Method and system for intelligently predicting production quality of electrolytic manganese process based on AI

Abstract

The invention relates to the technical field of industrial production intelligent process control, in particular to an AI-based electrolytic manganese process production quality intelligent prediction method and system, and the system comprises a drift calibration module, a structure matching module, a process parameter inversion module, a path correction module and a deviation verification module. According to the method, dynamic compensation of deviation between input parameter drift and output parameters is achieved through the response relation between input and output of all links, the actual influence degree of fluctuation of all the parameters on the target control parameters and product quality is effectively recognized according to correlation matching of the periodic change of the production process and the fluctuation frequency of the target control parameters, and the accuracy of the product quality is improved. Potential interference of key process parameters on stability of each process is locked in time, diffusion starting time and space of impurities are clearly traced through path judgment of residual liquid concentration trend and raw material adding position, and migration path parameters are further adjusted based on ion retention behaviors of boundary detection points.

Description

Technical Field

[0001] This invention relates to the field of intelligent process control technology in industrial production, specifically to an AI-based intelligent prediction method and system for the production quality of electrolytic manganese processes. Background Technology

[0002] AI-powered intelligent process control technology encompasses a comprehensive set of intelligent methods for real-time monitoring, data analysis, control decision-making, and automatic adjustment of key process parameters in industrial production. It aims to precisely regulate process variables such as flow rate, temperature, concentration, and current density by collecting process data through sensors, combining historical operating information with real-time status, and utilizing model prediction and feedback control. This enhances the automation level, control precision, and operational stability of industrial processes. Applied to continuous or complex coupled production processes in metallurgy, chemical engineering, power generation, and pharmaceuticals, it provides significant technical support for improving product quality and ensuring production safety. Traditional manganese smelting processes, such as leaching and purification, remain intermittent operations, resulting in low efficiency, high labor intensity, and poor working environments. Furthermore, key process control still relies heavily on human experience, leading to information lag and inefficient control, resulting in large fluctuations in actual control parameters, unstable product quality, and high production costs. Summary of the Invention

[0003] To address the problems existing in the control of current manganese smelting processes, this invention provides an AI-based intelligent prediction method and system for the production quality of electrolytic manganese processes. The technical solution is as follows: On the one hand, an AI-based intelligent prediction system for the production quality of electrolytic manganese processes is provided, which includes: The drift calibration module uses a temperature sensor to analyze the change trajectory between the temperature fluctuation trends of the injection chamber and the ambient temperature, evaluate the correspondence between temperature synchronization deviation and ion response difference, correct the current cycle concentration reading, and generate a concentration calibration value. The structure matching module uses the concentration calibration value to determine the synchronous correlation of the structure period by comparing the period of the lattice spacing change during the deposition process with the time frequency of the concentration fluctuation of each impurity ion, and identifies the influence of ion concentration fluctuation on the formation of a stable crystal configuration, thereby generating impurity response information. The process parameter inversion module calls the impurity response information, analyzes the positional relationship between the residual liquid element concentration trend and the raw material addition behavior, calls the raw material addition time period, current change direction and temperature change region, performs path judgment on the initial concentration change behavior of impurities, and generates the diffusion initiation concentration. The path correction module uses the diffusion initiation concentration to calculate the retention change direction between multiple detection points in the boundary region, determine the migration delay phenomenon of ions, correct the behavior offset parameters of the path segment, adjust the flow rate rhythm in the current path prediction, and generate the path delay coefficient.

[0004] As a further aspect of the present invention, the concentration calibration value includes the direction of temperature difference change, the magnitude of response delay, and the concentration correction ratio; the impurity response information specifically includes the periodic frequency matching factor, the structural correlation strength, and the degree of impurity influence; the diffusion initiation concentration includes the concentration fluctuation range, the initial behavior characteristics, and the direction of diffusion trend; and the path delay coefficient specifically refers to the amount of retention distribution difference, the amount of flow rate rhythm adjustment, and the path offset parameter.

[0005] As a further aspect of the present invention, the drift calibration module includes: The temperature difference detection submodule uses a temperature sensor to acquire data on the fluctuation trends of the sample injection chamber temperature and the ambient temperature, analyzes the trajectory of their changes, calculates the direction of temperature difference change, and establishes the temperature difference change range by combining the amplitude and direction of temperature difference change within each detection cycle. The synchronization determination submodule evaluates the trend of temperature synchronization deviation within the continuous sampling period based on the temperature difference change range, and judges the delayed response behavior of the injection chamber by combining the deviation duration and change direction, analyzes the continuity and consistency of the synchronization deviation, and generates synchronization deviation parameters. The response correction submodule calculates the correspondence between the degree of deviation and the difference in ion response in the concentration curve based on the synchronization deviation parameter, and corrects the current period concentration reading by the ratio between temperature difference change and concentration fluctuation, generating a concentration calibration value.

[0006] As a further aspect of the present invention, the specific formula for determining the delayed response behavior of the injection chamber is as follows: ; Calculate the trend index of synchronization deviation; in, As a trend indicator of synchronization deviation, This represents the normalized value of the injection chamber temperature collected during the i-th period. The normalized value of the ambient temperature collected during the i-th period is... This represents the normalized value of the injection chamber temperature collected during the (i-1)th cycle. The normalized value of the ambient temperature collected during the (i-1)th period. Let be the temperature-normalized stability constant. For symbolic functions, Let be the directional weight corresponding to the direction of temperature change within the i-th period. The total number of sampling periods. This is the index number of the current sampling period. This is the index number of the previous sampling period.

[0007] As a further aspect of the present invention, the structure matching module includes: The period comparison submodule obtains the concentration calibration value, collects the period of lattice spacing change during the deposition process, obtains the time frequency of impurity ion concentration fluctuation, compares the lattice spacing period and the frequency difference of each impurity ion concentration fluctuation, analyzes the consistency of period changes, and generates period matching differences. The synchronization analysis submodule determines the temporal synchronization correlation between the structural period and the fluctuation of impurity ion concentration based on the period matching difference, filters period segments with consistent changing trends, and establishes the correlation between the structural period and the fluctuation of ion concentration by combining the structural period synchronization parameters, thereby generating structural synchronization parameters. The influence identification submodule analyzes the trend of impurity ion concentration fluctuations on crystal deposition configuration changes based on the structural synchronization parameters, identifies the influence relationship of each impurity ion concentration fluctuation on the deposition structure, and generates impurity response information.

[0008] As a further aspect of the present invention, the specific formula for identifying the influence of the concentration fluctuation of each impurity ion on the deposition structure is as follows: ; Calculate the impact intensity value; in, Let be the intensity value of the influence of the k-th impurity ion on the crystal deposition configuration. This is the normalized value of the concentration variation of the k-th impurity ion in the j-th lattice period. Let be the difference between the frequency of the concentration fluctuation of the k-th impurity ion and the frequency of the spacing fluctuation in the j-th lattice period. This is the normalized value of the spacing offset of the structurally stable region caused by the k-th impurity within the j-th lattice period. This is the normalized value of the grain arrangement perturbation value in the lattice distortion region caused by the k-th impurity within the j-th lattice period. The lattice period number, Number the impurity types. This represents the total number of cycles.

