Green ball processing control method and device, electronic equipment and storage medium

Abstract

The present application relates to green ball processing control technical field, especially to a kind of green ball processing control method, device, electronic equipment and storage medium, the present application method first obtains the first distribution data of green ball diameter, powder characteristic data and the first moisture content of green ball;Then according to the first distribution data and the first moisture content, determine the first coefficient array of the first target item adjustment intensity characteristic;Then the first distribution data, the powder characteristic data and the first moisture content are substituted into green ball processing model, according to the obtained multiple outputs and the first coefficient array, obtain processing parameter score, and according to the processing parameter score, processing parameter is optimized, and obtains the first processing parameter array;Finally, according to the first processing parameter array, the multiple adjustment items of disc balling equipment are adjusted.The present application balanced adjustment control strategy, multiple processing indexes in green ball processing process are considered, and the quality and production efficiency of green ball are guaranteed.

Description

Technical Field

[0001] This invention relates to the field of green pellet processing control technology, and in particular to a green pellet processing control method, apparatus, electronic device and storage medium. Background Technology

[0002] In the pelletizing process of steel plants, "green pellets" are the core semi-finished product that transforms iron ore powder into qualified blast furnace feed (pellets). Their role is crucial throughout the entire pelletizing process, directly determining the quality of the final finished pellets and the efficiency of blast furnace ironmaking. This is because one of the core functions of green pellets is to transform the fine powder into uniformly sized, blocky semi-finished products (green pellets typically have a diameter of 8-16mm) through a "pelletizing" process (mixing fine iron ore powder with a binder and rolling it into pellets in a pelletizing machine). This lays the morphological foundation for subsequent roasting and furnace feeding.

[0003] Currently, the mainstream process for producing green pellets involves operators adding powder and moisture to the equipment based on the state of the green pellets in the pelletizing pan, and controlling the equipment's rotation speed and other parameters. In other words, the pelletizing process is entirely manual, and differences in individual responsibility and skill level can affect the quality of the green pellets, thus impacting the overall quality of the finished pellets.

[0004] Improving the yield of green pellets is an effective way to increase production while maintaining quality, as well as to reduce circulating pressure and equipment load, thereby lowering energy consumption and increasing production efficiency. Qualified green pellets require consideration of multiple parameters, including pellet diameter uniformity and moisture content, while simultaneously maintaining production efficiency. Since green pellet production equipment requires adjustment of numerous parameters, typically including powder feed rate, water addition, turntable speed, turntable angle, and scraper angle, automatically controlling and balancing multiple processing parameters is a significant challenge in the green pellet production process.

[0005] Therefore, it is necessary to develop a method for controlling the processing of green pellets. Summary of the Invention

[0006] The present invention provides a green pellet processing control method, apparatus, electronic device and storage medium to solve the problem of difficulty in balancing multiple processing indicators in green pellet processing in the prior art.

[0007] In a first aspect, embodiments of the present invention provide a method for controlling the processing of green pellets, comprising:

[0008] Obtain the first distribution data of green pellet diameter, powder characteristic data, and the first moisture content of green pellets;

[0009] Based on the first distribution data and the first moisture content, a first coefficient array characterizing the adjustment intensity of multiple target items is determined, wherein the multiple target items include: uniformity of green bulbs, moisture content of green bulbs, and diameter growth rate;

[0010] The first distribution data, the powder characteristic data, and the first moisture content are substituted into the green pellet processing model. Based on the multiple outputs and the first coefficient array, a processing parameter score is obtained, and the processing parameters are optimized based on the processing parameter score to obtain a first processing parameter array.

[0011] The multiple adjustment items of the disc ball-making equipment are adjusted according to the first processing parameter array.

[0012] In one possible implementation, determining a first coefficient array characterizing the adjustment intensity of multiple target items based on the first distribution data and the first moisture content includes:

[0013] Acquire the first distribution data, the standard deviation of the target green bulb diameter, the moisture content of the target green bulb, and the diameter growth conversion factor, wherein the first distribution data includes the mean diameter of the first green bulb and the standard deviation of the first green bulb diameter;

[0014] The diameter growth rate is determined based on the average diameter of the first green bulb and the average diameter of the second green bulb, wherein the time point corresponding to the average diameter of the second green bulb is earlier than that of the average diameter of the first green bulb.

[0015] Take the square root of the mean diameter of the first green bulb to obtain the first square root value;

[0016] The product of the first square root value and the diameter growth conversion factor is taken as the target diameter growth rate;

[0017] Calculate the difference between the target diameter growth rate and the diameter growth rate, and use it as the first growth rate difference;

[0018] Calculate the difference between the target green bulb moisture content and the first moisture content, and use it as the first moisture content difference;

[0019] The difference between the standard deviation of the target green bulb diameter and the standard deviation of the first green bulb diameter is calculated as the first green bulb uniformity difference;

[0020] Based on the first growth rate difference, the first moisture content difference, and the first bulb uniformity difference, a first coefficient array is constructed, wherein the first coefficient array includes: a coefficient characterizing the adjustment of diameter growth rate, a coefficient characterizing the adjustment of moisture content, and a coefficient characterizing the adjustment of bulb uniformity.

[0021] In one possible implementation, constructing the first coefficient array based on the first growth rate difference, the first moisture content difference, and the first green bulb uniformity difference includes:

[0022] Obtain the first ratio value and the second ratio value;

[0023] The growth rate difference, the moisture content difference, and the bulb uniformity difference are respectively added to the growth rate difference queue, the moisture content difference queue, and the bulb uniformity difference queue;

[0024] Based on the first formula, the growth rate difference queue, the moisture content difference queue, and the green bulb uniformity difference queue, the second growth rate difference coefficient, the second moisture content difference coefficient, and the second green bulb uniformity difference coefficient are determined respectively, wherein the first formula is:

[0025]

[0026] In the formula, This refers to the second growth rate difference coefficient, the second moisture content difference coefficient, or the second bulb uniformity difference coefficient. The first proportional value, The first in the growth rate difference queue, the moisture content difference queue, or the bulb uniformity difference queue One data point, This is the second proportional value. The total number of data points in the growth rate difference queue, moisture content difference queue, or green bulb uniformity difference queue;

[0027] The second growth rate difference coefficient, the second moisture content difference coefficient, and the second green ball uniformity difference coefficient are normalized, and the results are used to construct the first coefficient array.

[0028] In one possible implementation, the first distribution data, the powder characteristic data, and the first moisture content are substituted into the green pellet processing model. Based on the obtained multiple outputs and the first coefficient array, a processing parameter score is obtained. The processing parameters are then optimized based on the processing parameter score to obtain a first processing parameter array, including:

[0029] Obtain multiple second processing parameter arrays, wherein each second processing parameter array includes processing parameters corresponding to multiple adjustment items of the disc ball-making device;

[0030] For each second processing parameter array, the data in the second processing parameter array, the first distribution data, the powder characteristic data, and the first moisture content are substituted into the green pellet processing model;

[0031] The predicted values ​​of green bulb uniformity, green bulb moisture content, and diameter growth rate obtained each time are used to construct the model output array;

[0032] For each second processing parameter array, the data in the model output array are multiplied correspondingly with the data in the first coefficient array, and the resulting products are summed to obtain the second processing parameter score for the second processing parameter array.

