A method, apparatus, electronic device and storage medium for predicting the temperature of a rotary kiln.

Abstract

This invention discloses a method, device, electronic device, and storage medium for predicting the temperature of a rotary kiln. This method, applied in the field of rotary kiln technology, addresses the problems of low accuracy and poor timeliness in existing temperature prediction techniques. The proposed method includes: acquiring the current operating condition information of the rotary kiln; analyzing the current operating condition information using a pre-established rotary kiln temperature prediction model to obtain the current temperature field data of the rotary kiln; wherein the establishment process of the rotary kiln temperature prediction model is as follows: based on the historical operating condition information of the rotary kiln and the corresponding historical temperature field data, combined with a stacked self-encoded objective function modified by mutual information, a rotary kiln temperature prediction model is established. This invention can quickly predict the temperature of the rotary kiln, improve the real-time performance of rotary kiln temperature prediction, and facilitate better monitoring of the rotary kiln's operating status.

Description

Technical Field

[0001] This invention relates to the field of rotary kiln technology, and in particular to a method, apparatus, electronic device, and computer-readable storage medium for predicting the temperature of a rotary kiln. Background Technology

[0002] CFD (Computational Fluid Dynamics) simulation software is a commonly used tool for digital simulation of temperature fields. It has rich physical models, advanced numerical methods and powerful pre- and post-processing functions. It has wide applications in aerospace, automotive design, oil and gas and turbine design, and is also a commonly used technology for building digital twin models of industrial processes.

[0003] In industrial processes, many target parameters are difficult or impossible to measure, making it challenging to construct effective soft measurement models due to a lack of valid data. CFD-based numerical simulation tools can simulate the entire process or reaction by modeling the process mechanism, thereby obtaining simulated target data. The more detailed and comprehensive the mechanism model, the more realistic the simulation results. However, CFD numerical simulations typically involve large computational loads and long computation times; simulating a single operating condition can often take several days, and even longer for scenarios with large mesh sizes, falling far short of the real-time acquisition requirements for many difficult-to-measure variables in industrial settings.

[0004] Therefore, how to provide a more real-time rotary kiln temperature prediction method, device, electronic equipment, and computer-readable storage medium has become a problem that needs to be solved by those skilled in the art. Summary of the Invention

[0005] The purpose of this invention is to provide a method, apparatus, electronic device, and computer-readable storage medium for predicting the temperature of a rotary kiln. During use, the corresponding current temperature field data can be quickly predicted, improving the real-time performance of rotary kiln temperature prediction and facilitating better monitoring of the rotary kiln's operating status.

[0006] To address the aforementioned technical problems, the embodiments of the present invention provide the following technical solutions:

[0007] This invention provides a method for predicting the temperature of a rotary kiln, comprising:

[0008] Obtain the current operating status information of the rotary kiln;

[0009] The current operating condition information is analyzed using a pre-established rotary kiln temperature prediction model to obtain the current temperature field data of the rotary kiln; wherein, the process of establishing the rotary kiln temperature prediction model is as follows:

[0010] Based on the historical operating information of the rotary kiln and the corresponding historical temperature field data, a rotary kiln temperature prediction model is established by combining the stack autoencode objective function with mutual information correction.

[0011] In one exemplary embodiment, the step of establishing a rotary kiln temperature prediction model based on the historical operating condition information of the rotary kiln and the corresponding historical temperature field data, combined with a stacked autoencoder objective function modified by mutual information, includes:

[0012] Acquire multiple sets of historical operating condition information of the rotary kiln and historical temperature field data corresponding to the historical operating condition information;

[0013] The autoencoder is obtained by training a stacked autoencoder objective function with mutual information correction using multiple sets of historical operating condition information and corresponding historical temperature field data.

[0014] The autoencoder is used to extract features from each of the temperature field data to obtain historical temperature feature data corresponding to each of the historical operating conditions.

[0015] A shallow neural network was trained using multiple sets of historical operating condition information and corresponding historical temperature feature data to obtain a temperature feature prediction model.

[0016] A rotary kiln temperature prediction model is formed based on the temperature feature prediction model and the self-encoder.