[0009] As a further aspect of the present invention, the process parameter inversion module includes: The periodic analysis submodule calls the impurity response information to analyze the element concentration trend in the residual liquid at the end of each period, obtains the time position of each raw material addition, statistically analyzes the concentration change range of impurities within the period span, calculates the concentration fluctuation range within each period, and generates the periodic concentration range. The interval filtering submodule filters concentration segments with consistent fluctuation directions based on the periodic concentration interval, and, in conjunction with the impurity concentration change trend parameter, identifies concentration change segments consistent with periodic fluctuations, extracts fluctuation consistency markers, and generates consistent concentration segments. The path determination submodule calls the consistency concentration range, combines the raw material addition time period, current change direction and temperature change area to determine the initial concentration behavior of impurities in the concentration change range, analyzes the ratio and trend of each parameter, establishes the concentration value corresponding to the starting point of the diffusion path, and generates the diffusion starting point concentration.

[0010] As a further aspect of the present invention, the path correction module includes: The retention determination submodule calls the diffusion initiation concentration, calculates the direction of ion retention time change at each detection point within the boundary region, detects ion retention data at multiple detection points, determines the ion migration delay phenomenon, and generates migration delay parameters based on the retention time trend. The distribution comparison submodule compares the distribution differences of loitering behavior in multiple regions based on the migration delay parameter, analyzes the spatial distribution trend of loitering phenomena, and generates a loitering distribution coefficient. The behavior correction submodule calls the retention distribution coefficient to analyze the retention behavior of the boundary path and the migration trend of the ion continuous sampling segment, corrects the behavior offset parameter of the path segment, adjusts the flow rate rhythm in the current path prediction, and generates the path delay coefficient.

[0011] As a further aspect of the present invention, the system further includes: The deviation verification module calls the path delay coefficient to analyze the degree of deviation between the target feeding flow rate and the actual execution flow rate in the feeding instruction, compares the current flow characteristics of the feeding medium with the standard flow state, adjusts the flow rate control parameter range of the control signal, establishes the feedback relationship between viscosity and execution response offset, and generates feeding compensation configuration. The feeding compensation configuration includes flow rate control parameters, viscosity response relationship, and execution compensation coefficient.

[0012] As a further aspect of the present invention, the deviation verification module includes: The flow rate deviation submodule calls the path delay coefficient to analyze the target feed flow rate and the actual execution flow rate in the feed instruction, calculates the degree of difference, determines the deviation range, and generates the flow rate deviation range by combining the periodic flow rate data. The viscosity analysis submodule, based on the flow rate deviation range, calls the current flow characteristics and standard flow state of the feeding medium to analyze the impact of viscosity changes on the flow rate during the execution phase and generates a viscosity influence coefficient. The signal adjustment submodule calls the viscosity influence coefficient to adjust the flow rate control parameter range of the control signal, establishes a feedback relationship between viscosity and execution response offset, and generates a feeding compensation configuration.

[0013] On the other hand, an AI-based intelligent prediction method for the production quality of electrolytic manganese process is provided. This method is applied to an AI-based intelligent prediction system for the production quality of electrolytic manganese process, and includes: S1: Using a temperature sensor, analyze the change trajectory between the temperature fluctuation trends of the injection chamber and the ambient temperature, evaluate the correspondence between temperature synchronization deviation and ion response difference, correct the current cycle concentration reading, and generate a concentration calibration value. S2: Using the concentration calibration value, by comparing the period of lattice spacing change during the deposition process with the time frequency of each impurity ion concentration fluctuation, the synchronous correlation of the structural period is determined, the influence of ion concentration fluctuation on the formation of a stable crystal configuration is identified, and impurity response information is generated. S3: Call the impurity response information, analyze the positional relationship between the residual liquid element concentration trend and the raw material addition behavior, call the raw material addition time period, current change direction and temperature change area, make path judgment on the impurity initial concentration change behavior, and generate diffusion start concentration; S4: Using the diffusion initiation concentration, by calculating the retention change direction between multiple detection points in the boundary region, the migration delay phenomenon of ions is determined, the behavior offset parameter of the path segment is corrected, the flow rate rhythm in the current path prediction is adjusted, and the path delay coefficient is generated. S5: Call the path delay coefficient to analyze the deviation between the target feed flow rate and the actual execution flow rate in the feed instruction, call the current flow characteristics of the feed medium and the standard flow state for comparison, adjust the flow rate control parameter range of the control signal, establish the feedback relationship between viscosity and execution response offset, and generate feed compensation configuration.

[0014] The beneficial effects of the technical solutions provided in the embodiments of the present invention include at least the following: By utilizing the relationship between temperature fluctuations and concentration response, dynamic compensation is achieved for the deviation between temperature drift and ion concentration readings, reducing signal offset caused by temperature anomalies. Based on the correlation and matching between lattice period changes and impurity fluctuation frequencies, the actual impact of element fluctuations on the crystal structure is effectively identified, and potential interference of impurities on crystal structure stability is promptly identified. By judging the path between residual liquid concentration trends and raw material addition locations, the impurity diffusion starting point is accurately determined, and the diffusion initiation time and space of impurities are clearly traced. Furthermore, based on the ion retention behavior at boundary detection points, migration path parameters are adjusted. Based on actual flow rate and medium viscosity changes, the control parameter range of the feeding process is dynamically optimized, improving the accuracy of element monitoring and concentration control. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a system flowchart of the present invention; Figure 2 This is a schematic diagram of the system framework of the present invention. Detailed Implementation

[0017] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0018] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0019] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning.

[0020] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0021] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0022] This invention provides an AI-based intelligent prediction system for the production quality of electrolytic manganese processes. Please refer to [link / reference]. Figures 1 to 2 This invention provides a technical solution: an AI-based intelligent prediction system for the production quality of electrolytic manganese processes, comprising: The drift calibration module uses a temperature sensor to analyze the change trajectory between the temperature fluctuation trends of the injection chamber and the ambient temperature, evaluate the correspondence between temperature synchronization deviation and ion response difference, correct the current cycle concentration reading, and generate a concentration calibration value. The structure matching module uses concentration calibration values ​​to determine the synchronous correlation of structure period by comparing the period of lattice spacing change during the deposition process with the time frequency of concentration fluctuation of each impurity ion, identify the influence of ion concentration fluctuation on the formation of stable crystal configuration, and generate impurity response information. The process parameter inversion module calls impurity response information, analyzes the positional relationship between residual liquid element concentration trends and raw material addition behavior, calls raw material addition time period, current change direction and temperature change region, performs path judgment on the initial concentration change behavior of impurities, and generates diffusion initiation concentration. The path correction module uses the diffusion initiation concentration to calculate the retention change direction between multiple detection points in the boundary region, determines the migration delay phenomenon of ions, corrects the behavior offset parameters of the path segment, adjusts the flow rate rhythm in the current path prediction, and generates the path delay coefficient. The deviation verification module calls the path delay coefficient to analyze the degree of deviation between the target feeding flow rate and the actual execution flow rate in the feeding instruction. It compares the current flow characteristics of the feeding medium with the standard flow state, adjusts the flow rate control parameter range of the control signal, establishes the feedback relationship between viscosity and execution response offset, and generates the feeding compensation configuration.