[0033] If the iteration threshold is not reached, the multiple second processing parameter arrays are adjusted according to the scores of multiple second processing parameters, and the process jumps to the step of substituting the data in the second processing parameter array, the first distribution data, the powder characteristic data, and the first moisture content into the green ball processing model for each second processing parameter array.

[0034] Otherwise, the second processing parameter array with the highest score will be used as the first processing parameter array.

[0035] In one possible implementation, adjusting the array of multiple second processing parameters based on multiple second processing parameter scores includes:

[0036] Obtain multiple score queues, where each score queue corresponds to a second processing parameter array;

[0037] Add the score of the second processing parameter in each second processing parameter array to the score queue;

[0038] The historical second processing parameter array corresponding to the highest score in each score queue is used as the process optimal array;

[0039] The second processing parameter array with the largest second processing parameter score among the plurality of second processing parameter arrays is taken as the current optimal array;

[0040] For each second processing parameter array, adjustments are made based on the current optimal array and the corresponding process optimal array.

[0041] In one possible implementation, the adjustment for each second processing parameter array, based on the current optimal array and the corresponding process optimal array, includes:

[0042] For each second processing parameter array, adjustments are made according to the second formula, the current optimal array, and the corresponding process optimal array, wherein the second formula is:

[0043]

[0044] In the formula, For the adjusted second processing parameter array, the first One parameter, The second processing parameter array before adjustment One parameter, The first adjustment factor is... The first of the optimal arrays for the process One parameter, This is the second adjustment factor. The th of the current optimal array One parameter.

[0045] In one possible implementation, the green pellet processing model is constructed based on multiple historical processing datasets, including:

[0046] Multiple historical processing parameter datasets and multiple historical processing status datasets are obtained. Each historical processing parameter dataset corresponds to a historical processing status dataset. The historical processing parameter dataset includes: green pellet distribution data, powder characteristic data, green pellet moisture content data, and processing parameters corresponding to multiple adjustment items of the disc pelletizing equipment. The historical processing status dataset includes: green pellet diameter growth rate, green pellet moisture content, and green pellet uniformity.

[0047] The multiple historical processing parameter datasets are iteratively input into the model constructed according to the third formula to obtain multiple construction process output datasets, wherein the third formula is:

[0048]

[0049] In the formula, For transfer functions, It is a natural constant. For the first Line number The output of the column nodes, For the first The first node of the first row and first column One coefficient, For the first One input variable, The total number of input variables. This represents the total number of rows in the intermediate nodes. For the first Line number The first column node One coefficient, For the third formula One output, For the first The output node of the first One coefficient, The total number of columns in the intermediate nodes. This is the bias coefficient;

[0050] The output loss of the third formula is determined based on the multiple construction process output datasets and the multiple historical processing state datasets.

[0051] If the output loss of the third formula is greater than the loss threshold, then according to the output loss of the third formula, the gradient descent method is used to adjust multiple coefficients of the third formula, and the process jumps to the step of iteratively inputting the multiple historical processing parameter datasets into the model constructed according to the third formula to obtain multiple construction process output datasets.

[0052] Otherwise, the third formula shall be used as the green ball processing model.

[0053] In a second aspect, embodiments of the present invention provide a green pellet processing control device for implementing the green pellet processing control method as described in the first aspect or any possible implementation thereof, the green pellet processing control device comprising:

[0054] The green pellet status data acquisition module is used to acquire the first distribution data of green pellet diameter, powder characteristic data, and the first moisture content of green pellets;

[0055] The adjustment intensity equalization module is used to determine a first coefficient array characterizing the adjustment intensity of multiple target items based on the first distribution data and the first moisture content, wherein the multiple target items include: green ball uniformity, green ball moisture content and diameter growth rate.

[0056] The processing parameter optimization module is used to substitute the first distribution data, the powder characteristic data, and the first moisture content into the green pellet processing model, obtain a processing parameter score based on the multiple outputs and the first coefficient array, and optimize the processing parameters based on the processing parameter score to obtain a first processing parameter array.

[0057] as well as,

[0058] The green pellet processing parameter adjustment module is used to adjust multiple adjustment items of the disc pelletizing equipment according to the first processing parameter array.

[0059] Thirdly, embodiments of the present invention provide an electronic device, including a memory and a processor, wherein the memory stores a computer program executable on the processor, and the processor executes the computer program to implement the steps of the method as described in the first aspect or any possible implementation of the first aspect.

[0060] Fourthly, embodiments of the present invention provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the method as described in the first aspect or any possible implementation thereof.

[0061] The beneficial effects of the embodiments of the present invention compared with the prior art are as follows:

[0062] This invention discloses a green pellet processing control method. First, it acquires first distribution data of green pellet diameter, powder characteristic data, and first moisture content of the green pellets. Then, based on the first distribution data and the first moisture content, it determines a first coefficient array representing the adjustment intensity of multiple target items, including green pellet uniformity, green pellet moisture content, and diameter growth rate. Next, it substitutes the first distribution data, the powder characteristic data, and the first moisture content into a green pellet processing model. Based on the obtained multiple outputs and the first coefficient array, it obtains processing parameter scores and optimizes the processing parameters based on the scores to obtain a first processing parameter array. Finally, it adjusts multiple adjustment items of the disc pelletizing equipment according to the first processing parameter array. This invention sets coefficients for multiple target items based on the deviation between the current state and the target state. Based on the coefficients of multiple target items, it uses a balanced adjustment control strategy through model prediction and evaluation, weighted scoring and ranking, and iterative optimization convergence, taking into account multiple processing indicators during green pellet processing, thus ensuring green pellet quality and production efficiency. Attached Figure Description

[0063] To more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the 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.

[0064] Figure 1 This is a flowchart of the green pellet processing control method provided in the embodiments of the present invention;

[0065] Figure 2 This is a functional block diagram of the green pellet processing control device provided in an embodiment of the present invention;

[0066] Figure 3 This is a functional block diagram of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0067] In the following description, specific details such as particular system structures and techniques are set forth for illustrative purposes and not for limitation, so as to provide a thorough understanding of embodiments of the invention. However, those skilled in the art will understand that the invention can be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, and methods are omitted so as not to obscure the description of the invention with unnecessary detail.

[0068] To make the objectives, technical solutions, and advantages of the present invention clearer, specific embodiments will be described below in conjunction with the accompanying drawings.

[0069] The embodiments of the present invention will be described in detail below. This example is implemented based on the technical solution of the present invention, and provides detailed implementation methods and specific operation processes. However, the protection scope of the present invention is not limited to the following embodiments.

[0070] Figure 1 A flowchart of a green pellet processing control method provided for an embodiment of the present invention.

