation rate, resulting in low production efficiency. Summary of the invention

[0003] The embodiments of the present application provide a method, device, electronic device and storage medium for optimizing the calcination process of lithium mica, which can effectively improve production efficiency and enhance the intelligence and automation level of the calcination process of lithium mica. The above technical solutions are as follows: In a first aspect, an embodiment of the present application provides a method for optimizing a calcination process of lithium mica, the method comprising: Acquire multiple sample data, each of the multiple sample data includes a target parameter value corresponding to a decision parameter variable in a lithium mica roasting process, and indicator data corresponding to a preset indicator, wherein the preset indicator includes at least one of the following: lithium mica transformation rate, energy consumption intensity, and carbon emission intensity; Constructing a mapping function between the decision parameter variables and the preset indicators according to the plurality of sample data; Taking the indicator target corresponding to the above preset indicator as the optimization target, the above decision parameter variables are optimized based on the preset optimization algorithm and the above mapping function to obtain the target parameter set corresponding to the above decision parameter variables.

[0004] In a possible implementation, the method further includes: Acquire multiple groups of historical operation data within a preset time period, each group of the multiple groups of historical operation data including multiple parameter values ​​corresponding to multiple parameter variables in the calcination process of the lithium mica; Calculate the indicator data corresponding to each group of historical operation data based on the preset indicators; According to the above multiple groups of historical operation data and the indicator data corresponding to each group of historical operation data, the above multiple parameter variables are screened to obtain decision parameter variables and target parameter values ​​corresponding to the above decision parameter variables.

[0005] In a possible implementation, the above-mentioned multiple parameter variables include at least one of the following: lithium oxide content in the mixture, water content in the mixture, calcium content in the mixture, potassium content in the mixture, kiln tail temperature, natural gas flow, kiln head wind pressure, and oxygen concentration.

[0006] In a possible implementation, the above-mentioned multiple parameter variables are screened according to the above-mentioned multiple groups of historical operation data and the indicator data corresponding to the above-mentioned each group of historical operation data to obtain the decision parameter variables and the target parameter values ​​corresponding to the above-mentioned decision parameter variables, including: For each parameter variable among the plurality of parameter variables, determining a correlation degree between the parameter variable and each indicator data; When the correlation between the parameter variable and each indicator data is greater than a preset correlation threshold, the parameter variable is determined as a decision parameter variable; The parameter value corresponding to the above decision parameter variable is determined as the target parameter value corresponding to the above decision parameter variable.

[0007] In a possible implementation, the step of constructing a mapping function between the decision parameter variable and the preset index according to the plurality of sample data includes: Using the above decision parameter variables as input variables and the above preset indicators as output variables, an initial neural network model is constructed; Training the initial neural network model based on the multiple sample data to obtain a trained target neural network model; The target neural network model is determined as a mapping function between the decision parameter variables and the preset indicators.

[0008] In a possible implementation, the method further includes: In the case where the preset index includes the lepidolite transformation rate, maximizing the lepidolite transformation rate is used as the optimization target corresponding to the lepidolite transformation rate; In the case where the above-mentioned preset indicators include the above-mentioned energy consumption intensity, minimizing the above-mentioned energy consumption intensity is taken as the optimization target corresponding to the above-mentioned energy consumption intensity; In the case where the above-mentioned preset indicators include the above-mentioned carbon emission intensity, minimizing the above-mentioned carbon emission intensity is taken as the optimization target corresponding to the above-mentioned energy consumption intensity.

[0009] In a possible implementation, the preset optimization algorithm is a multi-objective genetic algorithm, the indicator target corresponding to the preset indicator is used as the optimization target, the decision parameter variables are optimized based on the preset optimization algorithm and the mapping function, and the target parameter set corresponding to the decision parameter variables is obtained, including: Determine the fitness function according to the above mapping function; Determine an initial population, wherein the initial population includes a plurality of candidate solutions; Determine the dominance relationship between the candidate solutions in the initial population based on preset rules; According to the above dominance relationship, the above candidate solutions are selected, crossed and mutated to generate new candidate solutions; Based on each of the above-mentioned new candidate solutions and the above-mentioned fitness function, the mapping indicator data corresponding to the above-mentioned parameter value is determined until the preset conditions are met, and the optimal solution of the above-mentioned decision parameter variables is output as the target parameter set corresponding to the above-mentioned decision parameter variables, wherein the above-mentioned preset conditions include that the number of iterations reaches the preset number of iterations or the above-mentioned mapping indicator data meets the preset threshold conditions.

[0010] In a second aspect, an embodiment of the present application provides a device for optimizing the calcination process of lithium mica, the device comprising: A first acquisition module is used to acquire a plurality of sample data, each of which includes a target parameter value corresponding to a decision parameter variable in a lithium mica roasting process, and indicator data corresponding to a preset indicator, wherein the preset indicator includes at least one of the following: lithium mica transformation rate, energy consumption intensity, and carbon emission intensity; A construction module, used to construct a mapping function between the decision parameter variables and the preset indicators according to the multiple sample data; The optimization module is used to optimize the above-mentioned decision parameter variables based on the preset optimization algorithm and the above-mentioned mapping function with the indicator target corresponding to the above-mentioned preset indicator as the optimization target, so as to obtain the target parameter set corresponding to the above-mentioned decision parameter variables.

[0011] In a possible implementation, the above device further includes: A second acquisition module is used to acquire multiple groups of historical operation data within a preset time period, each group of the multiple groups of historical operation data including multiple parameter values ​​corresponding to multiple parameter variables in the above-mentioned lithium mica roasting process; A calculation module, used to calculate the indicator data corresponding to each group of historical operation data based on preset indicators; The screening module is used to screen the above-mentioned multiple parameter variables according to the above-mentioned multiple groups of historical operation data and the indicator data corresponding to the above-mentioned groups of historical operation data, so as to obtain decision parameter variables and target parameter values ​​corresponding to the above-mentioned decision parameter variables.

[0012] In a possible implementation, the above-mentioned multiple parameter variables include at least one of the following: lithium oxide content in the mixture, water content in the mixture, calcium content in the mixture, potassium content in the mixture, kiln tail temperature, natural gas flow, kiln head wind pressure, and oxygen concentration.

[0013] In a possible implementation, the screening module includes: A first determining unit, configured to determine, for each parameter variable among the plurality of parameter variables, a correlation between the parameter variable and each indicator data; A second determining unit is used to determine the parameter variable as a decision parameter variable when the correlation between the parameter variable and each indicator data is greater than a preset correlation threshold; The third determining unit is used to determine the parameter value corresponding to the above-mentioned decision parameter variable as the target parameter value corresponding to the above-mentioned decision parameter variable.

[0014] In a possible implementation, the above-mentioned building blocks include: A construction unit, used to construct an initial neural network model using the above decision parameter variables as input variables and the above preset indicators as output variables; A training unit, used to train the initial neural network model based on the plurality of sample data to obtain a trained target neural network model; The fourth determination unit is used to determine the above-mentioned target neural network model as a mapping function between the above-mentioned decision parameter variables and the above-mentioned preset indicators.

[0015] In a possible implementation, the above device further includes: A first determination module is used for maximizing the lepidolite transformation rate as an optimization target corresponding to the lepidolite transformation rate when the preset index includes the lepidolite transformation rate; A second determination module is used for, when the preset index includes the energy consumption intensity, minimizing the energy consumption intensity as an optimization target corresponding to the energy consumption intensity; The third determination module is used to, when the preset index includes the carbon emission intensity, minimize the carbon emission intensity as the optimization target corresponding to the energy consumption intensity.