[0023] Concentration calibration values ​​include the direction of temperature difference change, response delay amplitude, and concentration correction ratio. Impurity response information specifically includes the periodic frequency matching factor, structural correlation strength, and degree of impurity influence. Diffusion starting concentration includes concentration fluctuation range, initial behavior characteristics, and diffusion trend direction. Path delay coefficient specifically refers to the amount of retention distribution difference, flow rate rhythm adjustment amount, and path offset parameter. Feeding compensation configuration includes flow rate control parameters, viscosity response relationship, and execution compensation coefficient.

[0024] The drift calibration module includes: The temperature difference detection submodule uses a temperature sensor to acquire data on the fluctuation trends of the sample injection chamber temperature and the ambient temperature, analyzes the trajectory of their changes, calculates the direction of temperature difference change, and establishes the temperature difference change range by combining the amplitude and direction of temperature difference change within each detection cycle. The temperature difference detection submodule uses a temperature sensor to collect the temperature values ​​inside the sample injection chamber and the external environment during the electrolysis process. The collection interval is set to once every 10 seconds, and 6 sets of data are collected continuously in each cycle, denoted as . ( (Representing the acquisition time), for example, the injection chamber temperature during the first acquisition cycle is recorded as follows: , , , , , The ambient temperatures were recorded as follows: , , , , , Based on the above data, trend analysis was performed on the temperature change trajectory. Using the acquisition time as the x-axis and temperature as the y-axis, temperature change curves for the injection chamber and the environment were plotted. By observing the trend of these curves and comparing the temperature change trajectories of the two, it was determined that the overall temperature trajectory of the injection chamber showed an upward trend. For example... If similar calculated values ​​are continuously positive, then the trend is upward. Simultaneously, analyze the ambient temperature curve to identify the upward trend of the ambient temperature trajectory, such as... The continuous value is positive. Further calculations are performed to determine the direction of temperature difference change between the injection chamber and the ambient temperature. Calculation, for example If the continuous temperature difference value changes All are positive values, for example Multiple consecutive positive values ​​indicate a continuously widening temperature difference trend. Continuous calculations determine the trend of change, and combine each The range of temperature difference variation is established as described above. If the temperature range within this collection period is 1.8℃-2.1℃, then 1.8℃ and 2.1℃ are used as the endpoints of the interval to generate the temperature difference range.

[0025] The synchronization determination submodule evaluates the trend of temperature synchronization deviation within a continuous sampling period based on the temperature difference change range. Combining the duration and direction of deviation, it judges the delayed response behavior of the injection chamber, analyzes the continuity and consistency of the synchronization deviation, and generates synchronization deviation parameters. The specific formula for determining the delayed response behavior of the injection chamber is as follows: ; Calculate the trend index of synchronization deviation; in, As a trend indicator of synchronization deviation, This represents the normalized value of the injection chamber temperature collected during the i-th period. The normalized value of the ambient temperature collected during the i-th period is... This represents the normalized value of the injection chamber temperature collected during the (i-1)th cycle. The normalized value of the ambient temperature collected during the (i-1)th period. Let be the temperature-normalized stability constant. For symbolic functions, Let be the directional weight corresponding to the direction of temperature change within the i-th period. The total number of sampling periods. This is the index number of the current sampling period. This is the index number of the previous sampling period.

[0026] formula: ; Detailed explanation of the formula and its calculation derivation: The formula is used to calculate the continuous consistency trend of temperature synchronization deviation, and the results are used to determine the continuity and significance of the influence of temperature drift trend on manganese electrolysis concentration measurement. Parameter meanings and settings: The normalized value of the injection chamber temperature collected in the i-th cycle is obtained by collecting actual injection chamber temperature data in real time and then normalizing it. The monitoring temperature in the second cycle is set to 35.6 degrees Celsius, and the normalization interval is set to the lowest actual monitoring value of 20 degrees Celsius and the highest actual monitoring value of 40 degrees Celsius. After normalization... ; The normalized value of the ambient temperature collected in the i-th period is calculated by monitoring the actual ambient temperature, which is 30.4 degrees Celsius in the same period. ; To prevent the denominator from being zero, the normalized value of the temperature stability constant is set to 0.01. This is a sign function used to determine whether the direction of the temperature difference in the current cycle is consistent with that in the previous cycle; The weight is the direction of temperature change in the i-th period. The weight standard is graded according to the magnitude of the absolute value of the temperature difference change. The weight value is set by the fluctuation range detected on site. The weight is 0.8 for below 0.2, 1.0 for 0.2-0.5, and 1.2 for above 0.5. Here, the absolute value of the temperature difference in the second period is |0.78-0.52|=0.26, corresponding to a weight of 1.0. To represent the total number of sampling periods, we set n=4; The monitoring sampling data is set as follows (all are normalized values): Period 1: , Weight 0.8; Second period: , Weight 1.0; 3rd period: , Weight 1.0; 4th period: , Weight 1.2; Substitute the parameters into the formula to calculate: Cycle 2: ; ; ; 3rd cycle: ; ; ; 4th cycle: ; ; ; The formula is to take the average: ; The result of 0.1820 indicates that the temperature synchronization deviation of the current cycle is relatively consistent. A value greater than 0 indicates that the temperature drift direction is consistent for most cycles and the change range is within the normal monitoring range. This trend value is used to generate the synchronization deviation parameter, which can provide a quantitative reference for subsequent automatic compensation or warning logic.

[0027] The response correction submodule calculates the correspondence between the degree of deviation and the difference in ion response in the concentration curve based on the synchronization deviation parameter, and corrects the current period concentration reading by the ratio between temperature difference change and concentration fluctuation, generating a concentration calibration value. The response correction submodule analyzes the correspondence between the synchronization deviation parameter and the ion concentration curve based on the synchronization deviation parameter, using the actual measured ion concentration data as the analysis object, such as the concentration data points of nickel ions continuously measured during the injection process. ( Representing the Using the detection time points as the analysis data, assuming that the concentration was measured at each of the 5 sampling points, the data is analyzed at each of the 5 detection time points. , , , , The synchronization deviation parameters of the temperature difference changes at each concentration point and the corresponding time point are calculated accordingly, such as the time point. The corresponding temperature difference is 2.1℃. A pre-set concentration correction reference value is invoked. When the temperature difference is between 2.0℃ and 2.2℃, the corresponding concentration correction coefficient is set to 0.95. This coefficient is pre-set based on actual production process experience. This correction coefficient is multiplied by the measured concentration point, for example, to obtain the corrected concentration. Similar calculations are performed sequentially for all concentration points to obtain the corrected concentration values ​​for each point. A corrected concentration curve is then established. The corrected concentration curve data for the current period is retrieved, and the average correction magnitude for each concentration point is calculated. For example... The average correction margin is the current periodic concentration calibration value.

