Method and device for determining operation index of coal mill and storage medium

Abstract

The application provides a coal mill operation index determination method, a coal mill operation index determination device and a computer readable storage medium. The coal mill operation index determination method comprises the following steps: obtaining a plurality of operating parameters of the coal mill. The plurality of operating parameters are selected from at least one of voltage, current and temperature; and inputting the plurality of operating parameters into a pre-trained first BP neural network model to determine the operation index of the coal mill. The operation index at least includes the unit load of the coal mill. The first BP neural network model is obtained based on training sample data, prediction value data and counterfactual explanation of the expected direction of the operation index of the XGB model. The application can be used to improve the prediction accuracy of nonlinear problems and the interpretability of the prediction results, and simplify the optimization process of the model, so as to improve the production efficiency of the mineral processing process.

Description

Technical Field

[0001] This invention relates to the field of mineral processing, and in particular to a method for determining the operating parameters of a coal mill, a device for determining the operating parameters of a coal mill, and a computer-readable storage medium. Background Technology

[0002] Mineral processing is a highly complex dynamic system involving multiple interconnected processes, multi-level coupling, and multi-temporal and spatial scale characteristics. To understand and control the status of these processing operations in real time, it is necessary to monitor a large number of operational indicators. Although advancements in sensor technology have enabled the real-time monitoring of most process indicators, some crucial indicators remain unavailable online and require model prediction. However, due to the inconsistent composition and quality of raw materials and frequent fluctuations in production conditions during coal mining, existing prediction models cannot effectively predict complex nonlinear relationships and thus struggle to match actual operating condition changes, preventing the maximization of production efficiency. Furthermore, existing prediction models are based on gradient optimization and have complex structures, resulting in complex optimization processes and high computational resource consumption.

[0003] In order to overcome the above-mentioned defects in the existing technology, there is an urgent need in the field for an improved method for determining the operating indicators of coal mills, which can improve the prediction accuracy and interpretability of nonlinear problems, and simplify the optimization process of the model, thereby enabling the adjustment of coal mill process parameters and optimization of coal mine processing operation to improve the production efficiency of mineral processing. Summary of the Invention

[0004] The following provides a brief overview of one or more aspects to offer a basic understanding of them. This overview is not an exhaustive summary of all conceived aspects, nor is it intended to identify key or decisive elements of all aspects, nor to define the scope of any or all aspects. Its sole purpose is to present some concepts of one or more aspects in a simplified form as a prelude to the more detailed descriptions that follow.

[0005] To overcome the aforementioned deficiencies in the existing technology, this invention provides a method for determining coal mill operating indicators, a device for determining coal mill operating indicators, and a computer-readable storage medium. It can replace the XGB model with a trained BP neural network model to improve the prediction accuracy and interpretability of nonlinear problems. Furthermore, by optimizing and training the prediction model based on counterfactual interpretation, it simplifies the model optimization process, thereby providing a reference for adjusting coal mill process parameters and optimizing coal mine processing operations, ultimately improving the production efficiency of mineral processing.

[0006] Specifically, the method for determining the operating indicators of a coal mill according to the first aspect of the present invention includes the following steps: acquiring multiple operating parameters of the coal mill. The multiple operating parameters are selected from at least one of voltage, current, and temperature; and inputting the multiple operating parameters into a pre-trained first BP neural network model to determine the operating indicators of the coal mill. The operating indicators include at least the unit load of the coal mill. The first BP neural network model is obtained by training on training sample data, predicted value data, and counterfactual interpretations of the expected directions corresponding to the operating indicators of the XGB model.

[0007] Further, in some embodiments of the present invention, the step of training the first BP neural network model includes: obtaining multiple sets of historical operating parameters X of the coal mill. Each set of historical operating parameters X includes the multiple operating parameters; sequentially inputting each set of historical operating parameters X into a pre-trained XGB model to obtain a corresponding first predicted value Y′; associating each set of historical operating parameters X with its corresponding first predicted value Y′ to construct a first training sample dataset; constructing an initial BP neural network model to be trained, and training the initial BP neural network model according to the first training sample dataset to obtain a second BP neural network model;

[0008] Determine the expected value Y″ of the aforementioned operating indicator:

[0009] Y″=Y′+ ε

[0010] Wherein, ε represents the desired direction of the operational metric; the historical operational parameters X from each group in the first training sample dataset are sequentially input into the second BP neural network model to obtain the second predicted value. And using the Adam optimizer, based on the second predicted value Loss between the expected value Y″ and the expected value Y″ The second BP neural network model is optimized to obtain the first BP neural network model.