[0071] like Figure 1 As shown, a flowchart illustrating the implementation of the green pellet processing control method provided by an embodiment of the present invention is presented, and is described in detail below:

[0072] In step 101, the first distribution data of green pellet diameter, powder characteristic data, and the first moisture content of green pellet are obtained.

[0073] In step 102, based on the first distribution data and the first moisture content, a first coefficient array characterizing the adjustment intensity of multiple target items is determined, wherein the multiple target items include: green bulb uniformity, green bulb moisture content, and diameter growth rate.

[0074] In some implementations, determining a first coefficient array characterizing the adjustment strength of multiple target items based on the first distribution data and the first moisture content includes:

[0075] Acquire the first distribution data, the standard deviation of the target green bulb diameter, the moisture content of the target green bulb, and the diameter growth conversion factor, wherein the first distribution data includes the mean diameter of the first green bulb and the standard deviation of the first green bulb diameter;

[0076] The diameter growth rate is determined based on the average diameter of the first green bulb and the average diameter of the second green bulb, wherein the time point corresponding to the average diameter of the second green bulb is earlier than that of the average diameter of the first green bulb.

[0077] Take the square root of the mean diameter of the first green bulb to obtain the first square root value;

[0078] The product of the first square root value and the diameter growth conversion factor is taken as the target diameter growth rate;

[0079] Calculate the difference between the target diameter growth rate and the diameter growth rate, and use it as the first growth rate difference;

[0080] Calculate the difference between the target green bulb moisture content and the first moisture content, and use it as the first moisture content difference;

[0081] The difference between the standard deviation of the target green bulb diameter and the standard deviation of the first green bulb diameter is calculated as the first green bulb uniformity difference;

[0082] Based on the first growth rate difference, the first moisture content difference, and the first bulb uniformity difference, a first coefficient array is constructed, wherein the first coefficient array includes: a coefficient characterizing the adjustment of diameter growth rate, a coefficient characterizing the adjustment of moisture content, and a coefficient characterizing the adjustment of bulb uniformity.

[0083] In some embodiments, constructing a first coefficient array based on the first growth rate difference, the first moisture content difference, and the first bulb uniformity difference includes:

[0084] Obtain the first ratio value and the second ratio value;

[0085] The growth rate difference, the moisture content difference, and the bulb uniformity difference are respectively added to the growth rate difference queue, the moisture content difference queue, and the bulb uniformity difference queue;

[0086] Based on the first formula, the growth rate difference queue, the moisture content difference queue, and the green bulb uniformity difference queue, the second growth rate difference coefficient, the second moisture content difference coefficient, and the second green bulb uniformity difference coefficient are determined respectively, wherein the first formula is:

[0087]

[0088] In the formula, This refers to the second growth rate difference coefficient, the second moisture content difference coefficient, or the second bulb uniformity difference coefficient. The first proportional value, The first in the growth rate difference queue, the moisture content difference queue, or the bulb uniformity difference queue One data point, This is the second proportional value. The total number of data points in the growth rate difference queue, moisture content difference queue, or green bulb uniformity difference queue;

[0089] The second growth rate difference coefficient, the second moisture content difference coefficient, and the second green ball uniformity difference coefficient are normalized, and the results are used to construct the first coefficient array.

[0090] For example, in the industrial production process of green pellet preparation, precise control of the core quality parameters of green pellets (uniformity, moisture content, and diameter growth rate) is a key link to ensure the stability of subsequent roasting processes and the quality of the finished ore. The following will elaborate on the process of obtaining and determining the green pellet adjustment coefficient array, clarifying the key points of operation, data meaning, and calculation logic of each step.

[0091] The basic data acquisition phase serves as the data source foundation for the entire adjustment logic. Its core objective is to obtain key quantitative indicators reflecting the current state of green pellet preparation, providing accurate and reliable input parameters for subsequent calculations of the adjustment coefficients. Specific data acquisition content and explanations are as follows:

[0092] The first distribution data of green bulb diameter: This data is the core indicator characterizing the dispersion and central tendency of the current green bulb diameter, and needs to be obtained through sampling and testing of a batch of green bulbs using statistical methods. It typically includes the "mean of the first green bulb diameter" (reflecting the overall central tendency of the current green bulb diameter) and the "standard deviation of the first green bulb diameter" (reflecting the degree of deviation of the current green bulb diameter from the mean, directly corresponding to the uniformity of the green bulbs). During collection, it is necessary to ensure the randomness and representativeness of the sampling. Generally, the sample size per batch should not be less than 50 bulbs, and it should cover green bulbs from different areas of the pelletizing machine to avoid data bias caused by local sampling.

[0093] Powder property data: As the raw material for green pellet preparation, the properties of the powder directly affect the molding quality and stability of the green pellets. These mainly include parameters such as particle size distribution (e.g., -200 mesh content), specific surface area, hydrophilicity, and mud content. Although these data do not directly participate in the calculation of the first coefficient array, they serve as important references for subsequent adjustment measures (for example, powders with high specific surface area require more precise moisture content control). Powder property data can be obtained through testing with professional equipment such as laser particle size analyzers and specific surface area analyzers, and must be retested when raw material batches are changed to ensure data timeliness.

[0094] The first moisture content of green pellets refers to the mass percentage of water in the currently sampled green pellets, and is a key factor affecting the strength, molding rate and diameter growth rate of green pellets.

[0095] The core task of determining the adjustment coefficient array is to determine the "first coefficient array" based on the basic data collected in step 101 and the target parameters of the production process through quantitative calculation. Each coefficient in this array represents the adjustment priority and intensity of the three target items: "green pellet uniformity, green pellet moisture content, and diameter growth rate". The larger the absolute value of the coefficient, the greater the deviation between the corresponding target item and the process requirements, and the more priority should be given to adjustment.

[0096] The core logic of coefficient array calculation

[0097] The determination of the first coefficient array essentially involves comparing the deviations between the "current production status parameters" and the "process target parameters," and combining this with the cumulative effect of historical data to quantify the adjustment needs of each target item. Its calculation process must follow the logic of "clear target parameters—deviation calculation—historical data weighting—normalization processing" to ensure the scientific validity and practicality of the coefficients.

[0098] Detailed calculation steps and instructions

[0099] In the specific implementation process, the operation of "determining the first coefficient array based on the first distribution data and the first moisture content" needs to be carried out step by step according to the following sub-steps. Each step must strictly follow the data logic to avoid errors caused by human intervention.

[0100] Define the core input parameters

[0101] This step requires integrating three types of input: "current testing data," "process target data," and "process experience parameters," specifically including:

[0102] The collected first distribution data (including the mean diameter of the first green pellets D1 and the standard deviation of the diameter of the first green pellets S1) and the first moisture content W1; process target parameters: target green pellet diameter standard deviation S0 (reflecting the ideal requirement for green pellet uniformity, determined by the feed port diameter tolerance of the subsequent roasting equipment), target green pellet moisture content W0 (the optimal moisture content preset according to the powder characteristics, ensuring that the green pellets have sufficient strength and are not prone to sticking). Process experience parameters: diameter growth conversion factor K (determined by equipment parameters such as pelletizer type, speed, and feed rate, as well as powder characteristics, obtained through fitting long-term production data; for example, the K value of a cylindrical pelletizer is usually between 0.8 and 1.2).