[0028] The structure matching module includes: The period comparison submodule obtains the concentration calibration value, collects the period of lattice spacing change during the deposition process, obtains the time frequency of impurity ion concentration fluctuation, compares the lattice spacing period and the frequency difference of each impurity ion concentration fluctuation, analyzes the consistency of period changes, and generates period matching differences. The periodic comparison submodule acquires concentration calibration values, specifically corrected concentration data of multiple impurity ions during electrolysis. Real-time data of the crystal lattice spacing during deposition is continuously acquired using an X-ray diffractometer; for example, the lattice spacing measured in six consecutive measurements is... , , , , , Calculate the change in lattice spacing between every two consecutive measurement points. For example, calculate the change from the first measurement to the second measurement. The process involves sequentially calculating the changes in lattice spacing to establish periodic data. A mass spectrometer is used to simultaneously measure the concentration fluctuation frequency of impurity ions, such as nickel ions. The interval between the highest concentration points within a period is defined as one fluctuation cycle. For example, the intervals between concentration peaks are 3 minutes, 3.2 minutes, and 3.1 minutes. The lattice cycle is compared with the nickel ion fluctuation cycle one by one. The cycle matching ratio is obtained by dividing the lattice cycle by the impurity ion cycle. A cycle matching is considered to occur when the matching ratio is between 0.95 and 1.05. For example, the matching ratio between a lattice cycle of 3.05 minutes and a nickel ion cycle of 3 minutes is 3.05 ÷ 3 = 1.017, satisfying the cycle matching condition. Multiple matching ratios are summarized and analyzed to calculate the matching difference. If the matching difference is between 0.95 and 1.05, the cycle matching difference is considered small; if it exceeds this range, the difference is considered large, generating a cycle matching difference.

[0029] The synchronization analysis submodule determines the temporal synchronization correlation between the structural period and the fluctuation of impurity ion concentration based on the period matching difference, filters period segments with consistent changing trends, and establishes the correlation between the structural period and the fluctuation of ion concentration by combining the structural period synchronization parameters, thereby generating structural synchronization parameters. The synchronous analysis submodule analyzes data based on period matching differences, using period groups with matching differences close to 1 as the base data. For example, it selects three data groups with period matching differences of 1.017, 0.998, and 1.004, and sequentially calls up the period data of lattice spacing and the fluctuation frequency of impurity ion concentration within the corresponding period group. Using the time coordinates corresponding to the start and end points of lattice changes within the period data as a benchmark, it determines whether the peak occurrence time of impurity ion concentration fluctuation falls within the start and end time range of the lattice period data. If the start and end time of the lattice period is 0 to 3.05 minutes, and the peak occurrence time of impurity ion concentration is 2.95 minutes, then it is judged... If the fluctuation time of impurity ion concentration is within the lattice period range, repeat the above steps to judge each set of data one by one, record the results of each set of judgments, and screen out the period groups in which the occurrence time of all impurity ion peaks falls within the lattice period range. For example, in a group with a period difference of 1.004, the lattice period range is 0 to 3.10 minutes, and the impurity ion peak is at 3.05 minutes, which falls completely within the lattice period range. Determine the structural period synchronization parameter by the proportion of the number of screened period groups to the total number of period groups. When this proportion exceeds 80%, it is confirmed that the structural period and impurity ion fluctuations have strong synchronization, and the structural synchronization parameter is established.

[0030] The influence identification submodule analyzes the trend of impurity ion concentration fluctuations on crystal deposition configuration changes based on structural synchronization parameters, identifies the influence relationship of each impurity ion concentration fluctuation on the deposition structure, and generates impurity response information. The specific formula for identifying the influence of concentration fluctuations of each impurity ion on the deposition structure is as follows: ; Calculate the impact intensity value; in, Let be the intensity value of the influence of the k-th impurity ion on the crystal deposition configuration. This is the normalized value of the concentration variation of the k-th impurity ion in the j-th lattice period. Let be the difference between the frequency of the concentration fluctuation of the k-th impurity ion and the frequency of the spacing fluctuation in the j-th lattice period. This is the normalized value of the spacing offset of the structurally stable region caused by the k-th impurity within the j-th lattice period. This is the normalized value of the grain arrangement perturbation value in the lattice distortion region caused by the k-th impurity within the j-th lattice period. The lattice period number, Number the impurity types. This represents the total number of cycles.

[0031] formula: ; Detailed explanation of the formula and its calculation derivation: The formula is used to calculate the intensity of the influence of each impurity ion on the crystal deposition configuration, and the results are used to determine the effect of impurity fluctuations on the deposition structure. Parameter meanings and settings: Let be the intensity value of the influence of the k-th impurity ion on the crystal deposition configuration. This is the normalized value of the concentration variation of the k-th impurity ion in the j-th lattice period. The value is obtained by normalizing the original value of the impurity concentration collected in real time by the ion concentration monitor and the value of the normalized value with the maximum concentration variation range in the same period. It is the difference between the concentration fluctuation frequency of the k-th impurity ion and the structural spacing fluctuation frequency in the j-th lattice period. The impurity fluctuation frequency is obtained by extracting the main peak from the high-frequency sampled concentration time series signal. This is the normalized value of the spacing shift of the structurally stable region caused by the k-th impurity within the j-th lattice period. The data is obtained by quantitatively measuring the lattice spacing change using an X-ray diffractometer. The normalized value of the grain arrangement perturbation value in the lattice distortion region caused by the k-th impurity in the j-th lattice period is obtained by statistical analysis of grain rearrangement in the distortion region using high-resolution electron microscopy. The total number of cycles is set to 5 cycles for lattice structure sampling and monitoring records. The lattice period number is taken from 1 to 5. Impurity types are numbered, with copper impurities numbered 1; the first cycle is defined as follows: , , , Second cycle: , , , , 3rd cycle: , , , , 4th cycle: , , , , 5th cycle: , , , ; Substitute the parameters into the formula to calculate: ; ; ; The results show that the influence intensity of copper impurities in the lattice structure deposition process of this batch is 0.0436, which represents the combined normalized influence of copper impurity fluctuations and structural disturbances. Corresponding to the impurity response information, it can be used to assess the risk contribution of copper impurities to the deposited structure in the quality prediction system.