[0017] In one exemplary implementation, the stack autoencoder objective function after mutual information correction is:

[0018] ;

[0019] Where T represents the objective function, N represents the number of input variable parameters, and M represents the number of hidden layers in the neural network. This represents the transfer function of the decoder. This represents the connection weight of the j-th input node of the decoder. This represents the connection parameter of the i-th input of the encoder. Let 'i' represent the i-th input variable of the encoder, and 'b' represent the encoder's connection parameters. This represents the connection threshold of the j-th input to the decoder. This represents the encoder's transfer function. This represents the joint probability distribution of the encoder's output and y. This represents the marginal probability distribution of the encoder's output. Let y represent the marginal probability distribution.

[0020] In one exemplary embodiment, the step of analyzing the current operating condition information using a pre-established rotary kiln temperature prediction model to obtain the current temperature field data of the rotary kiln includes:

[0021] The temperature feature prediction model in the rotary kiln temperature prediction model is used to process the current operating condition information to obtain various current temperature features corresponding to the current operating condition information.

[0022] The autoencoder in the rotary kiln temperature prediction model is used to reverse process each of the current temperature features to generate current temperature field data corresponding to the current operating condition information.

[0023] In one exemplary implementation, the operating information includes: gas volume, total air volume, secondary air inlet air volume ratio, rotary kiln speed, feed rate, and total coal ratio.

[0024] In one exemplary embodiment, acquiring multiple sets of historical operating condition information of the rotary kiln and historical temperature field data corresponding to the historical operating condition information includes:

[0025] Obtain multiple sets of historical operating condition information;

[0026] CFD numerical simulation was used to obtain historical temperature field data corresponding to each set of historical operating conditions.

[0027] Another aspect of the present invention provides a temperature prediction device for a rotary kiln, comprising:

[0028] The acquisition module is used to acquire the current operating condition information of the rotary kiln;

[0029] The prediction module is used to analyze the current operating condition information using a pre-established rotary kiln temperature prediction model to obtain the current temperature field data of the rotary kiln; wherein, the establishment module is used for:

[0030] Based on the historical operating information of the rotary kiln and the corresponding historical temperature field data, a rotary kiln temperature prediction model is established by combining the stack autoencode objective function with mutual information correction.

[0031] In one exemplary implementation, the establishment module includes:

[0032] The acquisition unit is used to acquire multiple sets of historical operating condition information of the rotary kiln and historical temperature field data corresponding to the historical operating condition information;

[0033] The first training unit is used to train an autoencoder by using multiple sets of historical operating condition information and corresponding historical temperature field data, combined with a stacked autoencoder objective function modified by mutual information.

[0034] The extraction unit is used to extract features from each of the temperature field data using the autoencoder to obtain historical temperature feature data corresponding to each of the historical operating conditions.

[0035] The second training unit is used to train a shallow neural network using multiple sets of historical operating condition information and corresponding historical temperature feature data to obtain a temperature feature prediction model.

[0036] A unit is established to form a rotary kiln temperature prediction model based on the temperature feature prediction model and the autoencoder.

[0037] Another aspect of the present invention provides an electronic device, comprising:

[0038] Memory, used to store computer programs;

[0039] A processor is used to execute the computer program to implement the steps of the rotary kiln temperature prediction method as described above.

[0040] In another aspect, the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the rotary kiln temperature prediction method described above.

[0041] As can be seen from the above technical solutions, the embodiments of the present invention have the following advantages:

[0042] This invention provides a method for predicting the temperature of a rotary kiln, comprising: acquiring the current operating condition information of the rotary kiln; analyzing the current operating condition information using a pre-established rotary kiln temperature prediction model to obtain the current temperature field data of the rotary kiln; wherein, the process of establishing the rotary kiln temperature prediction model is as follows: based on the historical operating condition information of the rotary kiln and the corresponding historical temperature field data, and combined with a stacked self-encoding objective function corrected by mutual information, a rotary kiln temperature prediction model is established.

[0043] Therefore, in this embodiment of the invention, mutual information can be used to modify the stack self-encoding objective function in advance to obtain the modified stack self-encoding objective function. Then, based on the historical operating information of the rotary kiln and the historical temperature field data corresponding to the historical operating information, a rotary kiln temperature prediction model is established in combination with the modified stack self-encoding objective function. When predicting the rotary kiln temperature, the current operating information of the rotary kiln is analyzed through the rotary kiln temperature prediction model, and the corresponding current temperature field data can be quickly predicted, thereby improving the real-time performance of the rotary kiln temperature prediction and facilitating better monitoring of the rotary kiln's operating status.