[0103] Calculate the diameter growth rate

[0104] The diameter growth rate V is a key indicator reflecting the efficiency of green pellet forming. Its calculation requires a comparison of the average green pellet diameter at different time points. The specific logic is as follows:

[0105] Select the "second average bulb diameter D2" (the time interval Δt between D2 and D1 needs to be fixed, usually 5-10 minutes, to ensure data comparability) earlier than the detection time point of the first average bulb diameter D1, and calculate the diameter growth rate V using the following formula:

[0106]

[0107] In the formula, the unit of V is usually mm / min. If the calculation result is positive, it means that the diameter of the green pellets is increasing. If it is negative or zero, it is necessary to check whether the feed rate, pelletizer speed and other parameters are abnormal.

[0108] Determine the target diameter growth rate

[0109] The target diameter growth rate V0 is an ideal rate determined by combining the final diameter requirement of the green ball with the forming efficiency. Its calculation needs to consider the non-linear growth characteristic of the average green ball diameter—the growth rate of the green ball diameter gradually slows down as the diameter increases. Therefore, the "first square root value" is introduced to correct D1. The specific formula is as follows:

[0110] First, take the square root of the mean diameter D1 of the first green bulb to obtain the first square root value. ;

[0111] The target diameter growth rate V0 is then calculated using the following formula:

[0112]

[0113] In the formula, K is the diameter growth conversion factor. Its value needs to be dynamically adjusted according to the fitting curve of "diameter-growth rate" in actual production to ensure that V0 meets the equipment capacity requirements and does not cause the green pellet structure to become loose due to excessive growth.

[0114] Calculate the deviation value of each target item.

[0115] Deviation value is a core indicator for quantifying the difference between the current state and the target state. It needs to be calculated separately for diameter growth rate, moisture content, and bulb uniformity, specifically including:

[0116] The first growth rate difference ΔV reflects the difference between the current growth rate of the growing bulb diameter and the ideal growth rate.

[0117] First moisture content difference ΔW: reflects the deviation between the current green bulb moisture content and the optimal value.

[0118] The first green ball uniformity difference ΔS: reflects the difference in green ball uniformity through the deviation of the standard deviation of diameter.

[0119] Construct a weighted deviation coefficient by combining historical data

[0120] To avoid misjudgments in adjustment caused by fluctuations in data at a single point in time, historical deviation data needs to be introduced for weighted calculation to ensure the stability of the adjustment coefficient. This step achieves the fusion of historical data through "queue storage + weighting formula," and the specific operation is as follows:

[0121] Parameter preparation: Obtain the first and second proportional values ​​'a' and 'b' of the process preset (a and b are both decimals between 0 and 1, and satisfy a + b ≤ 1, where a represents the cumulative weight of historical data and b represents the immediate weight of the latest data. For example, a = 0.3 and b = 0.7 means that more emphasis is placed on the changing trend of the latest data).

[0122] Queue storage: The currently calculated first growth rate difference ΔV, first moisture content difference ΔW, and first green bulb uniformity difference ΔS are added to the corresponding "growth rate difference queue QD_V", "moisture content difference queue QD_W", and "green bulb uniformity difference queue QD_S", respectively. Each queue follows the "first-in, first-out" principle. When the number of data in a queue reaches the preset maximum value qnmax (usually 5-10 groups, determined by the production rhythm), the earliest group of data is deleted to ensure that the queue data reflects the recent production status.

[0123] Weighted calculation: Use the first formula to calculate the "second deviation coefficient" (i.e., the weighted deviation coefficient after integrating historical data) for each of the three target items:

[0124]

[0125] In the formula, is the second growth rate difference coefficient, the second moisture content difference coefficient, or the second green pellet uniformity difference coefficient (the second growth rate difference coefficient , the second moisture content difference coefficient , or the second green pellet uniformity difference coefficient ), is the first proportional value, is the th data in the growth rate difference queue, the moisture content difference queue, or the green pellet uniformity difference queue, is the second proportional value, is the total number of data in the growth rate difference queue, the moisture content difference queue, or the green pellet uniformity difference queue.

[0126] Normalization processing to construct the first coefficient array

[0127] Since , , have different dimensions and numerical ranges, directly using them as adjustment coefficients will lead to unbalanced adjustment efforts. Therefore, normalization processing is required - converting each coefficient into a value between 0 and 1, and the sum of the three is 1, ensuring that the coefficients can directly reflect the adjustment priorities of each target item. The specific operations are as follows:

[0128] Calculate the sum of the absolute values of the three second deviation coefficients :

[0129]

[0130] Calculate the normalized value of each coefficient respectively as an element of the first coefficient array:

[0131] The coefficient for adjusting the diameter growth rate ;

[0132] The coefficient for adjusting the moisture content ;

[0133] The coefficient for adjusting the green pellet uniformity 。

[0134] The finally constructed first coefficient array is The larger the value of a certain element in the array, the more urgent the need to adjust the corresponding target item. For example, if the array is [0.2, 0.5, 0.3], the priority should be to adjust the moisture content of the green bulbs, followed by the uniformity of the green bulbs, and finally the diameter growth rate.

[0135] In step 103, the first distribution data, the powder characteristic data, and the first moisture content are substituted into the green pellet processing model. Based on the multiple outputs and the first coefficient array, a processing parameter score is obtained, and the processing parameters are optimized based on the processing parameter score to obtain a first processing parameter array.

[0136] In some embodiments, substituting the first distribution data, the powder characteristic data, and the first moisture content into the green pellet processing model, obtaining a processing parameter score based on the obtained multiple outputs and the first coefficient array, and optimizing the processing parameters based on the processing parameter score to obtain a first processing parameter array includes:

[0137] Obtain multiple second processing parameter arrays, wherein each second processing parameter array includes processing parameters corresponding to multiple adjustment items of the disc ball-making device;

[0138] For each second processing parameter array, the data in the second processing parameter array, the first distribution data, the powder characteristic data, and the first moisture content are substituted into the green pellet processing model;

[0139] The predicted values ​​of green bulb uniformity, green bulb moisture content, and diameter growth rate obtained each time are used to construct the model output array;

[0140] For each second processing parameter array, the data in the model output array are multiplied correspondingly with the data in the first coefficient array, and the resulting products are summed to obtain the second processing parameter score for the second processing parameter array.

[0141] If the iteration threshold is not reached, the multiple second processing parameter arrays are adjusted according to the scores of multiple second processing parameters, and the process jumps to the step of substituting the data in the second processing parameter array, the first distribution data, the powder characteristic data, and the first moisture content into the green ball processing model for each second processing parameter array.

[0142] Otherwise, the second processing parameter array with the highest score will be used as the first processing parameter array.