[0032] The process parameter inversion module includes: The periodic analysis submodule calls the impurity response information, analyzes the element concentration trend in the residual liquid at the end of each period, obtains the time position of each raw material addition, statistically analyzes the concentration change range of impurities within the period span, calculates the concentration fluctuation range within each period, and generates the periodic concentration range. The cycle analysis submodule calls the impurity response information, which is the data on the influence of different impurity ions on the crystal lattice established in the previous steps. It selects the concentrations of multiple impurity ions, such as nickel and copper ions, as the analysis objects. At the end of each electrolysis cycle, it collects the concentration data of nickel and copper ions in the residual liquid. For example, at the end of a certain cycle, the nickel ion concentration is measured to be 3.8 mg / L and the copper ion concentration is 2.4 mg / L. It also records the concentration values ​​of the same ions from the previous cycle, which are 3.2 mg / L for nickel ions and 1.9 mg / L for copper ions. It calculates the changing trend of each impurity ion concentration, specifically by subtracting the concentration from the previous cycle from the current cycle concentration. For example, nickel ion concentration is 3.8 mg / L - 3.2 mg / L = 0.6 mg / L, and copper ion concentration is 2.4 mg / L - 1.9 mg / L = 0.5 mg / L. It then calls the raw material addition control system to obtain the raw material addition time and records the corresponding electrolytic voltage. The starting point and position of the feed addition in the tank are determined, for example, the feed addition starts at the 10th minute and ends at the 15th minute of this cycle. This clarifies the temporal range of the cycle span. Based on the impurity concentration data of two consecutive cycles, the concentration fluctuation range of each ion within each cycle is statistically analyzed. Specifically, it is calculated by the difference between the maximum and minimum concentration values ​​measured in each cycle. For example, in the current cycle, the maximum nickel ion concentration is 4.0 mg / L and the minimum is 3.0 mg / L, with a fluctuation range of 4.0 mg / L - 3.0 mg / L = 1.0 mg / L. Based on this, the fluctuation range of copper ions is calculated to be 0.8 mg / L. The fluctuation range of each ion is recorded. Based on the data recording of the concentration fluctuation range in multiple cycles, the concentration fluctuation range of each ion in this cycle is determined. For example, the fluctuation range of nickel ions is 3.0 mg / L - 4.0 mg / L, and the fluctuation range of copper ions is 1.9 mg / L - 2.7 mg / L, thus generating the cycle concentration range.

[0033] The interval filtering submodule filters concentration segments with consistent fluctuation directions based on the periodic concentration interval. Combining the impurity concentration change trend parameter, it identifies concentration change segments consistent with the periodic fluctuation, extracts fluctuation consistency markers, and generates consistent concentration segments. The interval filtering submodule, based on the periodic concentration intervals and the concentration data of multiple recorded periods, calls the periodic concentration interval data of impurity ions such as nickel ions and copper ions, and determines the direction of concentration fluctuation of impurity ions period by period, using the order in which the maximum and minimum concentration values ​​occur as the basis. For example, if the maximum concentration value of nickel ions in period 1 occurs in the later part of the period, while the minimum concentration value occurs in the earlier part of the period, then the fluctuation trend of nickel ions in that period is determined to be an upward trend. Similar judgments on the fluctuation trend are performed for each ion in multiple periods, comparing whether the ion concentration fluctuation trend is consistent across multiple consecutive periods. To ensure consistency, the standard for consistent fluctuation direction is that the fluctuation direction of impurity ions is exactly the same within three consecutive periods. For example, if the concentration of nickel ions increases in consecutive periods 1, 2, and 3, and the concentration of copper ions increases in periods 1 and 2 but decreases in period 3, then nickel ions meet the standard for consistent fluctuation direction, while copper ions do not. All concentration fluctuation data segments that meet the standard for consistent fluctuation direction are filtered, and the concentration values ​​in each segment are marked. For example, the concentration segment of nickel ions from period 1 to period 3 is recorded as the consistent concentration segment. This marked segment is associated with the impurity concentration change trend parameter to generate the consistent concentration segment.

[0034] The path determination submodule calls the consistency concentration segment, combines the raw material addition time, current change direction and temperature change area, performs path determination on the initial concentration behavior of impurities in the concentration change segment, analyzes the ratio and trend of each parameter, establishes the concentration value corresponding to the diffusion path start point, and generates the diffusion start point concentration. The path determination submodule calls the consistency concentration segment, using the recorded raw material addition time period, current change direction, and temperature change area as a basis to compare the initial concentration behavior of impurity ions in the consistency concentration segment one by one. First, taking the raw material addition time period as a reference, it records the correspondence between the initial concentration data of each concentration segment and the starting time of raw material addition. For example, the initial concentration of nickel ions in the consistency concentration segment is 3.0 mg / L at the 10th minute, which is consistent with the starting time of raw material addition. It further analyzes the direction of current change in the electrolytic cell at that moment. For example, if the current is on an upward trend at that moment, it uses the actual electrolytic cell measured current data as a basis, such as 550A at the 10th minute and 550A at the 11th minute. The minute current is 560A, and the current rise rate is 560A-550A=10A / minute, confirming an upward trend in the current. Then, the temperature change data recorded by the temperature sensor is used to determine the temperature change trend in the electrolysis region at the starting moment. For example, if the temperature rises from 28℃ to 29℃ at that moment, the temperature change shows an upward trend. The above parameters are compared one by one. The standard is that the three parameters of the raw material addition start time, the current rise trend, and the temperature rise trend are all met simultaneously. The concentration at the starting point of the diffusion path is determined, and the initial concentration data is used as the concentration value corresponding to the starting point of the diffusion path. For example, if the initial concentration of nickel ions is 3.0mg / L, the diffusion starting point concentration is generated.

[0035] The path correction module includes: The retention determination submodule calls the diffusion initiation concentration, calculates the direction of ion retention time change at each detection point within the boundary region, detects ion retention data at multiple detection points, determines the ion migration delay phenomenon, and generates migration delay parameters by combining the retention time trend. The retention determination submodule calls the diffusion initiation concentration and uses multiple detection points set in the boundary area of ​​the electrolytic cell as the implementation objects. For example, if the detection points are A, B, and C, it records the time data of nickel ion concentration arriving at each detection point from the diffusion initiation. For instance, if the nickel ion arrives at detection point A at minute 8, at B at minute 10, and at C at minute 13, it calculates the retention time of nickel ions between each pair of adjacent detection points. For example, the retention time of nickel ions between A and B is 10 minutes minus 8 minutes, which is 2 minutes; the retention time between B and C is 13 minutes minus 10 minutes, which is 3 minutes. The retention times of each segment are then compared. The direction of change in residence time for each segment is calculated. For example, if the residence time from segment AB to segment BC increases from 2 minutes to 3 minutes, the direction of change is an increasing trend. The trend of change in multiple similar sampling data is statistically analyzed. If the residence time trend is increasing in three consecutive measurements, it is judged as a migration delay phenomenon. The residence data from multiple detection points are combined, and the residence trend of each segment obtained from each sampling is recorded and marked as either an increasing trend or a decreasing trend. The frequency of the trend between each detection point is statistically analyzed. An increasing trend accounting for more than 70% of the total trend frequency is used as the standard for judging migration delay. The proportion of the ion residence time increasing trend is calculated by combining the results of multiple measurements to clarify the degree of migration delay and generate migration delay parameters.