[0044] Furthermore, the present invention also provides a corresponding implementation device, electronic device, and computer-readable storage medium for the rotary kiln temperature prediction method, which further makes the method more practical, and the device, electronic device, and computer-readable storage medium have corresponding advantages.

[0045] It should be understood that the above general description and the following detailed description are merely exemplary and do not limit the invention. Attached Figure Description

[0046] To more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the prior art and embodiments 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.

[0047] Figure 1 A schematic flowchart of a rotary kiln temperature prediction method provided in an embodiment of the present invention;

[0048] Figure 2 A schematic diagram of a rotary kiln temperature prediction model provided in an embodiment of the present invention;

[0049] Figure 3 This is a structural diagram of a traditional depth SAE in the prior art;

[0050] Figure 4a for Figure 3 Internal structure diagram of AE1;

[0051] Figure 4b for Figure 3 Internal structure diagram of AE2;

[0052] Figure 5 This is a schematic diagram of a temperature feature prediction model provided in an embodiment of the present invention;

[0053] Figure 6 This is a structural diagram of a rotary kiln temperature prediction model provided in an embodiment of the present invention;

[0054] Figure 7 This is a schematic diagram of the structure of a temperature prediction device for a rotary kiln provided in an embodiment of the present invention;

[0055] Figure 8 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention;

[0056] Figure 9 This is a schematic diagram of the structure of a computer-readable storage medium provided in an embodiment of the present invention. Detailed Implementation

[0057] This invention provides a method, apparatus, electronic device, and computer-readable storage medium for predicting the temperature of a rotary kiln. During use, it can quickly predict the corresponding current temperature field data, improve the real-time performance of rotary kiln temperature prediction, and facilitate better monitoring of the rotary kiln's operating status.

[0058] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0059] Please refer to Figure 1 , Figure 1 This is a schematic flowchart of a rotary kiln temperature prediction method provided in an embodiment of the present invention.

[0060] S110: Obtain the current operating condition information of the rotary kiln;

[0061] It should be noted that, in this embodiment of the invention, a rotary kiln temperature prediction model can be pre-established. Specifically, this model can be established based on historical operating condition information and corresponding historical temperature field data. When predicting the temperature of the rotary kiln, the current operating condition information of the rotary kiln can be obtained. This current operating condition information may include a combination of one or more of the following: gas flow rate, total air flow rate, secondary air inlet air flow rate ratio, rotary kiln rotation speed, feed rate, and total coal ratio.

[0062] S120: The current operating condition information is analyzed using a pre-established rotary kiln temperature prediction model to obtain the current temperature field data of the rotary kiln; the process of establishing the rotary kiln temperature prediction model is as follows:

[0063] Based on the historical operating information of the rotary kiln and the corresponding historical temperature field data, a rotary kiln temperature prediction model is established by combining the objective function of the stack autoencoder with mutual information correction.

[0064] Specifically, in this embodiment of the invention, a rotary kiln temperature prediction model can be pre-established. During the establishment of this model, historical operating information of the rotary kiln and corresponding historical temperature field data can be used, combined with a stacked self-encoding objective function modified by mutual information, to establish the rotary kiln temperature prediction model. This model is then used to analyze the current operating information to obtain the predicted current temperature field data corresponding to the current operating information. Further, as... Figure 2As shown, the process of establishing a rotary kiln temperature prediction model in S120 above, based on the historical operating information of the rotary kiln and the corresponding historical temperature field data, combined with the stacked autoencoder objective function after mutual information correction, may specifically include:

[0065] S210: Obtain multiple sets of historical operating condition information of the rotary kiln and the historical temperature field data corresponding to the historical operating condition information;