[0143] In some implementations, adjusting the array of multiple second processing parameters based on the scores of multiple second processing parameters includes:

[0144] Obtain multiple score queues, where each score queue corresponds to a second processing parameter array;

[0145] Add the score of the second processing parameter in each second processing parameter array to the score queue;

[0146] The historical second processing parameter array corresponding to the highest score in each score queue is used as the process optimal array;

[0147] The second processing parameter array with the largest second processing parameter score among the plurality of second processing parameter arrays is taken as the current optimal array;

[0148] For each second processing parameter array, adjustments are made based on the current optimal array and the corresponding process optimal array.

[0149] In some implementations, the adjustment for each second processing parameter array, based on the current optimal array and the corresponding process optimal array, includes:

[0150] For each second processing parameter array, adjustments are made according to the second formula, the current optimal array, and the corresponding process optimal array, wherein the second formula is:

[0151]

[0152] In the formula, For the adjusted second processing parameter array, the first One parameter, The second processing parameter array before adjustment One parameter, The first adjustment factor is... The first of the optimal arrays for the process One parameter, This is the second adjustment factor. The th of the current optimal array One parameter.

[0153] In some implementations, the green pellet processing model is constructed based on multiple historical processing datasets, including:

[0154] Multiple historical processing parameter datasets and multiple historical processing status datasets are obtained. Each historical processing parameter dataset corresponds to a historical processing status dataset. The historical processing parameter dataset includes: green pellet distribution data, powder characteristic data, green pellet moisture content data, and processing parameters corresponding to multiple adjustment items of the disc pelletizing equipment. The historical processing status dataset includes: green pellet diameter growth rate, green pellet moisture content, and green pellet uniformity.

[0155] The multiple historical processing parameter datasets are iteratively input into the model constructed according to the third formula to obtain multiple construction process output datasets, wherein the third formula is:

[0156]

[0157] In the formula, For transfer functions, It is a natural constant. For the first Line number The output of the column nodes, For the first The first node of the first row and first column One coefficient, For the first One input variable, The total number of input variables. This represents the total number of rows in the intermediate nodes. For the first Line number The first column node One coefficient, For the third formula One output, For the first The output node of the first One coefficient, The total number of columns in the intermediate nodes. This is the bias coefficient;

[0158] The output loss of the third formula is determined based on the multiple construction process output datasets and the multiple historical processing state datasets.

[0159] If the output loss of the third formula is greater than the loss threshold, then according to the output loss of the third formula, the gradient descent method is used to adjust multiple coefficients of the third formula, and the process jumps to the step of iteratively inputting the multiple historical processing parameter datasets into the model constructed according to the third formula to obtain multiple construction process output datasets.

[0160] Otherwise, the third formula shall be used as the green ball processing model.

[0161] In step 104, multiple adjustment items of the disc ball-making equipment are adjusted according to the first processing parameter array.

[0162] For example, step 103 is the core execution link of the green pellet preparation control process. Its core objective is to combine the basic data obtained in the previous stage with the adjustment coefficient, and to achieve quantitative optimization of processing parameters through the predictive ability of the green pellet processing model.

[0163] Specifically, this step establishes a mapping relationship of "input data - predicted output - parameter score" through the green pellet processing model. Using the first coefficient array as a weight guide, it selects the optimal combination of processing parameters for improving the core quality of green pellets, and finally forms the first processing parameter array that can directly guide production.

[0164] The core logic of parameter optimization

[0165] The optimization logic of step 103 can be summarized as "multiple parameter trial calculations - model prediction and evaluation - weighted scoring and ranking - iterative optimization convergence". Among them, the green pellet processing model is the core tool for realizing the "parameter-quality" prediction, and the first coefficient array provides "priority weights" for the scoring - ensuring that the scoring results are tilted towards the core requirements of the three target items of green pellet uniformity, moisture content, and diameter growth rate, and avoiding imbalance of key quality indicators caused by indiscriminate optimization.

[0166] Detailed optimization steps and instructions

[0167] The process of "optimizing parameters through the green pellet processing model" must be executed step by step according to a strict procedure to ensure both the comprehensiveness of parameter calculations and the stability of optimization results through an iterative mechanism. The specific operation is as follows:

[0168] Construct multiple sets of second processing parameter arrays

[0169] The second processing parameter array is a "candidate parameter set" used for model trial calculations. Its construction must revolve around the core adjustment terms of the disc-shaped pelletizing equipment, ensuring coverage of all key adjustable dimensions of the equipment. Specific requirements include:

[0170] Each set of second processing parameter arrays corresponds to a complete set of adjustment items for the disc pelletizer. Typical adjustment items include: feed rate, water spray rate, disc rotation speed, disc tilt angle, scraper height, etc. The array dimensions and the number of adjustment items are perfectly matched. The values ​​of the second processing parameter arrays must be limited to the safe operating range of the equipment.

[0171] To ensure the comprehensiveness of the trial calculation, the second processing parameter array must include three categories: "baseline parameter group", "boundary parameter group" and "random parameter group". The baseline parameter group is based on the current production parameter settings, the boundary parameter group takes the upper and lower limits of the equipment adjustment range, and the random parameter group is randomly generated within the range. The total number is usually no less than 20 groups to cover the main parameter combinations.

[0172] Model input and prediction output

[0173] This step, which uses the green pellet processing model to transform "input data → quality prediction," is the core of parameter evaluation. Specific operations must meet the requirements of data integrity and matching.

[0174] Input data integration: For each set of second processing parameter arrays, it needs to be integrated with the "first distribution data, powder characteristic data (e.g., particle size distribution, specific surface area, etc.), and first moisture content" obtained in step 101 to form a complete model input dataset. Among them, the powder characteristic data needs to be standardized (e.g., converting the -200 mesh content into a normalized value of 0-1) to ensure consistency with the model input requirements.

[0175] Model computation: The integrated input dataset is fed into the green bulb processing model one by one. The model calculates through its internal algorithm and outputs the predicted green bulb quality under the corresponding parameter group.

[0176] Output array construction: The predicted values ​​of "uniformity of green bulbs, moisture content of green bulbs, and diameter growth rate" obtained from each prediction are constructed into a model output array in a fixed order. The array dimensions are completely matched with the first coefficient array, providing a basis for subsequent weighted scoring.

[0177] Weight-based parameter scoring calculation

[0178] The core function of parameter scoring is to quantify the overall improvement effect of each set of second processing parameters on the quality of green bulbs. Its calculation uses the first coefficient set as weights, highlighting the adjustment needs of core target items. The specific method is as follows:

[0179] Let the first coefficient array be (These correspond to the adjustment weights for diameter growth rate, moisture content, and uniformity, respectively). The model output array corresponding to a certain second processing parameter array is... The score for the second processing parameter of the array is calculated using the following formula:

[0180]

[0181] In the formula The reason for taking the reciprocal is that the uniformity of green balls is negatively correlated with the standard deviation (the smaller the standard deviation, the better the uniformity). Taking the reciprocal transforms "improved uniformity" into "increased numerical value," which is consistent with... , The optimization direction should be consistent to ensure a unified scoring logic. The larger the P-value, the better the overall optimization effect of that parameter group on the quality of the green bulbs.