[0011] Furthermore, in some embodiments of the present invention, the Adam optimizer is used based on the second predicted value. Loss between the expected value Y″ and the expected value Y″ The step of optimizing the second BP neural network model to obtain the first BP neural network model includes: based on the second predicted value Loss between the expected value Y″ and the expected value Y″ The historical operating parameters X of each group are adjusted to obtain the counterfactual explanation X′ of each group of historical operating parameters X; and the second BP neural network model is optimized based on the counterfactual explanation X′ to obtain the first BP neural network model.

[0012] Further, in some embodiments of the present invention, the step of training the initial BP neural network model based on the first training sample dataset to obtain a second BP neural network model includes: sequentially inputting each group of historical running parameters X from the first training sample dataset into the initial BP neural network model to obtain a third predicted value Y″′; determining a first model loss value of the initial BP neural network model based on the difference between the third predicted value Y″′ and the first predicted value Y′; adjusting the parameters of the initial BP neural network model and retraining the initial BP neural network model in response to the first model loss value being greater than a preset first loss threshold; and determining that the training of the initial BP neural network model is complete and obtaining the second BP neural network model in response to the first model loss value being less than or equal to the first loss threshold.

[0013] Furthermore, in some embodiments of the present invention, the step of training the XGB model includes: obtaining the true value Y of the operating index corresponding to each group of historical operating parameters X, and associating each group of historical operating parameters X with its corresponding true value Y of the operating index to construct a second training sample dataset; constructing the XGB model to be trained; and training the XGB model according to the second training sample dataset.

[0014] Furthermore, in some embodiments of the present invention, the step of constructing the XGB model to be trained includes: constructing the objective function of the XGB model:

[0015]

[0016] in, Let Ω(f) be the loss function for the i-th training sample. k The ) represents the complexity of the XGB model;

[0017] The loss function is expanded using a second-order Taylor series to optimize the objective function.

[0018]

[0019] Among them, g i Let h be the first derivative of the i-th training sample in the t-th learner. iThe second derivative of the first training sample in the t-th learner is given by the i-th sample; and the hyperparameters of the XGB model are searched using a random grid search method to minimize the objective function.

[0020] Further, in some embodiments of the present invention, the step of training the XGB model based on the second training sample dataset includes: preprocessing the second training sample dataset and dividing it into a training set and a test set; training the XGB model using the training set and testing the trained XGB model using the test set to determine a second model loss value of the XGB model; redefining the hyperparameters of the XGB model in response to the second model loss value being greater than a preset second loss threshold, and retraining the XGB model using the training set based on the redefined hyperparameters; and determining that the training of the XGB model is complete in response to the second model loss value being less than or equal to the second loss threshold.

[0021] Furthermore, in some embodiments of the present invention, the step of preprocessing the second training sample dataset includes: filtering the second training sample dataset; and normalizing the second training sample dataset.

[0022]

[0023] Where n is the number of first training samples in the second training sample dataset, and m is the number of variables in the first training samples.

[0024] Furthermore, in some embodiments of the present invention, the step of filtering the second training sample dataset includes: calculating the mutual information between the running parameters X and their corresponding true values ​​Y of the running indicators for each group.

[0025]

[0026] Where p(x) is the marginal probability density function of the operating parameter X. p(y) is the marginal probability density function of the operating index Y. p(x, y) is the joint probability density function of the operating parameter X and the operating index Y; and in response to any mutual information being less than a preset first threshold, the corresponding sample is deleted from the second training sample dataset.

[0027] Furthermore, the apparatus for determining the operating parameters of a coal mill according to the second aspect of the present invention includes a memory and a controller. The memory stores computer instructions. The controller is connected to the memory and configured to execute the computer instructions stored thereon to implement the method for determining the operating parameters of a coal mill as provided in the first aspect of the present invention.

[0028] Furthermore, the computer-readable storage medium provided in the third aspect of the present invention stores computer instructions thereon, characterized in that, when the computer instructions are executed by a processor, the method for determining counterfactual interpretations of coal mill parameters as provided in the first aspect of the present invention is implemented. Attached Figure Description

[0029] The above-described features and advantages of the present invention will be better understood after reading the following detailed description of embodiments of the present disclosure in conjunction with the accompanying drawings. In the drawings, components are not necessarily drawn to scale, and components having similar related characteristics or features may have the same or similar reference numerals.