[0036] The distribution comparison submodule compares the distribution differences of loitering behavior in multiple regions based on the migration delay parameter, analyzes the spatial distribution trend of loitering phenomenon, and generates loitering distribution coefficient; The distribution comparison submodule, based on the migration delay parameter, retrieves detection point data from different locations within the electrolytic cell boundary region. For example, regions I, II, and III each contain multiple detection points; detection points A1 and A2 in region I, B1 and B2 in region II, and C1 and C2 in region III. It calculates the average residence time of each detection point within each region. For instance, if detection points A1 in region I have a residence time of 2 minutes and A2 has a residence time of 2.5 minutes, the average residence time for this region is calculated as (2 + 2.5) / 2 = 2.25 minutes. Similarly, it calculates the average residence time for regions II and III. The module then compares the average residence times for each region. For example, if region I has a residence time of 2.25 minutes, region II has a residence time of 3 minutes, and region III has a residence time of 3.5 minutes, the comparison is complete. By direct comparison, the differences in retention behavior among different regions were confirmed, and region III was found to have the longest average retention time. Statistical calculations were performed on the average retention time data of different regions. The difference between the maximum and minimum average retention time between regions was calculated. For example, 3.5 minutes in region III minus 2.25 minutes in region I equals 1.25 minutes. This time difference was used as the basis for assessing the differences in retention distribution among different regions, and was divided by the regional average to determine the retention distribution coefficient, i.e., 1.25 minutes divided by (2.25+3+3.5) / 3, resulting in a retention distribution coefficient of approximately 0.43. The distribution trend of retention phenomena in the electrolytic cell space was confirmed based on the magnitude of the retention distribution coefficient, thus generating the retention distribution coefficient.

[0037] The behavior correction submodule calls the retention distribution coefficient to analyze the retention behavior of the boundary path and the migration trend of the ion continuous sampling segment, corrects the behavior offset parameters of the path segment, adjusts the flow rate rhythm in the current path prediction, and generates the path delay coefficient. The behavior correction submodule calls the retention distribution coefficient to record the migration path data of ions continuously sampled within the electrolyzer boundary area. For example, the starting point of the sampling path is the location corresponding to the diffusion initiation concentration, and the continuous sampling points along the path are detection points X, Y, and Z. It calculates the ion migration velocity between two adjacent sampling points. For instance, the migration velocity of ions from X to Y is the spatial distance (time of peak concentration at position Y minus time of peak concentration at position X) divided by the time difference. If the spatial distance from X to Y is 20 cm and the time difference between the peak concentrations is 2 minutes, then the migration velocity is 20 cm / 2 minutes = 10 cm / minute. Similarly, the migration velocity from Y to Z is calculated. If the velocity decreases segment by segment, such as from X to Y at 10 cm / minute... If the velocity decreases to 7 cm / min in the YZ segment, the migration trend is determined to be deceleration. The influence of the retention distribution coefficient on the ion migration trend in the boundary path is analyzed. This is done by multiplying the retention distribution coefficient by the actual measured migration velocity. For example, if the retention distribution coefficient is 0.43 and the actual migration velocity in the XY segment is 10 cm / min, the corrected velocity is 10 cm / min × (1 - 0.43) = 5.7 cm / min. The original flow rate rhythm in the path prediction is adjusted accordingly. For example, if the original predicted path velocity was 8 cm / min, the adjusted velocity is 5.7 cm / min. Through the above operations, all path segments are corrected one by one, ultimately generating the path hysteresis coefficient.

[0038] The deviation verification module includes: The flow rate deviation submodule calls the path delay coefficient, analyzes the target feed flow rate and the actual execution flow rate in the feed instruction, calculates the degree of difference, determines the deviation range, and generates the flow rate deviation range by combining the periodic flow rate data. The flow rate deviation submodule calls the path delay coefficient, using the target feed flow rate given by the electrolytic cell feeding command as a benchmark. For example, if the target feed flow rate is 20 liters per minute, the actual feed flow rate is measured in real time by a flow sensor. For instance, the measured actual flow rates are 18 liters in the first minute, 17 liters in the second minute, and 19 liters in the third minute. The difference between the target feed flow rate and the actual flow rate is calculated for each minute. For example, the flow rate deviation in the first minute is the target flow rate of 20 liters minus the actual flow rate of 18 liters, which is 2 liters. The deviations at multiple measurement moments are statistically analyzed, and the average flow rate deviation within each cycle is recorded. For example, averaging the deviations of 2 liters, 3 liters, and 1 liter at three measurement moments yields the average deviation within the cycle as (2 + 3). +1) ÷ 3 = 2 liters; Further, the aforementioned path delay coefficient is called, for example, the path delay coefficient is 0.3. By multiplying this coefficient by the average deviation within the cycle, i.e., 2 liters × 0.3 = 0.6 liters, the product value of 0.6 liters is used as the correction amount for the degree of flow velocity deviation. The percentage relationship between the correction amount and the target flow velocity is analyzed. For example, the proportion of 0.6 liters to the target flow velocity of 20 liters is 0.6 ÷ 20 = 3%. Based on this proportion, the degree of flow velocity deviation exceeding 5% is set as a high deviation range, and less than or equal to 5% is set as a normal deviation range. The 3% calculated in this cycle is included in the normal deviation range. The deviation range data of multiple cycles are determined by this method. The statistical data of the deviation range of each cycle are summarized to generate the flow velocity deviation range.

[0039] The viscosity analysis submodule, based on the flow rate deviation range, calls upon the current flow characteristics and standard flow state of the feeding medium to analyze the impact of viscosity changes on the flow rate during the execution phase and generates a viscosity influence coefficient. The viscosity analysis submodule, based on the flow rate deviation range, retrieves the flow characteristic data of the feeding medium in its current state. For example, if the medium is electrolyte, and the current viscosity is measured to be 5.5 mPa·s using a viscometer, it also retrieves the viscosity value of the electrolyte under standard flow conditions set in the standard process, for example, a standard viscosity value set to 5.0 mPa·s. The module records the viscosity values ​​of the current state and the standard state respectively. It calculates the difference between the current viscosity and the standard viscosity, specifically by subtracting the standard viscosity value from the current viscosity value. For example, the calculation result is 5.5 mPa·s - 5.0 mPa·s = 0.5 mPa·s. The module then analyzes the impact of this viscosity difference on the feeding flow rate during the actual execution phase, retrieving data from multiple cycles. The actual flow rate of the material is measured, and the deviation between the actual flow rate and the target flow rate in each cycle is calculated. For example, in the aforementioned actual flow rate measurement data, the deviation in the first cycle is 2 liters. The deviation data for each cycle are recorded as 2 liters, 3 liters, and 1 liter, respectively. The viscosity difference of 0.5 mPa·s is correlated with the actual flow rate deviation data. Specifically, the average value of the flow rate deviation in multiple cycles is calculated. For example, the average value is (2+3+1)÷3=2 liters. The viscosity difference is divided by the average flow rate deviation, i.e., 0.5 mPa·s÷2 liters=0.25 mPa·s / liter. This value is defined as the viscosity influence coefficient. This coefficient is used to quantify the degree of influence of viscosity change on the actual flow rate in the feeding execution stage, and the viscosity influence coefficient is generated.