[0182] Iterative optimization and convergence judgment

[0183] To avoid the limitations of single-round parameter calculations, an iterative mechanism is needed to gradually approach the optimal parameter combination. The specific process is as follows:

[0184] Iteration termination condition judgment: preset iteration number threshold (usually 10-20 times, which can be adjusted according to the production rhythm; if the rhythm is fast, the threshold should be small, such as 8 times; if the accuracy requirement is high, the threshold should be large, such as 20 times). If the current iteration number has not reached the threshold, the parameter adjustment stage is entered; if the threshold has been reached, the iteration is terminated.

[0185] Parameter adjustment: Based on the scoring results of all second processing parameter arrays in this round, retain the core characteristics of high-scoring parameter groups (such as high-scoring groups usually have the characteristics of "medium-speed feeding and precise water spraying"), and make targeted modifications to low-scoring parameter groups (such as deleting parameter groups where the water spray volume deviates in the opposite direction to the predicted moisture content value), and generate a new batch of second processing parameter arrays.

[0186] Looping operation: Substitute the adjusted second processing parameter array back into the model for prediction and scoring, repeat the "prediction-scoring-adjustment" process until the threshold of the number of iterations is reached;

[0187] Optimal parameter determination: After the iteration terminates, the second processing parameter array with the highest score among all rounds is selected as the final first processing parameter array, which can be directly sent to the control system of the disc ball-making equipment.

[0188] Detailed implementation method for parameter adjustment

[0189] During the iteration process, "adjusting the second processing parameter array based on the score" is key to ensuring optimization efficiency. Its core is to avoid blind parameter adjustments by using a two-dimensional reference of "historical best + current best." The specific operation is as follows:

[0190] Build a score queue to store historical data

[0191] To track the historical performance of each group of second processing parameter arrays, a separate score queue needs to be created for each group:

[0192] The score queue corresponds one-to-one with the second processing parameter array, and the queue stores the score of the second processing parameter in each iteration of the parameter group.

[0193] The score queue adopts a "first-in, first-out" constraint mechanism. When the queue length reaches a preset value (usually 5 rounds, i.e., retaining the 5 most recent scores), the earliest score data is deleted to ensure that the queue reflects the recent optimization trend of the parameter group.

[0194] Determine the optimal parameter array in two dimensions

[0195] Based on the score queue and the current round score, two types of optimal parameter arrays are selected as adjustment benchmarks:

[0196] Process-optimal array: For each group of second processing parameter arrays, find the record with the highest score in its corresponding score queue, and use the historical parameter group corresponding to that record as the "process-optimal array" of that array (denoted as...). This reflects the historical best state of the parameter set;

[0197] Current optimal array: Among all the second processing parameter arrays in the current round, the parameter set with the highest score is selected as the "current optimal array" (denoted as ). This reflects the global optimal state in the current iteration round.

[0198] Parameter tuning based on double-optimal arrays

[0199] Parameter adjustment employs a fusion strategy of "historical experience + current trends," achieving quantitative adjustment through a second formula. This ensures that the adjusted parameters both inherit their inherent advantages and approach the global optimum. The second formula is as follows:

[0200]

[0201] In the formula, For the adjusted second processing parameter array, the first One parameter, The second processing parameter array before adjustment One parameter, The first adjustment factor is... The first of the optimal arrays for the process One parameter, This is the second adjustment factor. The th of the current optimal array One parameter.

[0202] For example, the water spray volume parameters of a certain parameter group The optimal array corresponds to the parameters in the process. The parameters corresponding to the current optimal array ,Pick , Then adjust the water spray volume It retains its historical advantages while moving closer to the current optimal state.

[0203] Construction method of green pellet processing model

[0204] The green bulb processing model is the core tool of step 103. Essentially, it's a predictive model trained on historical production data, capable of accurately mapping the non-linear relationship between "input data" and "green bulb quality." This model typically employs a neural network architecture (as can be determined from the transfer function characteristics of the third formula), and the specific construction process is as follows:

[0205] Historical dataset collection and preprocessing

[0206] Data sets are the foundation of model building and must meet the requirements of "data completeness, accurate correspondence, and comprehensive coverage":

[0207] Data collection: Collect at least 3 months of continuous production data to construct two major sets: "historical processing parameter dataset" and "historical processing status dataset". The two sets of data must correspond one-to-one according to the production timestamp.

[0208] Dataset composition:

[0209] Historical processing parameter dataset (input data): includes green pellet distribution data (historical mean diameter, standard deviation), powder characteristic data (particle size distribution, specific surface area, etc.), green pellet moisture content data, and adjustment parameters of the disc pelletizing equipment (feed rate, water spray rate, etc.).

[0210] Historical processing status dataset (output labels): includes the green bulb diameter growth rate, actual moisture content, and actual uniformity (expressed as diameter standard deviation) for the corresponding timestamp.

[0211] Data preprocessing: Remove data under abnormal operating conditions such as equipment failure and raw material change; standardize data of different dimensions (e.g., convert feed rate into a normalized value of 0-1); divide the data into training set, validation set and test set in a ratio of 7:2:1.

[0212] Model structure and formula definition

[0213] The model employs a multi-layer network structure, using the third formula to achieve signal transmission from the input layer to the intermediate layer and then to the output layer. This third formula is the core computational logic of the model, and its overall expression and the meanings of its sub-formulas are as follows:

[0214]

[0215] In the formula, For transfer functions, It is a natural constant. For the first Line number The output of the column nodes, For the first The first node of the first row and first column One coefficient, For the first One input variable, The total number of input variables. This represents the total number of rows in the intermediate nodes. For the first Line number The first column node One coefficient, For the third formula One output, For the first The output node of the first One coefficient, The total number of columns in the intermediate nodes. This is the bias coefficient.

[0216] Model training and convergence verification

[0217] The model converges through a cycle of "training-loss calculation-weight adjustment," the specific process of which is as follows:

[0218] Model training: The training set data is fed into the model built based on the third formula one by one, and the output dataset of the construction process is obtained through calculation logic;

[0219] Loss Calculation: The mean squared error (MSE) is used to calculate the output loss between the output dataset of the construction process and the historical processing state dataset corresponding to the training set. The formula is as follows: ( For the sample size, (This is the actual output value).

[0220] Weight adjustment: If the output loss is greater than the preset loss threshold (usually set to 0.01, adjusted according to accuracy requirements), the gradient descent method (such as the Adam optimizer) is used to adjust the connection weights and bias coefficients in the direction of backpropagation of loss to reduce the model prediction error.

[0221] Iterative validation: Substitute the adjusted model back into the training set for calculation, repeat the "training-loss calculation-weight adjustment" steps until the output loss is less than the loss threshold;

[0222] Model determination: Use the validation set and test set to verify the generalization ability of the model (the prediction accuracy of the test set must be greater than 90%). If the requirement is met, the third formula with the current parameter configuration will be used as the final raw ball processing model; if not, the data preprocessing method or model structure needs to be re-optimized.