[0030] Figure 1 A schematic diagram of the process for training a first BP neural network model according to some embodiments of the present invention is shown.

[0031] Figure 2 A line graph showing historical operational data and counterfactual interpretations provided according to some embodiments of the present invention is shown.

[0032] Figure 3 A flowchart illustrating a method for determining coal mill operating parameters according to some embodiments of the present invention is shown.

[0033] Figure 4 The following graph shows the results of XGB model prediction of operating indicators provided according to some embodiments of the present invention.

[0034] Figure 5 The diagram shows the results of a BP neural network model predicting performance indicators according to some embodiments of the present invention. Detailed Implementation

[0035] The following specific embodiments illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. Although the description of the present invention is presented in conjunction with preferred embodiments, this does not mean that the features of the invention are limited to these embodiments. On the contrary, the purpose of describing the invention in conjunction with embodiments is to cover other options or modifications that may be derived based on the claims of the present invention. To provide a thorough understanding of the invention, many specific details will be included in the following description. The invention may also be implemented without using these details. Furthermore, to avoid confusion or obscuring the focus of the invention, some specific details will be omitted in the description.

[0036] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0037] Furthermore, the terms "upper," "lower," "left," "right," "top," "bottom," "horizontal," and "vertical" used in the following description should be understood as the orientations shown in the relevant paragraphs and accompanying drawings. These relative terms are for illustrative purposes only and do not imply that the described apparatus must be manufactured or operated in a specific orientation, and therefore should not be construed as limiting the invention.

[0038] It is understood that although terms such as "first," "second," and "third" may be used herein to describe various components, regions, layers, and / or parts, these components, regions, layers, and / or parts should not be limited by these terms, and these terms are only used to distinguish different components, regions, layers, and / or parts. Therefore, the first components, regions, layers, and / or parts discussed below may be referred to as second components, regions, layers, and / or parts without departing from some embodiments of the present invention.

[0039] As mentioned above, due to issues such as inconsistent raw material composition and quality, and frequent fluctuations in production conditions during coal mining, existing prediction models are unable to predict complex nonlinear relationships and struggle to effectively match actual changes in operating conditions, resulting in maximizing production efficiency. Furthermore, existing prediction models are based on gradient optimization and have complex structures, leading to complex optimization processes and high computational resource consumption.

[0040] To overcome the aforementioned deficiencies in the existing technology, this invention provides a method for determining coal mill operating indicators, a device for determining coal mill operating indicators, and a computer-readable storage medium. It can replace the XGB model with a trained BP neural network model to improve the prediction accuracy and interpretability of nonlinear problems. Furthermore, by optimizing and training the prediction model based on counterfactual interpretation, it simplifies the model optimization process, thereby providing a reference for adjusting coal mill process parameters and optimizing coal mine processing operations, ultimately improving the production efficiency of mineral processing.

[0041] In some non-limiting embodiments, the method for determining coal mill operating parameters provided in the first aspect of the present invention can be implemented based on the apparatus for determining coal mill operating parameters provided in the second aspect of the present invention. Specifically, the apparatus for determining coal mill operating parameters may be configured with a memory and a processor. The memory includes, but is not limited to, the computer-readable storage medium described in the third aspect of the present invention, on which computer instructions are stored. The processor is connected to the memory and configured to execute the computer instructions stored in the memory to implement the method for determining coal mill operating parameters provided in the first aspect of the present invention.

[0042] Please refer to further details. Figure 1 . Figure 1 A schematic diagram of the process for training a first BP neural network model according to some embodiments of the present invention is shown.

[0043] like Figure 1 As shown, during the training of the first BP neural network model, the processor can acquire multiple sets of historical operating parameters X of the coal mill and input each set of historical operating parameters X into the pre-trained XGB model in sequence to obtain the corresponding first predicted value Y′. Here, each set of historical operating parameters X includes multiple operating parameters.

[0044] Specifically, during the training of the XGB model, the processor can obtain the true values ​​of the operating metrics Y corresponding to each set of historical operating parameters X, and associate each set of historical operating parameters X with its corresponding true values ​​of operating metrics Y to construct the second training sample dataset D = {X, Y}.

[0045] Next, the processor can construct the XGB model to be trained. Specifically, the processor can first construct the objective function of the XGB model:

[0046]

[0047] in, Let Ω(f) be the loss function for the i-th first training sample. k ) represents the complexity of the XGB model.