[0040] The signal adjustment submodule calls the viscosity influence coefficient to adjust the flow rate control parameter range of the control signal, establishes the feedback relationship between viscosity and execution response offset, and generates the feeding compensation configuration. The signal adjustment submodule calls the viscosity influence coefficient, using the original feed flow rate control parameter range in the control signal as the adjustment basis. For example, if the original flow rate control parameter range was set to 18 to 22 liters per minute, the viscosity influence coefficient of 0.25 mPa·s / L is used to calculate the flow rate parameter adjustment corresponding to the viscosity change. Specifically, this is the product of the viscosity influence coefficient and the viscosity change, i.e., 0.25 mPa·s / L × 0.5 mPa·s = 0.125 liters. This adjustment amount is used to adjust the upper and lower limits of the original flow rate control parameter range separately. For example, the upper limit flow rate of 22 liters minus the adjustment amount of 0.125 liters results in an adjusted flow rate of 21.875 liters, and the lower limit flow rate of 18 liters minus the adjustment amount of 0.125 liters results in an adjusted flow rate of 21.875 liters. The flow rate was then adjusted to 17.875 liters, completing the adjustment of the flow rate control parameter range. Further, based on the adjusted flow rate control parameter range, the actual measured viscosity and flow rate data were retrieved, and the data correlation between the measured viscosity change and the actual executed flow rate deviation was compared one by one. For example, a viscosity deviation of 0.5 mPa·s corresponds to an actual flow rate deviation of 2 liters. This establishes a feedback data relationship between viscosity and the executed response deviation. Specifically, a viscosity change of 0.1 mPa·s corresponds to an actual flow rate deviation of 0.4 liters. Multiple similar viscosity-flow rate data records were recorded and summarized to form a corresponding relationship dataset. Finally, the adjusted flow rate control parameter range and the corresponding relationship dataset were used to generate the feeding compensation configuration.

[0041] An AI-based intelligent prediction method for the production quality of electrolytic manganese processes is proposed. This method is applied to an AI-based intelligent prediction system for the production quality of electrolytic manganese processes. The method includes: S1: Using a temperature sensor, analyze the change trajectory between the temperature fluctuation trends of the injection chamber and the ambient temperature, evaluate the correspondence between temperature synchronization deviation and ion response difference, correct the current cycle concentration reading, and generate a concentration calibration value. S2: Using the concentration calibration value, by comparing the period of lattice spacing change during the deposition process with the time frequency of each impurity ion concentration fluctuation, the synchronous correlation of the structural period is determined, the influence of ion concentration fluctuation on the formation of a stable crystal configuration is identified, and impurity response information is generated. S3: Call the impurity response information, analyze the positional relationship between the residual liquid element concentration trend and the raw material addition behavior, call the raw material addition time period, current change direction and temperature change area, make path judgment on the impurity initial concentration change behavior, and generate diffusion start concentration; S4: Using the diffusion initiation concentration, by calculating the retention change direction between multiple detection points in the boundary region, the migration delay phenomenon of ions is determined, the behavior offset parameter of the path segment is corrected, the flow rate rhythm in the current path prediction is adjusted, and the path delay coefficient is generated. S5: Call the path delay coefficient to analyze the deviation between the target feed flow rate and the actual execution flow rate in the feed instruction, call the current flow characteristics of the feed medium and the standard flow state for comparison, adjust the flow rate control parameter range of the control signal, establish the feedback relationship between viscosity and execution response offset, and generate feed compensation configuration.

[0042] The above embodiments can be implemented, in whole or in part, by software, hardware (such as circuits), firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of the present invention are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can 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 sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0043] It should be understood that the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. A and B can be singular or plural. Additionally, the character " / " in this article generally indicates an "or" relationship between the preceding and following related objects, but it can also represent an "and / or" relationship. Please refer to the context for a more accurate understanding.

[0044] In this invention, "at least one" means one or more, and "more than one" means two or more. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of a single item or a plurality of items. For example, at least one of a, b, or c can represent: a, b, c, ab, ac, bc, or abc, where a, b, and c can be a single item or multiple items.

[0045] It should be understood that, in various embodiments of the present invention, the order of the above-mentioned process numbers does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0046] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0047] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the devices, apparatuses, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0048] In the several embodiments provided by this invention, it should be understood that the disclosed devices, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another device, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0049] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0050] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0051] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0052] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

Claims

1. An AI-based intelligent prediction system for the production quality of electrolytic manganese processes, characterized in that, The system includes: The drift calibration module uses a temperature sensor to analyze the change trajectory between the temperature fluctuation trends of the injection chamber and the ambient temperature, evaluate the correspondence between temperature synchronization deviation and ion response difference, correct the current cycle concentration reading, and generate a concentration calibration value. The structure matching module uses the concentration calibration value to determine the synchronous correlation of the structure period by comparing the period of the lattice spacing change during the deposition process with the time frequency of the concentration fluctuation of each impurity ion, identify the influence of ion concentration fluctuation on the formation of a stable crystal configuration, and generate impurity response information. The process parameter inversion module calls the impurity response information, analyzes the positional relationship between the residual liquid element concentration trend and the raw material addition behavior, calls the raw material addition time period, current change direction and temperature change region, performs path judgment on the initial concentration change behavior of impurities, and generates the diffusion starting point concentration. The path correction module uses the diffusion initiation concentration to calculate the retention change direction between multiple detection points in the boundary region, determine the migration delay phenomenon of ions, correct the behavior offset parameters of the path segment, adjust the flow rate rhythm in the current path prediction, and generate the path delay coefficient.

2. The AI-based intelligent prediction system for electrolytic manganese production quality according to claim 1, characterized in that, The concentration calibration value includes the direction of temperature difference change, the magnitude of response delay, and the concentration correction ratio. The impurity response information specifically includes the periodic frequency matching factor, the structural correlation strength, and the degree of impurity influence. The diffusion initiation concentration includes the concentration fluctuation range, the initial behavior characteristics, and the direction of diffusion trend. The path delay coefficient specifically refers to the amount of retention distribution difference, the amount of flow rate rhythm adjustment, and the path offset parameter.

3. The AI-based intelligent prediction system for electrolytic manganese production quality according to claim 1, characterized in that, The drift calibration module includes: The temperature difference detection submodule uses a temperature sensor to acquire data on the fluctuation trends of the sample injection chamber temperature and the ambient temperature, analyzes the trajectory of their changes, calculates the direction of temperature difference change, and establishes the temperature difference change range by combining the amplitude and direction of temperature difference change within each detection cycle. The synchronization determination submodule evaluates the trend of temperature synchronization deviation within the continuous sampling period based on the temperature difference change range, and judges the delayed response behavior of the injection chamber by combining the deviation duration and change direction, analyzes the continuity and consistency of the synchronization deviation, and generates synchronization deviation parameters. The response correction submodule calculates the correspondence between the degree of deviation and the difference in ion response in the concentration curve based on the synchronization deviation parameter, and corrects the current period concentration reading by the ratio between temperature difference change and concentration fluctuation, and generates a concentration calibration value.