[0223] Key Explanation

[0224] The process parameters involved in this process need to be set specifically according to the specific green pellet preparation process, equipment model and powder characteristics, and continuously optimized through feedback from long-term production data to ensure the accuracy of the coefficient array. The timeliness and accuracy of data acquisition are the core prerequisites of this method. It is recommended to use automated detection equipment (such as online particle size analyzer and moisture content sensor) to realize real-time data acquisition and transmission, and reduce the error of manual detection.

[0225] When a sudden abnormality occurs during the preparation of green pellets (such as interruption of powder supply or equipment failure), the calculation of this process must be suspended, the abnormal situation should be dealt with first, and data acquisition and coefficient calculation should be restarted after production has returned to stability.

[0226] The green pellet processing model needs to establish a regular update mechanism (such as updating it monthly) to adjust the model coefficients based on new production data, so as to avoid a decrease in the model's prediction accuracy due to factors such as changes in raw material characteristics and equipment aging.

[0227] The present invention provides an implementation method for controlling green pellet processing. First, it acquires first distribution data of green pellet diameter, powder characteristic data, and first moisture content of the green pellets. Then, based on the first distribution data and the first moisture content, it determines a first coefficient array characterizing the adjustment intensity of multiple target items, including green pellet uniformity, green pellet moisture content, and diameter growth rate. Next, it substitutes the first distribution data, the powder characteristic data, and the first moisture content into a green pellet processing model. Based on the obtained multiple outputs and the first coefficient array, it obtains processing parameter scores and optimizes the processing parameters based on these scores to obtain a first processing parameter array. Finally, it adjusts multiple adjustment items of the disc pelletizing equipment according to the first processing parameter array. This invention sets coefficients for multiple target items based on the deviation between the current state and the target state. Based on these coefficients, it uses a balanced adjustment control strategy through model prediction and evaluation, weighted scoring and ranking, and iterative optimization convergence, taking into account multiple processing indicators during green pellet processing and ensuring both green pellet quality and production efficiency.

[0228] It should be understood that the sequence number of each step in the above embodiments 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.

[0229] The following are embodiments of the apparatus of the present invention. For details not described in detail, please refer to the corresponding method embodiments described above.

[0230] Figure 2 This is a functional block diagram of the green pellet processing control device provided in the embodiments of the present invention, with reference to... Figure 2 The green ball processing control device includes: a green ball status data acquisition module 201, an adjustment force balancing module 202, a processing parameter optimization module 203, and a green ball processing parameter adjustment module 204, wherein:

[0231] The green pellet status data acquisition module 201 is used to acquire the first distribution data of green pellet diameter, powder characteristic data, and the first moisture content of green pellets;

[0232] The equalization adjustment module 202 is used to determine a first coefficient array characterizing the adjustment intensity of multiple target items based on the first distribution data and the first moisture content, wherein the multiple target items include: uniformity of green bulbs, moisture content of green bulbs, and diameter growth rate.

[0233] The processing parameter optimization module 203 is used to substitute the first distribution data, the powder characteristic data and the first moisture content into the green pellet processing model, obtain a processing parameter score based on the multiple outputs and the first coefficient array, and optimize the processing parameters based on the processing parameter score to obtain a first processing parameter array.

[0234] The green pellet processing parameter adjustment module 204 is used to adjust multiple adjustment items of the disc pelletizing equipment according to the first processing parameter array.

[0235] Figure 3 This is a functional block diagram of the electronic device provided in an embodiment of the present invention. For example... Figure 3 As shown, the electronic device 3 of this embodiment includes a processor 300 and a memory 301, wherein the memory 301 stores a computer program 302 that can run on the processor 300. When the processor 300 executes the computer program 302, it implements the steps of the various green pellet processing control methods and embodiments described above, for example... Figure 1 Steps 101 to 104 are shown.

[0236] For example, the computer program 302 may be divided into one or more modules / units, which are stored in the memory 301 and executed by the processor 300 to complete the present invention.

[0237] The electronic device 3 can be a desktop computer, laptop, handheld computer, cloud server, or other computing device. The electronic device 3 may include, but is not limited to, a processor 300 and a memory 301. Those skilled in the art will understand that... Figure 3 This is merely an example of electronic device 3 and does not constitute a limitation on electronic device 3. It may include more or fewer components than shown, or combine certain components, or different components. For example, electronic device 3 may also include input / output devices, network access devices, buses, etc.

[0238] The processor 300 may be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor.

[0239] The memory 301 can be an internal storage unit of the electronic device 3, such as a hard disk or memory. The memory 301 can also be an external storage device of the electronic device 3, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc., equipped on the electronic device 3. Furthermore, the memory 301 can include both internal and external storage units of the electronic device 3. The memory 301 is used to store the computer program 302 and other programs and data required by the electronic device 3. The memory 301 can also be used to temporarily store data that has been output or will be output.

[0240] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is merely an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments 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. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this application. The specific working process of the units and modules in the above system can be referred to the corresponding process in the aforementioned method embodiments, and will not be repeated here.

[0241] In the above embodiments, the descriptions of each embodiment have their own emphasis. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0242] 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.

[0243] In the embodiments provided by this invention, it should be understood that the disclosed devices / electronic devices and methods can be implemented in other ways. For example, the device / electronic device embodiments described above are merely illustrative. For instance, the division of modules or 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 system, or some features may be ignored or not executed. Furthermore, the mutual 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.

[0244] 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, depending on actual needs.

[0245] Furthermore, 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. The integrated unit can be implemented in hardware or as a software functional unit.

[0246] If the integrated module / unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the above-described embodiments can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various methods and apparatus embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc.