[0016] In a possible implementation, the preset optimization algorithm is a multi-objective genetic algorithm, and the optimization module includes: A fifth determining unit, used to determine a fitness function according to the mapping function; A sixth determining unit, configured to determine an initial population, wherein the initial population includes a plurality of candidate solutions; A seventh determination unit, configured to determine a dominance relationship between candidate solutions in the initial population based on a preset rule; A processing unit, configured to perform selection, crossover and mutation processing on the candidate solutions according to the dominance relationship to generate new candidate solutions; The eighth determination unit is used to determine the mapping indicator data corresponding to the above-mentioned parameter value based on each of the above-mentioned new candidate solutions and the above-mentioned fitness function until the preset conditions are met, and output the optimal solution of the above-mentioned decision parameter variables as the target parameter set corresponding to the above-mentioned decision parameter variables, wherein the above-mentioned preset conditions include the number of iterations reaching the preset number of iterations or the above-mentioned mapping indicator data meeting the preset threshold conditions.

[0017] In a third aspect, an embodiment of the present application provides an electronic device, including: a processor and a memory; The above-mentioned memory stores a computer program, and the above-mentioned computer program is suitable for being loaded by the above-mentioned processor and executing the steps of the method provided by the first aspect of the embodiment of the present application or any possible implementation method of the first aspect.

[0018] In a fourth aspect, an embodiment of the present application provides a computer storage medium, which stores multiple instructions, and the instructions are suitable for being loaded by a processor and executing the steps of the method provided by the first aspect of the embodiment of the present application or any possible implementation method of the first aspect.

[0019] The embodiment of the present application obtains multiple sample data, each of which includes the target parameter value corresponding to the decision parameter variable in the lithium mica roasting process, and the index data corresponding to the preset index, and the preset index includes at least one of the following: lithium mica transformation rate, energy consumption intensity and carbon emission intensity; the mapping function between the decision parameter variable and the preset index is constructed according to the multiple sample data; the index target corresponding to the preset index is used as the optimization target, and the decision parameter variable is optimized based on the preset optimization algorithm and the mapping function to obtain the target parameter set corresponding to the decision parameter variable. By constructing the mapping function between the decision parameter variable and the preset index, the influence of different decision parameters on the index result can be quantified and clearly described, and the preset index (such as lithium mica transformation rate, energy consumption intensity, carbon emission intensity) is used as the optimization target, and the decision parameter is adjusted through the optimization algorithm, and multiple indicators can be considered at the same time to achieve comprehensive optimization, so as to better guide the actual production, help maintain the stability and controllability of the production process, reduce the fluctuation caused by improper parameter adjustment, effectively improve the production efficiency, and improve the intelligence and automation level of the lithium mica roasting process. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0021] Figure 1 A schematic diagram of the structure of a lithium mica roasting process optimization system provided for an exemplary embodiment of the present application; Figure 2 A schematic flow chart of a method for optimizing a calcination process of lithium mica provided as an exemplary embodiment of the present application; Figure 3 A flowchart of a method for determining a decision parameter variable provided by an exemplary embodiment of the present application; Figure 4 A schematic diagram of the structure of a target neural network model provided for an exemplary embodiment of the present application; Figure 5 A schematic diagram of the effect of multi-objective optimization provided by an exemplary embodiment of the present application; Figure 6 A schematic diagram of the structure of a device for optimizing the calcination process of lithium mica provided in an exemplary embodiment of the present application; Figure 7 A schematic structural diagram of an electronic device provided as an exemplary embodiment of the present application. DETAILED DESCRIPTION

[0022] The technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application.

[0023] The terms "first", "second", "third", etc. in the specification and claims of this application and the above-mentioned drawings are used to distinguish different objects, rather than to describe a specific order. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but optionally includes steps or units that are not listed, or optionally includes other steps or units inherent to these processes, methods, products or devices.

[0024] Please refer to the following Figure 1 , Figure 1 A structural schematic diagram of a lithium mica roasting process optimization system provided in an embodiment of the present application, wherein the lithium mica roasting process optimization system includes a terminal 10 and a server 20, etc.; the terminal 10 and the server 20 are connected via a network, for example, via a wired or wireless network connection, etc.

[0025] Among them, the terminal 10 can be used to display a graphical user interface. Among them, the terminal 10 is used to interact with the user through the graphical user interface, for example, downloading and installing the corresponding client through the terminal 10 and running it. In the embodiment of the present application, the terminal 10 can also be used by relevant personnel to upload multiple sample data, and the terminal 10 can send the multiple sample data to the server 20, so that the server 20 constructs the mapping function between the above-mentioned decision parameter variables and the above-mentioned preset indicators based on the multiple sample data, and feeds back the obtained mapping function to the terminal 10.

[0026] Optionally, the server 20 may also store the mapping function locally, and after obtaining the plurality of sample data and the mapping function, optimize the decision parameter variables based on the preset optimization algorithm and the mapping function with the indicator target corresponding to the preset indicator as the optimization target, and obtain the target parameter set corresponding to the decision parameter variable. After determining the target parameter set, the server 20 may store, distribute, and perform other processing on the target parameter set.

[0027] Optionally, the terminal 10 itself can also directly construct a mapping function between the above-mentioned decision parameter variables and the above-mentioned preset indicators based on multiple sample data uploaded by related objects, and take the indicator target corresponding to the above-mentioned preset indicators as the optimization target, optimize the above-mentioned decision parameter variables based on the preset optimization algorithm and the above-mentioned mapping function, and obtain the target parameter set corresponding to the above-mentioned decision parameter variables.

[0028] Optionally, the terminal 10 may also send the mapping function it constructs to the server 20, so that after the server 20 obtains multiple sample data, it takes the indicator target corresponding to the above-mentioned preset indicator as the optimization target, optimizes the above-mentioned decision parameter variables based on the preset optimization algorithm and the above-mentioned mapping function, and obtains the target parameter set corresponding to the above-mentioned decision parameter variables.

[0029] The following uses the server 20 for constructing a mapping function as an example to illustrate the present solution. Specifically, the server 20 can be used to: obtain multiple sample data, each of the multiple sample data includes a target parameter value corresponding to a decision parameter variable in the lithium mica roasting process, and indicator data corresponding to a preset indicator, and the preset indicator includes at least one of the following: lithium mica transformation rate, energy consumption intensity, and carbon emission intensity; construct a mapping function between the above-mentioned decision parameter variables and the above-mentioned preset indicators based on the above-mentioned multiple sample data; take the indicator target corresponding to the above-mentioned preset indicator as the optimization target, optimize the above-mentioned decision parameter variables based on the preset optimization algorithm and the above-mentioned mapping function, and obtain the target parameter set corresponding to the above-mentioned decision parameter variables.

[0030] An exemplary embodiment of the present application provides a method for optimizing the calcination process of lithium mica. The method for optimizing the calcination process of lithium mica can be applied to the above-mentioned terminal. For details, please refer to Figure 2 , which exemplarily shows a flow chart of a method for optimizing the calcination process of lithium mica provided in an embodiment of the present application. Figure 2 As shown, the lithium mica roasting process optimization method includes the following S21-S23: S21. Acquire multiple sample data, each of the multiple sample data includes a target parameter value corresponding to a decision parameter variable in the lithium mica roasting process, and indicator data corresponding to a preset indicator, and the preset indicator includes at least one of the following: lithium mica transformation rate, energy consumption intensity and carbon emission intensity.