4. The AI-based intelligent prediction system for electrolytic manganese production quality according to claim 3, characterized in that, The specific formula for determining the delayed response behavior of the injection chamber is as follows: ； Calculate the trend index of synchronization deviation; in, As a trend indicator of synchronization deviation, This represents the normalized value of the injection chamber temperature collected during the i-th period. The normalized value of the ambient temperature collected during the i-th period is... This represents the normalized value of the injection chamber temperature collected during the (i-1)th cycle. The normalized value of the ambient temperature collected during the (i-1)th period. Let be the temperature-normalized stability constant. For symbolic functions, Let be the directional weight corresponding to the direction of temperature change within the i-th period. The total number of sampling periods. This is the index number of the current sampling period. This is the index number of the previous sampling period.

5. The AI-based intelligent prediction system for electrolytic manganese production quality according to claim 3, characterized in that, The structure matching module includes: The period comparison submodule obtains the concentration calibration value, collects the period of lattice spacing change during the deposition process, obtains the time frequency of impurity ion concentration fluctuation, compares the lattice spacing period and the frequency difference of each impurity ion concentration fluctuation, analyzes the consistency of period changes, and generates period matching differences. The synchronization analysis submodule determines the temporal synchronization correlation between the structural period and the fluctuation of impurity ion concentration based on the period matching difference, filters period segments with consistent changing trends, and establishes the correlation between the structural period and the fluctuation of ion concentration by combining the structural period synchronization parameters, thereby generating structural synchronization parameters. The influence identification submodule analyzes the trend of impurity ion concentration fluctuations on crystal deposition configuration changes based on the structural synchronization parameters, identifies the influence relationship of each impurity ion concentration fluctuation on the deposition structure, and generates impurity response information.

6. The AI-based intelligent prediction system for electrolytic manganese production quality according to claim 5, characterized in that, The specific formula for identifying the influence of concentration fluctuations of each impurity ion on the deposition structure is as follows: ； Calculate the influence intensity value; in, Let be the intensity value of the influence of the k-th impurity ion on the crystal deposition configuration. This is the normalized value of the concentration variation of the k-th impurity ion in the j-th lattice period. Let be the difference between the frequency of the concentration fluctuation of the k-th impurity ion and the frequency of the spacing fluctuation in the j-th lattice period. This is the normalized value of the spacing offset of the structurally stable region caused by the k-th impurity within the j-th lattice period. This is the normalized value of the grain arrangement perturbation value in the lattice distortion region caused by the k-th impurity within the j-th lattice period. The lattice period number, Number the impurity types. This represents the total number of cycles.

7. The AI-based intelligent prediction system for electrolytic manganese production quality according to claim 5, characterized in that, The process parameter inversion module includes: The periodic analysis submodule calls the impurity response information to analyze the element concentration trend in the residual liquid at the end of each period, obtains the time position of each raw material addition, statistically analyzes the concentration change range of impurities within the period span, calculates the concentration fluctuation range within each period, and generates the periodic concentration range. The interval filtering submodule filters concentration segments with consistent fluctuation directions based on the periodic concentration interval, and, in conjunction with the impurity concentration change trend parameter, identifies concentration change segments consistent with periodic fluctuations, extracts fluctuation consistency markers, and generates consistent concentration segments. The path determination submodule calls the consistency concentration range, combines the raw material addition time period, current change direction and temperature change area to determine the initial concentration behavior of impurities in the concentration change range, analyzes the ratio and trend of each parameter, establishes the concentration value corresponding to the starting point of the diffusion path, and generates the diffusion starting point concentration.

8. The AI-based intelligent prediction system for electrolytic manganese production quality according to claim 7, characterized in that, The path correction module includes: The retention determination submodule calls the diffusion initiation concentration, calculates the direction of ion retention time change at each detection point within the boundary region, detects ion retention data at multiple detection points, determines the ion migration delay phenomenon, and generates migration delay parameters based on the retention time trend. The distribution comparison submodule compares the distribution differences of loitering behavior in multiple regions based on the migration delay parameter, analyzes the spatial distribution trend of loitering phenomena, and generates a loitering distribution coefficient. The behavior correction submodule calls the retention distribution coefficient to analyze the retention behavior of the boundary path and the migration trend of the ion continuous sampling segment, corrects the behavior offset parameter of the path segment, adjusts the flow rate rhythm in the current path prediction, and generates the path delay coefficient.

9. The AI-based intelligent prediction system for electrolytic manganese production quality according to claim 1, characterized in that, The system also includes: The deviation verification module calls the path delay coefficient to analyze the degree of deviation between the target feeding flow rate and the actual execution flow rate in the feeding instruction, compares the current flow characteristics of the feeding medium with the standard flow state, adjusts the flow rate control parameter range of the control signal, establishes the feedback relationship between viscosity and execution response offset, and generates feeding compensation configuration. The feeding compensation configuration includes flow rate control parameters, viscosity response relationship, and execution compensation coefficient; The deviation verification module includes: The flow rate deviation submodule calls the path delay coefficient to analyze the target feed flow rate and the actual execution flow rate in the feed instruction, calculates the degree of difference, determines the deviation range, and generates the flow rate deviation range by combining the periodic flow rate data. The viscosity analysis submodule, based on the flow rate deviation range, calls the current flow characteristics and standard flow state of the feeding medium to analyze the impact of viscosity changes on the flow rate during the execution phase and generates a viscosity influence coefficient. The signal adjustment submodule calls the viscosity influence coefficient to adjust the flow rate control parameter range of the control signal, establishes a feedback relationship between viscosity and execution response offset, and generates a feeding compensation configuration.

10. An AI-based intelligent prediction method for the production quality of electrolytic manganese processes, characterized in that, The method is used to implement the AI-based intelligent prediction system for electrolytic manganese process production quality as described in any one of claims 1-9, the method comprising: S1: Using a temperature sensor, analyze the change trajectory between the temperature fluctuation trends of the injection chamber and the ambient temperature, evaluate the correspondence between temperature synchronization deviation and ion response difference, correct the current cycle concentration reading, and generate a concentration calibration value. S2: Using the concentration calibration value, by comparing the period of lattice spacing change during the deposition process with the time frequency of each impurity ion concentration fluctuation, the synchronous correlation of the structural period is determined, the influence of ion concentration fluctuation on the formation of a stable crystal configuration is identified, and impurity response information is generated. S3: Call the impurity response information, analyze the positional relationship between the residual liquid element concentration trend and the raw material addition behavior, call the raw material addition time period, current change direction and temperature change area, make path judgment on the impurity initial concentration change behavior, and generate diffusion start concentration; S4: Using the diffusion initiation concentration, by calculating the retention change direction between multiple detection points in the boundary region, the migration delay phenomenon of ions is determined, the behavior offset parameter of the path segment is corrected, the flow rate rhythm in the current path prediction is adjusted, and the path delay coefficient is generated. S5: Call the path delay coefficient to analyze the deviation between the target feed flow rate and the actual execution flow rate in the feed instruction, call the current flow characteristics of the feed medium and the standard flow state for comparison, adjust the flow rate control parameter range of the control signal, establish the feedback relationship between viscosity and execution response offset, and generate feed compensation configuration.

Images