[0247] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A green ball processing control method characterized by, include: Obtain the first distribution data of green pellet diameter, powder characteristic data, and the first moisture content of green pellets; Based on the first distribution data and the first moisture content, a first coefficient array characterizing the adjustment intensity of multiple target items is determined, wherein the multiple target items include: green bulb uniformity, green bulb moisture content, and diameter growth rate; The first distribution data, the powder characteristic data, and the first moisture content are substituted into the green pellet processing model. Based on the multiple outputs and the first coefficient array, a processing parameter score is obtained, and the processing parameters are optimized based on the processing parameter score to obtain a first processing parameter array. The multiple adjustment items of the disc pelletizing equipment are adjusted according to the first processing parameter array; The step of determining a first coefficient array characterizing the adjustment intensity of multiple target items based on the first distribution data and the first moisture content includes: Acquire the first distribution data, the standard deviation of the target green bulb diameter, the moisture content of the target green bulb, and the diameter growth conversion factor, wherein the first distribution data includes the mean diameter of the first green bulb and the standard deviation of the first green bulb diameter; The diameter growth rate is determined based on the average diameter of the first green bulb and the average diameter of the second green bulb, wherein the time point corresponding to the average diameter of the second green bulb is earlier than that of the average diameter of the first green bulb. Take the square root of the mean diameter of the first green bulb to obtain the first square root value; The product of the first square root value and the diameter growth conversion factor is taken as the target diameter growth rate; Calculate the difference between the target diameter growth rate and the diameter growth rate, and use it as the first growth rate difference; Calculate the difference between the target green bulb moisture content and the first moisture content, and use it as the first moisture content difference; The difference between the standard deviation of the target green bulb diameter and the standard deviation of the first green bulb diameter is calculated as the first green bulb uniformity difference; Based on the first growth rate difference, the first moisture content difference, and the first bulb uniformity difference, a first coefficient array is constructed, wherein the first coefficient array includes: a coefficient characterizing the adjustment of diameter growth rate, a coefficient characterizing the adjustment of moisture content, and a coefficient characterizing the adjustment of bulb uniformity. The step of constructing a first coefficient array based on the first growth rate difference, the first moisture content difference, and the first bulb uniformity difference includes: Obtain the first ratio value and the second ratio value; The first growth rate difference, the first moisture content difference, and the first green bulb uniformity difference are respectively added to the growth rate difference queue, the moisture content difference queue, and the green bulb uniformity difference queue; Based on the first formula, the growth rate difference queue, the moisture content difference queue, and the green bulb uniformity difference queue, the second growth rate difference coefficient, the second moisture content difference coefficient, and the second green bulb uniformity difference coefficient are determined respectively, wherein the first formula is: In the formula, This refers to the second growth rate difference coefficient, the second moisture content difference coefficient, or the second bulb uniformity difference coefficient. The first proportional value, The first in the growth rate difference queue, the moisture content difference queue, or the bulb uniformity difference queue One data point, This is the second proportional value. The total number of data points in the growth rate difference queue, moisture content difference queue, or green bulb uniformity difference queue; The second growth rate difference coefficient, the second moisture content difference coefficient, and the second green ball uniformity difference coefficient are normalized, and the results are used to construct the first coefficient array. Specifically, the first distribution data, the powder characteristic data, and the first moisture content are substituted into the green pellet processing model. Based on the obtained multiple outputs and the first coefficient array, a processing parameter score is obtained. The processing parameters are then optimized based on the processing parameter score to obtain a first processing parameter array, including: Obtain multiple second processing parameter arrays, wherein each second processing parameter array includes processing parameters corresponding to multiple adjustment items of the disc ball-making device; For each second processing parameter array, the data in the second processing parameter array, the first distribution data, the powder characteristic data, and the first moisture content are substituted into the green pellet processing model; The predicted values ​​of green bulb uniformity, green bulb moisture content, and diameter growth rate obtained each time are used to construct the model output array; For each second processing parameter array, the data in the model output array are multiplied correspondingly with the data in the first coefficient array, and the resulting products are summed to obtain the second processing parameter score for the second processing parameter array. If the iteration threshold is not reached, the multiple second processing parameter arrays are adjusted according to the scores of multiple second processing parameters, and the process jumps to the step of substituting the data in the second processing parameter array, the first distribution data, the powder characteristic data, and the first moisture content into the green ball processing model for each second processing parameter array. Otherwise, the second processing parameter array with the highest score will be used as the first processing parameter array; The green pellet processing model is constructed based on multiple historical processing datasets, including: Multiple historical processing parameter datasets and multiple historical processing status datasets are obtained. Each historical processing parameter dataset corresponds to a historical processing status dataset. The historical processing parameter dataset includes: green pellet distribution data, powder characteristic data, green pellet moisture content data, and processing parameters corresponding to multiple adjustment items of the disc pelletizing equipment. The historical processing status dataset includes: green pellet diameter growth rate, green pellet moisture content, and green pellet uniformity. The multiple historical processing parameter datasets are iteratively input into the model constructed according to the third formula to obtain multiple construction process output datasets, wherein the third formula is: In the formula, For transfer functions, It is a natural constant. For the first Line 1 The output of the column nodes, For the first The first node of the first row and first column One coefficient, For the first One input variable, The total number of input variables. This represents the total number of rows in the intermediate nodes. For the first Line 1 The first column node One coefficient, For the third formula One output, For the first The output node of the first One coefficient, The total number of columns for the intermediate nodes. This is the bias coefficient; The output loss of the third formula is determined based on the multiple construction process output datasets and the multiple historical processing state datasets. If the output loss of the third formula is greater than the loss threshold, then according to the output loss of the third formula, the gradient descent method is used to adjust multiple coefficients of the third formula, and the process jumps to the step of iteratively inputting the multiple historical processing parameter datasets into the model constructed according to the third formula to obtain multiple construction process output datasets. Otherwise, the third formula shall be used as the green ball processing model.

2. The green ball processing control method according to claim 1, characterized by, The step of adjusting the array of multiple second processing parameters based on the scores of multiple second processing parameters includes: Obtain multiple score queues, where each score queue corresponds to a second processing parameter array; Add the score of the second processing parameter in each second processing parameter array to the score queue; The historical second processing parameter array corresponding to the highest score in each score queue is used as the process optimal array; The second processing parameter array with the largest second processing parameter score among the plurality of second processing parameter arrays is taken as the current optimal array; For each second processing parameter array, adjustments are made based on the current optimal array and the corresponding process optimal array.

3. The green ball processing control method according to claim 2, characterized by, For each second processing parameter array, adjustments are made based on the current optimal array and the corresponding process optimal array, including: For each second processing parameter array, adjustments are made according to the second formula, the current optimal array, and the corresponding process optimal array, wherein the second formula is: In the formula, For the adjusted second processing parameter array, the first One parameter, The second processing parameter array before adjustment One parameter, The first adjustment factor is... The first of the optimal arrays for the process One parameter, This is the second adjustment factor. The th of the current optimal array One parameter.

4. A green ball processing control device characterized by comprising: For implementing the green pellet processing control method as described in any one of claims 1-3, the green pellet processing control device comprises: The green pellet status data acquisition module is used to acquire the first distribution data of green pellet diameter, powder characteristic data, and the first moisture content of green pellets; The adjustment intensity equalization module is used to determine a first coefficient array characterizing the adjustment intensity of multiple target items based on the first distribution data and the first moisture content, wherein the multiple target items include: green ball uniformity, green ball moisture content and diameter growth rate. The processing parameter optimization module is used to substitute the first distribution data, the powder characteristic data, and the first moisture content into the green pellet processing model, obtain a processing parameter score based on the multiple outputs and the first coefficient array, and optimize the processing parameters based on the processing parameter score to obtain a first processing parameter array. as well as, The green pellet processing parameter adjustment module is used to adjust multiple adjustment items of the disc pelletizing equipment according to the first processing parameter array.

5. An electronic device comprising a memory and a processor, said memory having stored therein a computer program operable on said processor, characterized in that, When the processor executes the computer program, it implements the steps of the method as described in any one of claims 1 to 3 above.

6. A computer-readable storage medium storing a computer program, the computer program comprising instructions that, when executed by a computer, cause the computer to perform the method of any one of claims 1 to 5. When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1 to 3 above.

Images