[0031] In some embodiments, decision parameter variables refer to variables that can be actively adjusted and optimized by operators or related control systems during industrial production processes (such as lithium mica roasting). The adjustment of these variables is intended to achieve specific production goals, such as improving product quality, reducing energy consumption, reducing carbon emissions, etc.

[0032] In some embodiments, the preset index is a key parameter pre-set during the lepidolite roasting process for measuring and evaluating production efficiency, quality, environmental impact, etc. During the lepidolite rotary kiln roasting process, the preset index is used to detect and optimize the roasting process to ensure the achievement of production goals.

[0033] In some embodiments, the indicator data corresponding to the preset indicator is the actual value corresponding to the preset indicator recorded during the calcination of lithium mica, that is, the operating value, for example, the operating value of the lithium mica transformation rate, the operating value of the energy consumption intensity, and the operating value of the carbon emission intensity.

[0034] Among them, the lithium mica conversion rate refers to the proportion of lithium mica raw materials successfully converted into the required products during the rotary kiln roasting process. Energy consumption intensity refers to the amount of energy consumed to produce a unit of product (lithium). Carbon emission intensity refers to the amount of carbon dioxide emissions generated by producing a unit of product (lithium) within a certain period of time.

[0035] S22. Constructing a mapping function between the decision parameter variables and the preset indicators according to the plurality of sample data.

[0036] In some embodiments, the mapping function between the decision parameter variables and the preset indicators can be a mathematical or computational model for describing how the decision parameter variables affect the preset indicators. The mapping function can be linear, nonlinear, or a complex model based on machine learning.

[0037] S23. Taking the indicator target corresponding to the above-mentioned preset indicator as the optimization target, the above-mentioned decision parameter variables are optimized based on the preset optimization algorithm and the above-mentioned mapping function to obtain the target parameter set corresponding to the above-mentioned decision parameter variables.

[0038] In some embodiments, when the preset index includes the lepidolite transformation rate, maximizing the lepidolite transformation rate is used as the optimization target corresponding to the lepidolite transformation rate.

[0039] In some embodiments, when the above-mentioned preset indicators include the above-mentioned energy consumption intensity, minimizing the above-mentioned energy consumption intensity is used as the optimization target corresponding to the above-mentioned energy consumption intensity.

[0040] In some embodiments, when the above-mentioned preset indicators include the above-mentioned carbon emission intensity, minimizing the above-mentioned carbon emission intensity is used as the optimization target corresponding to the above-mentioned energy consumption intensity.

[0041] In some embodiments, the relationship between the decision parameter variables and the preset indicators is relatively complex and nonlinear, and a global optimization algorithm such as a genetic algorithm or a particle swarm optimization can be selected to optimize the above decision parameter variables to obtain a target parameter set corresponding to the above decision parameter variables. The target parameter set corresponding to the above decision parameter variables can be a better combination of decision parameter variables, so that the preset indicators are as close to their target values ​​as possible.

[0042] Optimizing any one or more of the lithium mica transformation rate, energy intensity, and carbon emission intensity can improve the performance of the entire production system as a whole. Using optimization algorithms to consider these goals can also identify bottlenecks and room for improvement in the production system, thereby achieving more refined control of process parameters.

[0043] The embodiment of the present application obtains multiple sample data, each of which includes the target parameter value corresponding to the decision parameter variable in the lithium mica roasting process, and the index data corresponding to the preset index, and the preset index includes at least one of the following: lithium mica transformation rate, energy consumption intensity and carbon emission intensity; the mapping function between the decision parameter variable and the preset index is constructed according to the multiple sample data; the index target corresponding to the preset index is used as the optimization target, and the decision parameter variable is optimized based on the preset optimization algorithm and the mapping function to obtain the target parameter set corresponding to the decision parameter variable. By constructing the mapping function between the decision parameter variable and the preset index, the influence of different decision parameters on the index result can be quantified and clearly described, and the preset index (such as lithium mica transformation rate, energy consumption intensity, carbon emission intensity) is used as the optimization target, and the decision parameter is adjusted through the optimization algorithm, and multiple indicators can be considered at the same time to achieve comprehensive optimization, so as to better guide the actual production, help maintain the stability and controllability of the production process, reduce the fluctuation caused by improper parameter adjustment, effectively improve the production efficiency, and improve the intelligence and automation level of the lithium mica roasting process.

[0044] Figure 3The following is a flow chart of a method for determining a decision parameter variable provided by an exemplary embodiment of the present application. Figure 3 , with respect to the determination of the above decision parameter variables, the above method further includes S31-S33: S31, acquiring multiple groups of historical operation data within a preset time period, wherein each group of the multiple groups of historical operation data includes multiple parameter values ​​corresponding to multiple parameter variables in the above-mentioned lithium mica roasting process.

[0045] In some embodiments, the above-mentioned multiple groups of historical operation data may be actual data of continuous operation of the lepidolite rotary kiln roasting, and these historical operation data may include relevant key parameters.

[0046] Among them, the above-mentioned multiple parameter variables include at least one of the following: lithium oxide content in the mixture, water content in the mixture, calcium content in the mixture, potassium content in the mixture, kiln tail temperature, natural gas flow, kiln head wind pressure, and oxygen concentration.

[0047] In the actual operation of the rotary kiln, the equipment may be suspended or adjusted due to maintenance, overhaul, failure, etc. In these cases, the collected data may be abnormal, thus generating abnormal data. Optionally, such abnormal data can be eliminated in combination with relevant operation logs. After eliminating abnormal data, data points may be missing. For this, the Lagrange interpolation method can be used to further supplement and improve the missing data, so as to ensure the continuity and accuracy of historical operation data to a certain extent, making the supplemented sample data more complete, and providing a reliable data basis for subsequent data analysis and model establishment.

[0048] Among them, the operation log is a log file that records information such as equipment operating status, operator behavior, equipment maintenance and repair, etc.

[0049] In the embodiment of the present application, the above parameter variables cover the material composition, heat supply, combustion conditions and reaction environment in the roasting process, providing a comprehensive description of the entire roasting process. A more comprehensive selection of control variables can better reflect the status of each link in the process, help to more accurately adjust the production process, thereby improving the utilization efficiency of resources.

[0050] S32. Calculate the index data corresponding to each group of historical operation data based on the preset index.

[0051] In some embodiments, when the above-mentioned preset indicators include the transformation rate of lithium mica, the ratio of the content of soluble lithium in the clinker obtained after the lithium mica is roasted to the content of soluble lithium in the mixture before the lithium mica is roasted can be determined as the operating value (%) of the transformation rate of lithium mica.

[0052] In some embodiments, when the above-mentioned preset indicators include energy consumption intensity, the ratio of natural gas consumption during the lithium mica roasting process to the lithium mica mixture processing volume can be determined as the operating value of energy consumption intensity, in tons / ton (t / t).

[0053] In some embodiments, when the above-mentioned preset indicators include carbon emission intensity, the emission factor method in the relevant technology is used to determine the operating value of the carbon emission intensity, and the unit is ton / ton (t / t).

[0054] S33. Screen the plurality of parameter variables according to the plurality of groups of historical operation data and the indicator data corresponding to each group of historical operation data to obtain decision parameter variables and target parameter values ​​corresponding to the decision parameter variables.

[0055] In some embodiments, by analyzing multiple groups of historical operation data and their corresponding index data, key parameter variables that have a significant impact on the optimization objectives (such as maximizing the lithium mica transformation rate, minimizing energy consumption, etc.) can be screened out. These parameter variables are decision parameter variables, which reflect important factors in the production process, thereby effectively simplifying the complexity of the model and making subsequent analysis more targeted.