[0048] Next, the processor can perform a second-order Taylor expansion on the loss function to optimize the objective function:

[0049]

[0050] Among them, g i Let h be the first derivative of the i-th first training sample in the t-th learner. i Let be the second derivative of the i-th first training sample in the t-th learner.

[0051] Please refer to Table 1, which shows the parameter results of the random grid search method provided according to some embodiments of the present invention.

[0052] Table 1. Parameter Results of the Random Grid Search Method

[0053]

[0054] As shown in Table 1, the processor can use a random grid search method to search for the hyperparameters of the XGB model to minimize the objective function. Since the XGB model contains a large number of parameters, manual verification would be extremely labor-intensive and prone to errors. Therefore, the processor can use a random grid search method to randomly sample parameter subspaces within the parameter space for four parameters in the XGB model: the number of weak classifiers ("n_estimators"), the sampling ratio ("subsample"), the strength of L1 regularization in the objective function ("reg_alpha"), and the strength of L2 regularization in the objective function ("reg_lambda"). The optimal hyperparameters for the XGB model are then searched within these subspaces.

[0055] Next, the processor can filter the second training sample dataset. Using the coal mill's current as a benchmark, the processor can identify the coal mill's operating status and delete samples indicating abnormal operation. During normal operation, the coal mill's current ranges from approximately 40A to 50A. When the coal mill's current is below zero or significantly abnormal, the processor can determine that the coal mill is in an abnormal operating state. After this filtering process, the processor can reduce the original 58,155 data entries in the second training sample dataset to 25,587 data entries.

[0056] Furthermore, in some preferred embodiments, during the screening of the second training sample dataset, the processor can also calculate the mutual information between each group of running parameters X and their corresponding true values ​​of running metrics Y:

[0057]

[0058] Where p(x) is the marginal probability density function of the operating parameter X, p(y) is the marginal probability density function of the operating index Y, and p(x, y) is the joint probability density function of the operating parameter X and the operating index Y. In response to any mutual information being less than a preset first threshold, the processor can delete the corresponding sample from the second training sample dataset.

[0059] The processor can then normalize the second training sample dataset:

[0060]

[0061] Where n is the number of second training samples in the second training sample dataset, and m is the number of variables in the second training samples. For example, if the second training sample dataset after the above filtering has 25,587 sample data and 30 feature variables, then n = 25,587 and m = 30.

[0062] Next, the processor can divide the aforementioned second training sample dataset into a training set and a test set. The processor can use the training set to train the XGB model and use the test set to test the trained XGB model to determine the second model loss value of the XGB model.

[0063] Subsequently, in response to the second model loss value being greater than the preset second loss threshold, the processor can redefine the hyperparameters of the XGB model and retrain the XGB model using the training set based on the redefined hyperparameters.

[0064] Subsequently, in response to the second model loss value being less than or equal to the second loss threshold, the processor can determine that the training of the XGB model is complete.

[0065] Then, the processor can associate each set of historical running parameters X with the corresponding first predicted value Y′ output by the XGB model to construct the first training sample dataset D′={X,Y′}.

[0066] Next, the processor can construct an initial BP neural network model to be trained, and train the initial BP neural network model based on the first training sample dataset to obtain a second BP neural network model. Here, the initial BP neural network model consists of an input layer, hidden layers, and an output layer.

[0067] Specifically, during the training of the initial BP neural network model, the processor can sequentially input each set of historical running parameters X from the first training sample dataset D′={X,Y′} into the initial BP neural network model to obtain the third predicted value Y″′. Then, based on the difference between the third predicted value Y″′ and the first predicted value Y′, the first model loss value of the initial BP neural network model is determined.

[0068] Subsequently, in response to the first model loss value being greater than the preset first loss threshold, the processor can return the error signal along the original path of the output layer, hidden layer, and input layer, adjust the weights, biases, and other parameters of the initial BP neural network model, and retrain the initial BP neural network model.

[0069] Subsequently, in response to the first model loss value being less than or equal to the first loss threshold, the processor can determine that the training of the initial BP neural network model is complete and obtain the second BP neural network model.

[0070] After obtaining the second BP neural network model described above, the processor can determine the expected value Y″ of the operating metric based on the expected direction of the operating metric:

[0071] Y″=Y′+ε

[0072] Where ε represents the expected direction of the operating indicator.

[0073] Furthermore, in some embodiments, ε > 0 when the expected direction of the operating metric is upward, and ε < 0 when the expected direction of the operating metric is downward.

[0074] Next, the processor can sequentially input the historical running parameters X from the first training sample dataset D′={X,Y′} into the second BP neural network model to obtain the second predicted value.