[0056] In an embodiment of the present application, by screening the obtained decision parameter variables, the input dimension in the model can be reduced from the original multiple parameter variables to fewer decision parameter variables, which not only reduces the complexity of the model but also improves the calculation efficiency.

[0057] In some embodiments, in S33, the above-mentioned multiple parameter variables are screened according to the above-mentioned multiple groups of historical operation data and the indicator data corresponding to the above-mentioned each group of historical operation data to obtain the decision parameter variables and the target parameter values ​​corresponding to the above-mentioned decision parameter variables, including S331-S333: S331. For each parameter variable among the plurality of parameter variables, determine the correlation between the parameter variable and each indicator data.

[0058] In some embodiments, in S331, for each parameter variable in the above-mentioned multiple parameter variables, the correlation between the above-mentioned parameter variable and each indicator data is determined, including: for each parameter variable in the above-mentioned multiple parameter variables, the correlation between the parameter variable and each indicator data is calculated by a grey correlation method.

[0059] Among them, the grey correlation method is an analytical method used to measure the strength of the relationship between variables. In this application, the grey correlation coefficients between different process parameters and lithium mica transformation rate, energy consumption intensity and carbon emission intensity are calculated to determine the correlation between them.

[0060] In some embodiments, in S331, for each parameter variable in the plurality of parameter variables, determining the correlation between the parameter variable and each indicator data includes S3311-S3316: S3311. Determine a feature sequence based on indicator data.

[0061] Among them, the characteristic sequence is a sequence established based on indicator data (such as lithium mica transformation rate, energy consumption intensity, carbon emission intensity, etc.).

[0062] S3312. Determine multiple comparison sequences according to the multiple parameter variables, wherein the comparison sequences correspond to the parameter variables one by one.

[0063] S3313. Perform dimensionless processing on the above-mentioned indicator data and the above-mentioned multiple parameter variables.

[0064] Optionally, dimensionless processing can be performed through range standardization, mean standardization, etc.

[0065] S3314. Calculate the absolute difference between the feature sequence and each comparison sequence at each time point.

[0066] The purpose of calculating the absolute difference is to quantify the degree of deviation between the feature sequence (i.e., indicator data) and each comparison sequence (i.e., parameter variable) at each time point.

[0067] S3315. Calculate the correlation coefficient of the parameter variable relative to the indicator data using a preset correlation coefficient formula; optionally, the preset correlation coefficient formula can be expressed as:

[0068] in, Represents the value of the feature sequence at the kth time point. Represents the value of the comparison sequence (parameter variable) at the kth time point. ρ is the resolution coefficient, which usually takes a value between 0 and 1, such as 0.5, to adjust the impact of the minimum difference on the result.

[0069] By calculating the correlation coefficient, the strength of the correlation between each parameter variable and the preset indicator can be quantitatively measured. The larger the value, the stronger the correlation.

[0070] S3316. Calculate the average value of the grey correlation coefficient.

[0071] Specifically, the correlation coefficient of each parameter variable is averaged at all time points to obtain the grey correlation degree of each parameter variable.

[0072] S332: When the correlation between the parameter variable and each indicator data is greater than a preset correlation threshold, the parameter variable is determined as a decision parameter variable.

[0073] Among them, the preset correlation threshold is a pre-set value used to determine whether the correlation strength between the parameter variable and the preset indicator is significant enough. The preset correlation threshold can be set according to the specific application scenario and data characteristics. For example, the above preset correlation threshold can be 0.5.

[0074] S333. Determine the parameter value corresponding to the above-mentioned decision parameter variable as the target parameter value corresponding to the above-mentioned decision parameter variable.

[0075] For example, when the preset indicators include lithium mica transformation rate, energy consumption intensity and carbon emission intensity, the decision parameter variables and the corresponding grey correlation degree can be obtained as shown in Table 1: Table 1 Decision parameter variables and corresponding grey relational degrees

[0076] As shown in Table 1 above, the decision parameter variables screened include: lithium oxide content of the mixture, moisture content of the mixture, calcium content of the mixture, kiln tail temperature, natural gas flow, kiln head wind pressure and oxygen concentration. The correlation between the above decision parameter variables and each indicator data corresponding to the preset indicators is greater than the preset correlation threshold of 0.5.

[0077] In the embodiment of the present application, by calculating the grey correlation degree, it is possible to more clearly understand the correlation between the parameter variables and the indicator data, and then identify the decision parameter variables that have a more significant impact on the target indicators, thereby focusing the optimization and control on these decision parameter variables, effectively improving the efficiency and effect of the optimization.

[0078] In some embodiments, in S22, the mapping function between the decision parameter variable and the preset index is constructed according to the plurality of sample data, including S221-S223: S221. Using the above decision parameter variables as input variables and the above preset indicators as output variables, an initial neural network model is constructed.

[0079] In some embodiments, the initial neural network model can be used to describe the relationship between input variables and output variables, that is, to understand and predict how each decision parameter variable affects a preset indicator.

[0080] S222. Train the initial neural network model based on the multiple sample data to obtain a trained target neural network model.

[0081] In S222, multiple sample data can be used as training data and input into the initial neural network model. The model is trained by adjusting the parameters (such as weights and biases) inside the neural network, and finally a trained target neural network model is obtained.

[0082] For example, suppose there are 500 sets of sample data, each set of data includes multiple parameter values ​​corresponding to multiple parameter variables in the actual lithium mica roasting process. These data are input into the initial neural network model, and the model parameters are continuously adjusted until the model can better predict these output results to obtain the target neural network model.

[0083] Figure 4 A schematic diagram of the structure of a target neural network model provided by an exemplary embodiment of the present application is shown in FIG. Figure 4 , the target neural network model includes input layer, hidden layer and output layer. Among them, the input layer is used to input decision parameter variables X1-X i ,The output layer is used to output the mapping indicator data corresponding to the preset indicators, such as lithium mica transformation rate Y1, energy consumption intensity Y2 and carbon emission intensity Y3.

[0084] S223, determining the target neural network model as a mapping function between the decision parameter variables and the preset indicators.

[0085] The mapping function refers to a mathematical function that describes the relationship between input variables and output variables. The trained neural network model is determined as the final mapping function, which is used to describe the complex relationship between various decision parameter variables (such as lithium oxide content in the mixture, water content in the above mixture, calcium content in the above mixture, potassium content in the above mixture, kiln tail temperature, natural gas flow, kiln head wind pressure, oxygen concentration, etc.) and preset indicators (such as lithium mica transformation rate, energy consumption intensity, carbon emission intensity).

[0086] In some embodiments, the initial neural network model can be a particle swarm optimization back propagation neural network processor (PSOBP-NNP) or a back propagation neural network processor (BP-NNP). Compared with the BP-NNP model, the PSOBP-NNP model has better performance and more obvious improvement in the prediction error of lithium mica transformation rate, energy consumption intensity and carbon emission intensity.

[0087] Specifically, the BP-NNP model and the PSOBP-NNP model can be constructed respectively with the decision parameter variables as input variables and the above-mentioned preset indicators as output variables, and the preset indicators (lepidolite transformation rate, energy consumption intensity and carbon emission intensity, etc.) of each sample data can be predicted by the BP-NNP model and the PSOBP-NNP model respectively, so as to obtain the first prediction indicator data output by the BP-NNP model and the second prediction indicator data output by the PSOBP-NNP model.