[0075] Please refer to Figure 2 . Figure 2 A line graph showing historical operational data and counterfactual interpretations provided according to some embodiments of the present invention is shown.

[0076] After that, such as Figure 2 As shown, the processor can utilize the Adam optimizer based on the second predicted value. Loss between the expected value Y″ The historical operating parameters X for each group are adjusted to obtain a counterfactual explanation X′ for each group of historical operating parameters X. Then, the processor can optimize the second BP neural network model based on the counterfactual explanation X′ to obtain the first BP neural network model.

[0077] Thus, the processor can use the aforementioned second predicted value Loss between the expected value Y″ The counterfactual interpretations of the historical operating parameters for each group are calculated to make the operating indicators change in the desired direction. Thus, the model is optimized based on the counterfactual interpretations without calculating gradients, which simplifies the optimization process of the first BP neural network model and saves computing resources.

[0078] Please refer to the details. Figure 3 . Figure 3 A flowchart illustrating a method for determining coal mill operating parameters according to some embodiments of the present invention is shown.

[0079] like Figure 3As shown, after training and optimizing the first BP neural network model, the processor can acquire various operating parameters of the coal mill. These parameters are selected from at least one of voltage, current, and temperature. The processor can then input these operating parameters into the pre-trained first BP neural network model to determine the operating indicators of the coal mill. These operating indicators include at least the unit load of the coal mill.

[0080] Furthermore, in order to verify the performance of the optimized first BP neural network model and the XGB model, the present invention also performed statistical analysis on their prediction values ​​respectively.

[0081] Please refer to the reference. Figure 4 , Figure 5 Tables 2 and 3. Figure 4 The following graph shows the results of XGB model prediction of operating indicators provided according to some embodiments of the present invention. Figure 5 Table 1 shows the results of predicting the performance metrics of the first BP neural network model provided according to some embodiments of the present invention. Table 2 shows the performance metrics of the XGB model provided according to some embodiments of the present invention. Table 3 shows the performance metrics of the first BP neural network model provided according to some embodiments of the present invention.

[0082] Table 2 Performance Indicators of XGB Model

[0083] RMSE <![CDATA[R 2 ]]> 0.139 0.993

[0084] Table 3 Performance of the First BP Neural Network Model

[0085] RMSE <![CDATA[R 2 ]]> 0.499 0.906

[0086] like Figure 4 , Figure 5 As shown in Tables 2 and 3, the XGB model exhibits higher accuracy in predicting operational indicators, with a root mean square error (RMSE) of 0.139 and a coefficient of determination (R²) of 0.993. The first BP neural network model also has a RMSE of 0.139 and a R² of determination of 0.906. Therefore, the first BP neural network model can improve the prediction accuracy and interpretability of nonlinear problems while simplifying the optimization process, achieving prediction accuracy similar to the XGB model. Consequently, it is more suitable for data processing and result prediction of nonlinear problems.

[0087] In summary, the method, apparatus, and computer-readable storage medium for determining coal mill operating indicators provided by this invention can all replace the XGB model by training a BP neural network model to improve the prediction accuracy and interpretability of nonlinear problems. Furthermore, by optimizing and training the prediction model based on counterfactual interpretation, the optimization process of the model can be simplified, thereby providing a reference for adjusting coal mill process parameters and optimizing the coal mine processing operation, thus improving the production efficiency of the mineral processing process.

[0088] Although the methods described above are illustrated and depicted as a series of actions for the sake of simplicity, it should be understood and appreciated that these methods are not limited by the order of the actions, as some actions may occur in a different order and / or concurrently with other actions from the illustrations and descriptions herein or not illustrated and described herein but which may be understood by those skilled in the art, according to one or more embodiments.