[0088] The following example is taken as an example of the case where the preset indicator includes the lithium mica transformation rate. For each sample data, the first prediction indicator data includes the first predicted lithium mica transformation rate (output by the BP-NNP model), and the second prediction indicator data includes the second predicted lithium mica transformation rate (output by the PSOBP-NNP model). Then, the first predicted lithium mica transformation rate, the second predicted lithium mica transformation rate and the actual value of the lithium mica transformation rate included in the sample data (i.e., the operating condition value) can be compared and analyzed. The comparison results can be seen in Table 2: Table 2 Comparison results of predicted lithium mica transformation rate and operating conditions

[0089] Further, the mean square error (MSE), root mean square error (RMSE) and mean absolute percentage error (MAPE) between the first predicted lithium mica transformation rate and the operating value of the lithium mica transformation rate corresponding to each sample data in sample data 1-10 can be calculated, and the mean square error, root mean square error and mean absolute percentage error between the second predicted lithium mica transformation rate and the operating value of the lithium mica transformation rate of each sample data can be calculated, and compared and analyzed. The comparison results of the lithium mica transformation rate error can be seen in Table 3: Table 3 Comparison results of the error in transformation rate of lithium mica

[0090] It can be seen that the error evaluation index data corresponding to PSOBP-NNP are smaller than those corresponding to BP-NNP, that is, PSOBP-NNP is better than BP-NNP in the prediction performance of lithium mica transformation rate.

[0091] The following example is taken as an example to illustrate the case where the preset index includes energy consumption intensity. For each sample data, the first prediction index data includes the first predicted energy consumption intensity (output by the BP-NNP model), and the second prediction index data includes the second predicted energy consumption intensity (output by the PSOBP-NNP model), and then the first predicted energy consumption intensity, the second predicted energy consumption intensity and the actual value (i.e., the operating condition value) of the energy consumption intensity included in the sample data can be compared and analyzed. The comparison results can be seen in Table 4.

[0092] Table 4 Comparison results of predicted energy consumption intensity and operating condition values

[0093] Furthermore, the mean square error, root mean square error and average absolute percentage error between the first predicted energy consumption intensity of each sample data and the operating condition value of the energy consumption intensity can be calculated, and the mean square error, root mean square error and average absolute percentage error between the second predicted energy consumption intensity of each sample data and the operating condition value of the energy consumption intensity can be calculated, and compared and analyzed. The energy consumption intensity error comparison results can be seen in Table 5.

[0094] Table 5 Comparison results of energy consumption intensity error

[0095] It can be seen that the error evaluation index data corresponding to PSOBP-NNP are smaller than those corresponding to BP-NNP, that is, PSOBP-NNP is better than BP-NNP in the prediction performance of energy consumption intensity.

[0096] The following example is taken as an example of the case where the preset index includes carbon emission intensity. For each sample data, the first prediction index data includes the first predicted carbon emission intensity (output by the BP-NNP model), and the second prediction index data includes the second predicted carbon emission intensity (output by the PSOBP-NNP model), and then the first predicted carbon emission intensity, the second predicted carbon emission intensity and the actual value of the carbon emission intensity included in the sample data (i.e., the operating condition value) can be compared and analyzed. The comparison results can be seen in Table 6.

[0097] Table 6 Comparison results between predicted carbon emission intensity and operating condition values

[0098] Furthermore, the mean square error, root mean square error and average absolute percentage error between the first predicted carbon emission intensity of each sample data and the operating value of the carbon emission intensity can be calculated, and the mean square error, root mean square error and average absolute percentage error between the second predicted carbon emission intensity of each sample data and the operating value of the energy consumption intensity can be calculated, and compared and analyzed. The energy consumption intensity error comparison results can be seen in Table 7: Table 7 Comparison results of carbon emission intensity errors

[0099] It can be seen that the mean square error corresponding to PSOBP-NNP is smaller than that corresponding to BP-NNP, the root mean square error corresponding to PSOBP-NNP is smaller than that corresponding to BP-NNP, and only the mean absolute percentage error corresponding to BP-NNP is smaller than that corresponding to PSOBP-NNP.

[0100] In summary, the overall prediction accuracy and stability of the PSOBP-NNP model are higher. It optimizes the weights and thresholds of the neural network through the particle swarm optimization algorithm, reduces the possibility of local minima, and improves the global search capability, thus making the error smaller and the prediction effect more stable.

[0101] In some embodiments, the preset optimization algorithm is a nondominated sorting genetic algorithm II (NSGA-II). In S23, the indicator target corresponding to the preset indicator is used as the optimization target, and the decision parameter variable is optimized based on the preset optimization algorithm and the mapping function to obtain the target parameter set corresponding to the decision parameter variable, including S231-S235: S231. Determine a fitness function according to the above mapping function.

[0102] In some embodiments, the above mapping function can be used as a fitness function. When the preset indicators include lithium mica transformation rate, energy consumption intensity and carbon emission intensity, the fitness function can be expressed as:

[0103] in, , , are mapping functions corresponding to the lithium mica transformation rate, energy consumption intensity and carbon emission intensity respectively; They correspond to the decision parameter variables in Table 1 respectively. , and Respectively represent the optimization targets of lithium mica transformation rate, energy consumption intensity and carbon emission intensity. , and The three objective values ​​are used as the fitness values ​​of the solution to guide the evolution of the NSGA-II algorithm.

[0104] S232: Determine an initial population, where the initial population includes multiple candidate solutions.

[0105] The initial population may refer to a set of candidate solutions generated in the first step of the optimization process. These candidate solutions (i.e., individuals) constitute the population at the beginning of the algorithm. Specifically, the initial population in the embodiment of the present application includes the initial values ​​of the target parameter set corresponding to the decision parameter variables.

[0106] In some embodiments, when generating the initial population, a random generation method may be used to assign a random value within its feasible range to each decision parameter variable.

[0107] In some embodiments, when generating an initial population, a feasibility check may be performed on candidate solutions in the initial population based on preset constraints. If the generated candidate solutions do not meet the constraints, corrections may be made to ensure that the candidate solutions in the initial population are as feasible as possible.

[0108] In some embodiments, the constraints may include material balance conditions and / or heat balance conditions.

[0109] Alternatively, the material balance constraint can be expressed as:

[0110] in, Indicates the material input of the rotary kiln, including the amount of lithium mica mixed material added, natural gas consumption, and air consumption. It is used for the production and discharge of rotary kiln, including dust collection, kiln slag and flue gas.

[0111] Alternatively, the thermal equilibrium condition can be expressed as:

[0112] in, The heat input of the rotary kiln includes the heat of fuel combustion, the physical sensible heat of air blast, the physical sensible heat of furnace charge and the heat released by chemical reaction; It is the heat expenditure of rotary kiln, including heat absorbed by chemical reaction, heat dissipated by kiln surface, heat brought out by kiln slag, heat brought out by smoke and dust, heat brought out by flue gas, heat of water evaporation, heat of incomplete combustion and other heat losses.

[0113] S233: Determine the dominance relationship between the candidate solutions in the initial population based on preset rules.

[0114] The preset rules include at least one of the following rules: Rule (1): If solution i is a feasible solution and solution j is an infeasible solution, then solution i is better than solution j.

[0115] Rule (2): If solution i and solution j are both infeasible solutions, but the constraint violation degree of solution i is smaller, then solution i is better than solution j.

[0116] Rule (3): If solution i and solution j are both feasible solutions, and solution i dominates solution j, then solution i is better than solution j.

[0117] Among them, i and j represent two candidate solutions (i.e., individuals) in the initial population.

[0118] S234: According to the above-mentioned dominance relationship, the above-mentioned candidate solutions are selected, crossed and mutated to generate new candidate solutions.

[0119] In some embodiments, in S234, according to the above-mentioned dominance relationship, each candidate solution in the above-mentioned initial population is selected, crossed and mutated to generate a new candidate solution, including: S2341. Perform non-dominated sorting on the candidate solutions in the initial population to determine the dominance level corresponding to each candidate solution.

[0120] The dominance level indicates the non-dominated level of each candidate solution in the population, reflecting the quality of the solution in multi-objective optimization.

[0121] S2342. For candidate solutions of the same non-dominated level, calculate the congestion distance of each candidate solution.

[0122] Among them, the crowding distance is used to measure the density of solutions around the candidate solution, and can be used to sort the solutions within the same non-dominated level to give priority to those solutions that are sparsely distributed in the solution space. Specifically, the crowding distance can be calculated using the relevant formula in the multi-objective genetic algorithm.

[0123] S2343. According to the crowding distance, select at least part of the candidate solutions from the initial population to generate a parent population.

[0124] Optionally, the candidate solutions can be sorted according to the crowding distance, and solutions with larger crowding distance are preferentially selected to enter the parent population to retain solutions with sparse distribution in the solution space, thereby maintaining the diversity and coverage of the population.

[0125] S2344. Perform crossover and mutation processing on each candidate solution in the above parent population to generate a new candidate solution.

[0126] In some embodiments, in S2344, when the candidate solutions in the above-mentioned parent population are cross-processed, a new solution can be generated from the selected candidate solutions by combining their features. For example, the kiln tail temperature of one candidate solution and the natural gas flow rate of another candidate solution are combined to obtain a new candidate solution. Specifically, it can be achieved by single-point cross-over, multi-point cross-over, uniform cross-over, etc.

[0127] In some embodiments, in S2344, when mutating each candidate solution in the above-mentioned parent population, it can be achieved by performing small-amplitude random changes, Gaussian mutations, non-uniform mutations, etc. on some candidate solutions.

[0128] S235. Determine the mapping index data corresponding to the parameter value based on each of the new candidate solutions and the fitness function until the preset conditions are met, and output the optimal solution of the decision parameter variable as the target parameter set corresponding to the decision parameter variable, wherein the preset conditions include the number of iterations reaching the preset number of iterations or the mapping index data meeting the preset threshold conditions.

[0129] Among them, the target parameter set corresponding to the above decision parameter variables can be a Pareto optimal process parameter solution set, and the target parameter set can be one or more. Taking 10 groups of target parameter sets as an example, the specific target parameter sets can be referred to Table 8: Table 8 Pareto optimal process parameter solutions for 10 groups of lithium mica rotary kiln roasting

[0130] Table 8 above shows 10 sets of Pareto optimal solutions (i.e., target parameter sets) obtained through multi-objective optimization in the process of lithium mica rotary kiln roasting, corresponding to different process parameter combinations. These target parameter sets represent the optimal parameter configurations obtained under multi-objective optimization (such as maximizing the transformation rate, minimizing the energy intensity and carbon emission intensity), which are used to guide the optimization of the process.

[0131] Figure 5 A schematic diagram of the effect of multi-objective optimization provided by an exemplary embodiment of the present application is shown in FIG. Figure 5 The three-dimensional coordinate axes in the figure correspond to three different preset indicators: the x-axis represents the lithium mica transformation rate, the y-axis represents the energy intensity, and the z-axis represents the carbon emission intensity. The points in the marked areas A, B, and C in the figure represent different target parameter sets (or candidate solutions). Among them, area A represents a target parameter set with low energy intensity and carbon emission intensity, but a low lithium mica transformation rate. Area B represents a target parameter set that is more balanced among the three, such as a relatively compromised target parameter set. Area C represents a target parameter set with a higher lithium mica transformation rate, but the corresponding energy intensity and carbon emission intensity may be higher. This three-dimensional graph shows the Pareto optimal solution set found in the multi-objective optimization process, in which the solutions in different areas perform differently among the three objectives. Through this multi-objective optimization, multiple options can be provided to decision makers to balance the needs of increasing the transformation rate, reducing energy consumption, and reducing carbon emissions.

[0132] In the embodiments of the present application, the quality of candidate solutions is judged by preset rules, especially by screening the solution set through the dominating relationship, so that the algorithm can give priority to retaining solutions that perform better in multi-objective optimization, and continuously generate new solutions through crossover and mutation operations, thereby improving the diversity and exploratory nature of solutions and avoiding premature convergence to local optimal solutions; through multiple iterations of the optimization process, based on the mapping index data of each candidate solution, a set of solutions that meet the preset conditions is finally obtained. This multi-objective optimization method can find a Pareto optimal solution that balances different objectives, and provide relevant personnel with a variety of feasible solutions, so as to facilitate the selection of appropriate parameter sets according to different priorities in practical applications, and can achieve sustainable development of production while optimizing the lithium mica transformation rate, which meets the environmental protection requirements of current industrial development.

[0133] Please refer to the following Figure 6 , which is a schematic diagram of the structure of a lithium mica roasting process optimization device provided by an exemplary embodiment of the present application. Figure 6 As shown, the above-mentioned lithium mica roasting process optimization device 600 includes: A first acquisition module 601 is used to acquire a plurality of sample data, each of which includes a target parameter value corresponding to a decision parameter variable in a lithium mica roasting process, and indicator data corresponding to a preset indicator, wherein the preset indicator includes at least one of the following: lithium mica transformation rate, energy consumption intensity, and carbon emission intensity; A construction module 602 is used to construct a mapping function between the decision parameter variables and the preset indicators according to the plurality of sample data; The optimization module 603 is used to optimize the above-mentioned decision parameter variables based on the preset optimization algorithm and the above-mentioned mapping function with the indicator target corresponding to the above-mentioned preset indicator as the optimization target, so as to obtain the target parameter set corresponding to the above-mentioned decision parameter variables.

[0134] In a possible implementation, the apparatus 600 further includes: A second acquisition module is used to acquire multiple groups of historical operation data within a preset time period, each group of the multiple groups of historical operation data including multiple parameter values ​​corresponding to multiple parameter variables in the above-mentioned lithium mica roasting process; A calculation module, used to calculate the indicator data corresponding to each group of historical operation data based on preset indicators; The screening module is used to screen the above-mentioned multiple parameter variables according to the above-mentioned multiple groups of historical operation data and the indicator data corresponding to the above-mentioned groups of historical operation data, so as to obtain decision parameter variables and target parameter values ​​corresponding to the above-mentioned decision parameter variables.

[0135] In a possible implementation, the above-mentioned multiple parameter variables include at least one of the following: lithium oxide content in the mixture, water content in the mixture, calcium content in the mixture, potassium content in the mixture, kiln tail temperature, natural gas flow, kiln head wind pressure, and oxygen concentration.

[0136] In a possible implementation, the screening module includes: A first determining unit, configured to determine, for each parameter variable among the plurality of parameter variables, a correlation between the parameter variable and each indicator data; A second determining unit is used to determine the parameter variable as a decision parameter variable when the correlation between the parameter variable and each indicator data is greater than a preset correlation threshold; The third determining unit is used to determine the parameter value corresponding to the above-mentioned decision parameter variable as the target parameter value corresponding to the above-mentioned decision parameter variable.

[0137] In a possible implementation, the building module 602 includes: A construction unit, used to construct an initial neural network model using the above decision parameter variables as input variables and the above preset indicators as output variables; A training unit, used to train the initial neural network model based on the plurality of sample data to obtain a trained target neural network model; The fourth determination unit is used to determine the above-mentioned target neural network model as a mapping function between the above-mentioned decision parameter variables and the above-mentioned preset indicators.