[0089] The prior description of this disclosure is provided to enable any person skilled in the art to make or use this disclosure. Various modifications to this disclosure will be apparent to those skilled in the art, and the general principles defined herein may be applied to other variations without departing from the spirit or scope of this disclosure. Therefore, this disclosure is not intended to be limited to the examples and designs described herein, but should be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for determining the operating parameters of a coal mill, characterized in that, Includes the following steps: Obtain multiple operating parameters of the coal mill, wherein the multiple operating parameters are selected from at least one of voltage, current, and temperature; and The various operating parameters are input into a pre-trained first BP neural network model to determine the operating indicators of the coal mill, wherein the operating indicators include at least the unit load of the coal mill. The step of training the first BP neural network model includes: obtaining multiple sets of historical operating parameters of the coal mill. Each set of historical operating parameters The data includes various operating parameters; the historical operating parameters in each group are... The pre-trained XGB model is input sequentially to obtain the corresponding first predicted value. ; associate the historical operating parameters of each group Its corresponding first predicted value The process involves: constructing a first training sample dataset; constructing an initial BP neural network model to be trained; and training the initial BP neural network model based on the first training sample dataset to obtain a second BP neural network model; and determining the expected value of the operating metric. : in, Indicates the desired direction of the operational metrics; the historical operational parameters from each group in the first training sample dataset... The second BP neural network model is input sequentially to obtain the second predicted value. ; and using the Adam optimizer, based on the second predicted value With the expected value Losses between The second BP neural network model is optimized to obtain the first BP neural network model.

2. The determination method as described in claim 1, characterized in that, The Adam optimizer is used based on the second predicted value. With the expected value Losses between The steps of optimizing the second BP neural network model to obtain the first BP neural network model include: According to the second predicted value With the expected value Losses between For each group of historical operating parameters Adjustments were made to obtain the historical operating parameters for each group. Counterfactual interpretation ;as well as According to the counterfactual interpretation The second BP neural network model is optimized to obtain the first BP neural network model.

3. The determination method as described in claim 1, characterized in that, The step of training the initial BP neural network model based on the first training sample dataset to obtain the second BP neural network model includes: The historical running parameters in each group of the first training sample dataset The initial BP neural network model is input sequentially to obtain the third predicted value. ; According to the third predicted value Compared with the first predicted value The difference is used to determine the first model loss value of the initial BP neural network model; In response to the first model loss value exceeding a preset first loss threshold, the parameters of the initial BP neural network model are adjusted, and the initial BP neural network model is retrained; and In response to the first model loss value being less than or equal to the first loss threshold, it is determined that the training of the initial BP neural network model is complete, and the second BP neural network model is obtained.

4. The determination method as described in claim 1, characterized in that, The steps for training the XGB model include: Obtain the historical operating parameters of each group. Corresponding actual values ​​of operating indicators And associate the historical operating parameters of each group. The actual values ​​of the corresponding operating indicators To construct a second training sample dataset; Construct the XGB model to be trained; and The XGB model is trained based on the second training sample dataset.

5. The determination method as described in claim 4, characterized in that, The steps for constructing the XGB model to be trained include: Construct the objective function of the XGB model: in, For the first The loss function for the first training sample. This indicates the complexity of the XGB model; The loss function is expanded using a second-order Taylor series to optimize the objective function. in, For the first The first training sample in the _ The first derivative in a learner For the first The first training sample in the _ The second derivative in each learner; and The hyperparameters of the XGB model are searched using a random grid search method to minimize the objective function.

6. The determination method as described in claim 5, characterized in that, The step of training the XGB model based on the second training sample dataset includes: The second training sample dataset is preprocessed and divided into a training set and a test set. The XGB model is trained using the training set and tested using the test set to determine the second model loss value of the XGB model. In response to the second model loss value exceeding a preset second loss threshold, the hyperparameters of the XGB model are redefined, and the XGB model is retrained using the training set based on the redefined hyperparameters; and When the second model loss value is less than or equal to the second loss threshold, the training of the XGB model is determined to be complete.

7. The determination method as described in claim 6, characterized in that, The preprocessing steps for the second training sample dataset include: The second training sample dataset is then filtered; and Normalize the second training sample dataset: in, The number of second training samples in the second training sample dataset. The number of variables in the second training sample.

8. The determination method as described in claim 7, characterized in that, The step of filtering the second training sample dataset includes: Calculate the operating parameters for each group. and the corresponding actual values ​​of operating indicators Mutual information between them: in, For the operating parameters The marginal probability density function, For the aforementioned operating indicators The marginal probability density function, For the operating parameters With the aforementioned operating indicators The joint probability density function; and In response to any of the mutual information being less than a preset first threshold, the corresponding sample is deleted from the second training sample dataset.

9. A device for determining the operating parameters of a coal mill, characterized in that, include: Memory, on which computer instructions are stored; and A controller is connected to the memory and configured to execute computer instructions stored thereon to implement the method for determining the operating parameters of the coal mill as described in any one of claims 1 to 8.

10. A computer-readable storage medium storing computer instructions thereon, characterized in that, When the computer instructions are executed by the processor, the method for determining the operating parameters of the coal mill as described in any one of claims 1 to 8 is implemented.

Images