[0138] In a possible implementation, the apparatus 600 further includes: A first determination module is used for maximizing the lepidolite transformation rate as an optimization target corresponding to the lepidolite transformation rate when the preset index includes the lepidolite transformation rate; A second determination module is used for, when the preset index includes the energy consumption intensity, minimizing the energy consumption intensity as an optimization target corresponding to the energy consumption intensity; The third determination module is used to, when the preset index includes the carbon emission intensity, minimize the carbon emission intensity as the optimization target corresponding to the energy consumption intensity.

[0139] In a possible implementation, the preset optimization algorithm is a multi-objective genetic algorithm, and the optimization module includes: A fifth determining unit, used to determine a fitness function according to the mapping function; A sixth determining unit, configured to determine an initial population, wherein the initial population includes a plurality of candidate solutions; A seventh determination unit, configured to determine a dominance relationship between candidate solutions in the initial population based on a preset rule; A processing unit, configured to perform selection, crossover and mutation processing on the candidate solutions according to the dominance relationship to generate new candidate solutions; The eighth determination unit is used to determine the mapping indicator data corresponding to the above-mentioned parameter value based on each of the above-mentioned new candidate solutions and the above-mentioned fitness function until the preset conditions are met, and output the optimal solution of the above-mentioned decision parameter variables as the target parameter set corresponding to the above-mentioned decision parameter variables, wherein the above-mentioned preset conditions include the number of iterations reaching the preset number of iterations or the above-mentioned mapping indicator data meeting the preset threshold conditions.

[0140] The division of each module in the above-mentioned lithium mica roasting process optimization device 600 is only for illustration. In other embodiments, the lithium mica roasting process optimization device can be divided into different modules as needed to complete all or part of the functions of the above-mentioned lithium mica roasting process optimization device. The implementation of each module in the lithium mica roasting process optimization device provided in the embodiment of this specification can be in the form of a computer program. The computer program can be run on a terminal or a server. The program modules constituted by the computer program can be stored in the memory of the terminal or the server. When the computer program is executed by the processor, all or part of the steps of the lithium mica roasting process optimization method described in the embodiment of this specification are implemented.

[0141] See next Figure 7 , which is a schematic diagram of the structure of an electronic device provided by an exemplary embodiment of the present application. Figure 7 As shown, the electronic device 700 may include: a processor 710 and a memory 720 , and may also include a user interface 730 , a network interface 740 and a communication bus 750 .

[0142] Among them, the processor 710 may include one or more processing cores. The processor 710 uses various interfaces and lines to connect various parts of the entire electronic device 700, and executes various functions and processes data of the electronic device 700 by running or executing instructions, programs, code sets or instruction sets stored in the memory 720, and calling data stored in the memory 720. Optionally, the processor 710 can be implemented in at least one hardware form of digital signal processing (Digital Signal Processing, DSP), field programmable gate array (Field-Programmable Gate Array, FPGA), and programmable logic array (Programmable Logic Array, PLA). The processor 710 can integrate one or a combination of a central processing unit (Central Processing Unit, CPU), a graphics processing unit (Graphics Processing Unit, GPU) and a modem. Among them, the CPU mainly processes the operating system and application programs; the GPU is responsible for rendering and drawing the content to be displayed on the display screen; the modem is used to process wireless communications. It can be understood that the above-mentioned modem may not be integrated into the processor 710, and can be implemented separately through a chip.

[0143] Among them, the memory 720 may include a random access memory (Random Access Memory, RAM) and may also include a read-only memory (Read-Only Memory). Optionally, the memory 720 includes a non-transitory computer-readable storage medium. The memory 720 can be used to store instructions, programs, codes, code sets or instruction sets. The memory 720 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function (such as a receiving function, a control function, etc.), instructions for implementing the above-mentioned various method embodiments, etc.; the data storage area may store data involved in the above-mentioned various method embodiments, etc. The memory 720 may also be optionally at least one storage device located away from the aforementioned processor 710. As Figure 7 As shown, the memory 720 as a computer storage medium may include an operating system, a network communication module, a user interface module, and program instructions.

[0144] Optionally, the communication bus 750 is used to realize the connection and communication between these components. The user interface 730 may include a display screen (Display), a camera (Camera), and may also include a standard wired interface and a wireless interface; the network interface 740 may optionally include a standard wired interface and a wireless interface (such as a WI-FI interface).

[0145] exist Figure 7 In the electronic device 700 shown, the processor 710 may be used to call program instructions stored in the memory 720 and specifically perform the following operations: Acquire multiple sample data, each of the multiple sample data includes a target parameter value corresponding to a decision parameter variable in a lithium mica roasting process, and indicator data corresponding to a preset indicator, wherein the preset indicator includes at least one of the following: lithium mica transformation rate, energy consumption intensity, and carbon emission intensity; Constructing a mapping function between the decision parameter variables and the preset indicators according to the plurality of sample data; Taking the indicator target corresponding to the above preset indicator as the optimization target, the above decision parameter variables are optimized based on the preset optimization algorithm and the above mapping function to obtain the target parameter set corresponding to the above decision parameter variables.

[0146] In a possible implementation, the method further includes: Acquire multiple groups of historical operation data within a preset time period, each group of the multiple groups of historical operation data including multiple parameter values ​​corresponding to multiple parameter variables in the calcination process of the lithium mica; Calculate the indicator data corresponding to each group of historical operation data based on the preset indicators; According to the above multiple groups of historical operation data and the indicator data corresponding to each group of historical operation data, the above multiple parameter variables are screened to obtain decision parameter variables and target parameter values ​​corresponding to the above decision parameter variables.

[0147] In a possible implementation, the above-mentioned multiple parameter variables include at least one of the following: lithium oxide content in the mixture, water content in the mixture, calcium content in the mixture, potassium content in the mixture, kiln tail temperature, natural gas flow, kiln head wind pressure, and oxygen concentration.

[0148] In a possible implementation, the above-mentioned multiple parameter variables are screened according to the above-mentioned multiple groups of historical operation data and the indicator data corresponding to the above-mentioned each group of historical operation data to obtain the decision parameter variables and the target parameter values ​​corresponding to the above-mentioned decision parameter variables, including: For each parameter variable among the plurality of parameter variables, determining a correlation degree between the parameter variable and each indicator data; When the correlation between the parameter variable and each indicator data is greater than a preset correlation threshold, the parameter variable is determined as a decision parameter variable; The parameter value corresponding to the above decision parameter variable is determined as the target parameter value corresponding to the above decision parameter variable.

[0149] In a possible implementation, the step of constructing a mapping function between the decision parameter variable and the preset index according to the plurality of sample data includes: Using the above decision parameter variables as input variables and the above preset indicators as output variables, an initial neural network model is constructed; Training the initial neural network model based on the multiple sample data to obtain a trained target neural network model; The target neural network model is determined as a mapping function between the decision parameter variables and the preset indicators.

[0150] In a possible implementation, the method further includes: In the case where the preset index includes the lepidolite transformation rate, maximizing the lepidolite transformation rate is used as the optimization target corresponding to the lepidolite transformation rate; In the case where the above-mentioned preset indicators include the above-mentioned energy consumption intensity, minimizing the above-mentioned energy consumption intensity is taken as the optimization target corresponding to the above-mentioned energy consumption intensity; In the case where the above-mentioned preset indicators include the above-mentioned carbon emission intensity, minimizing the above-mentioned carbon emission intensity is taken as the optimization target corresponding to the above-mentioned energy consumption intensity.

[0151] In a possible implementation, the preset optimization algorithm is a multi-objective genetic algorithm, the indicator target corresponding to the preset indicator is used as the optimization target, the decision parameter variables are optimized based on the preset optimization algorithm and the mapping function, and the target parameter set corresponding to the decision parameter variables is obtained, including: Determine the fitness function according to the above mapping function; Determine an initial population, wherein the initial population includes a plurality of candidate solutions; Determine the dominance relationship between the candidate solutions in the initial population based on preset rules; According to the above dominance relationship, the above candidate solutions are selected, crossed and mutated to generate new candidate solutions; Based on each of the above-mentioned new candidate solutions and the above-mentioned fitness function, the mapping indicator data corresponding to the above-mentioned parameter value is determined until the preset conditions are met, and the optimal solution of the above-mentioned decision parameter variables is output as the target parameter set corresponding to the above-mentioned decision parameter variables, wherein the above-mentioned preset conditions include that the number of iterations reaches the preset number of iterations or the above-mentioned mapping indicator data meets the preset threshold conditions.

[0152] The present application also provides a computer-readable storage medium, which stores instructions, and when the instructions are executed on a computer or a processor, the computer or the processor executes one or more steps in the above embodiment. If the components of the above-mentioned lithium mica roasting process optimization device are implemented in the form of software functional units and sold or used as independent products, they can be stored in the above-mentioned computer-readable storage medium.

[0153] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The above computer program product includes one or more computer instructions. When the above computer program instructions are loaded and executed on a computer, the above process or function according to the embodiment of the present application is generated in whole or in part. The above computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The above computer instructions may be stored in a computer-readable storage medium or transmitted through the above computer-readable storage medium. The above computer instructions may be transmitted from a website site, computer, server or data center to another website site, computer, server or data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (Digital Subscriber Line, DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The above computer-readable storage medium may be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more available media integrated. The above-mentioned available media can be magnetic media (for example, floppy disks, hard disks, tapes), optical media (for example, digital versatile discs (DVD)), or semiconductor media (for example, solid state drives (SSD)), etc.

[0154] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program, and the program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. The aforementioned storage medium includes: ROM, RAM, magnetic disk or optical disk and other media that can store program codes. In the absence of conflict, the technical features in this embodiment and the implementation scheme can be combined arbitrarily.

[0155] The above-mentioned embodiments are merely preferred embodiments of the present application and are not intended to limit the scope of the present application. Without departing from the design spirit of the present application, various modifications and improvements made to the technical solutions of the present application by ordinary technicians in this field should fall within the protection scope determined by the claims of the present application.

Claims

1. A method for optimizing the calcination process of lithium mica, characterized in that: include: Acquire a plurality of sample data, each of the plurality of sample data comprising a target parameter value corresponding to a decision parameter variable in a calcination process of lithium mica, and indicator data corresponding to a preset indicator, wherein the preset indicator comprises at least one of the following: lithium mica transformation rate, energy consumption intensity, and carbon emission intensity; Constructing a mapping function between the decision parameter variable and the preset index according to the plurality of sample data; Taking the indicator target corresponding to the preset indicator as the optimization target, the decision parameter variable is optimized based on the preset optimization algorithm and the mapping function to obtain the target parameter set corresponding to the decision parameter variable.

2. The method according to claim 1, characterized in that The method further comprises: Acquire multiple groups of historical operation data within a preset time period, each group of the multiple groups of historical operation data including multiple parameter values ​​corresponding to multiple parameter variables in the lepidolite roasting process; Calculate the index data corresponding to each group of historical operation data based on the preset index; According to the multiple groups of historical operation data and the indicator data corresponding to each group of historical operation data, the multiple parameter variables are screened to obtain decision parameter variables and target parameter values ​​corresponding to the decision parameter variables.

3. The method according to claim 2, characterized in that The multiple parameter variables include at least one of the following: lithium oxide content in the mixture, water content in the mixture, calcium content in the mixture, potassium content in the mixture, kiln tail temperature, natural gas flow, kiln head wind pressure, and oxygen concentration.

4. The method according to claim 2, characterized in that The step of screening the plurality of parameter variables according to the plurality of groups of historical operation data and the indicator data corresponding to each group of historical operation data to obtain decision parameter variables and target parameter values ​​corresponding to the decision parameter variables includes: For each parameter variable among the multiple parameter variables, determining a correlation degree between the parameter variable and each indicator data; In the case where the correlation between the parameter variable and each indicator data is greater than a preset correlation threshold, determining the parameter variable as a decision parameter variable; The parameter value corresponding to the decision parameter variable is determined as the target parameter value corresponding to the decision parameter variable.

5. The method according to claim 1, characterized in that The step of constructing a mapping function between the decision parameter variable and the preset index according to the plurality of sample data includes: Taking the decision parameter variables as input variables and the preset indicators as output variables, an initial neural network model is constructed; Training the initial neural network model based on the multiple sample data to obtain a trained target neural network model; The target neural network model is determined as a mapping function between the decision parameter variables and the preset indicators.

6. The method according to claim 1, characterized in that The method further comprises: In the case where the preset index includes the lepidolite transformation rate, maximizing the lepidolite transformation rate is used as an optimization target corresponding to the lepidolite transformation rate; In the case where the preset indicator includes the energy consumption intensity, minimizing the energy consumption intensity is used as the optimization target corresponding to the energy consumption intensity; When the preset index includes the carbon emission intensity, minimizing the carbon emission intensity is used as the optimization target corresponding to the energy consumption intensity.

7. The method according to claim 1, characterized in that The preset optimization algorithm is a multi-objective genetic algorithm, and the indicator target corresponding to the preset indicator is used as the optimization target. The decision parameter variable is optimized based on the preset optimization algorithm and the mapping function to obtain the target parameter set corresponding to the decision parameter variable, including: Determine a fitness function according to the mapping function; Determine an initial population, wherein the initial population includes a plurality of candidate solutions; Determining the dominance relationship between candidate solutions in the initial population based on a preset rule; According to the dominance relationship, the candidate solutions are selected, crossed and mutated to generate new candidate solutions; Based on each of the new candidate solutions and the fitness function, the mapping indicator data corresponding to the parameter value is determined until a preset condition is met, and the optimal solution of the decision parameter variable is output as the target parameter set corresponding to the decision parameter variable, wherein the preset condition includes that the number of iterations reaches a preset number of iterations or the mapping indicator data meets a preset threshold condition.

8. A device for optimizing the calcination process of lithium mica, characterized in that: include: A first acquisition module is used to acquire a plurality of sample data, each of which includes a target parameter value corresponding to a decision parameter variable in a lithium mica roasting process, and indicator data corresponding to a preset indicator, wherein the preset indicator includes at least one of the following: lithium mica transformation rate, energy consumption intensity, and carbon emission intensity; A construction module, used to construct a mapping function between the decision parameter variable and the preset index according to the plurality of sample data; The optimization module is used to optimize the decision parameter variables based on the preset optimization algorithm and the mapping function with the indicator target corresponding to the preset indicator as the optimization target, so as to obtain the target parameter set corresponding to the decision parameter variables.

9. An electronic device, characterized in that: include: Processor and memory; The memory stores a computer program, and the computer program is suitable for being loaded by the processor and executing the steps of the method according to any one of claims 1 to 7.

10. A computer storage medium, characterized in that: The computer storage medium stores a plurality of instructions, and the instructions are suitable for being loaded by a processor and executing the steps of the method according to any one of claims 1 to 7.

Images