[0066] It should be noted that in this embodiment of the invention, multiple sets of historical operating condition information of the rotary kiln and the historical temperature field data of the rotary kiln corresponding to each set of historical operating condition information can be obtained. Specifically, CFD numerical simulation models can be used to obtain the historical temperature field data corresponding to the historical operating condition information. Specifically, based on the orthogonal experimental method, the number of variables affecting the operating condition of the research object can be determined, an appropriate number of levels can be selected, i.e., the corresponding values ​​of the variables, and a reasonable experimental framework can be constructed to traverse the value range of the variables. For example, according to the orthogonal experimental method, six sets of variable factors can be designed: gas volume, total air volume, secondary air outlet air volume ratio, rotary kiln speed, feed rate, and total coal ratio, with 3 or 5 levels for CFD simulation requirements. The number of levels corresponds to the specific values ​​of the variables. The number of experimental groups can be determined according to the specific experiment, with 18 or 25 groups respectively. The experimental table is as follows. Each group achieves a steady-state temperature field based on the specific values ​​of the variable factors. The optimal parameter group includes a gas volume of 12 Nm³ corresponding to level 3. 3 / h, the secondary air volume of level 3 accounts for 50%, the rotary kiln speed of level 4 is 110s / revolution, the feed rate of level 5 is 180kg / h, and the total coal ratio of level 1 is 22%.

[0067] Table 1 Experimental Table

[0068]

[0069] Within this experimental framework, we further expanded the combination of operating conditions and used the resources of the supercomputing center to run the CFD simulation model in parallel at multiple points to obtain temperature field data under different operating conditions, which served as the training set.

[0070] S220: The autoencoder is trained by using multiple sets of historical operating condition information and corresponding historical temperature field data, combined with a stacked autoencoder objective function modified by mutual information.

[0071] Specifically, mutual information can be used to modify the stack autoencoder objective function to obtain the modified stack autoencoder objective function. Then, the training set is divided into a training set and a test set in a 7:3 ratio, and the modified stack autoencoder objective function is used for training to obtain the trained autoencoder.

[0072] The process of determining the modified stack auto-encoding target function in real time in this invention is as follows:

[0073] It should be noted that traditional SAE (Stacked Autoencoder) algorithms do not consider the correlation with the target variable during feature extraction. Therefore, the extracted features may be unrelated to the target variable, making it impossible to establish a relationship model between the extracted features and the target parameters. Please refer to... Figure 3 And Figure 4, Figure 3 Figure 4 shows the structure of a traditional depth SAE. Figure 3 Internal structure of AE1 and AE2:

[0074] The input to AE1 is multiple industrial process data. ,in, to Let these represent the input variables of AE1 and each node in the hidden layer H of AE1, respectively. Where h represents the output of each node, and f represents the activation function. This represents the connection weights between the input and the hidden layer. This represents the connection threshold between the input and the hidden layer. Each layer of the Advanced Effects (AE) extracts corresponding features. ,in, to These represent the outputs of the hidden layer, H, which extracts features from the input data. The learning objective of the model is to make the input vector X and the output vector X equal. The principle of minimizing the mean square error is to ensure that the extracted features can be decoded to reconstruct the data as much as possible.

[0075] (1);

[0076] Where N represents the number of input variable parameters, and T represents the objective function.

[0077] It is evident that the learning objective of traditional SAE models is unrelated to the target variable. Therefore, this invention designs an autoencoder MI-AE based on mutual information (MI) correction to address the correlation between extracted features and target parameters. Specifically:

[0078] Mutual information (MI) is a measure in information theory used to evaluate the degree of dependence between two random variables. For variables X and Y, mutual information is the inherent dependence between the joint distribution of variables X and Y relative to the joint distribution assuming X and Y are independent. In other words, mutual information measures the degree to which the uncertainty about one variable is reduced when the other is known. When the two variables are uncorrelated, their mutual information is 0; when they are correlated, their mutual information is positive, and the stronger the correlation, the larger the value of the mutual information. When the two variables are perfectly correlated, the mutual information is 1. Unlike Pearson correlation, mutual information (MI) can measure non-monotonic and even more complex relationships between multiple variables.

[0079] For two continuous random variables X and Y, their mutual information is:

[0080] (2);

[0081] in, Represents the marginal probability distribution of variable X. This represents the marginal probability distribution of variable Y. This represents the joint probability distribution of variables X and Y. When calculating mutual information for discrete variables, the integral in the above formula is replaced by a summation operation as follows:

[0082] (3);

[0083] To more intuitively analyze the correlation between two variables expressed by mutual information, mutual information is usually represented in an entropy-based form. From the definition of entropy, the entropy of variable X can be expressed as:

[0084] (4);

[0085] The entropy of a variable measures its uncertainty; the more discrete the distribution of X, the higher the value of H(X). The conditional entropy between two variables is:

[0086] (5);

[0087] H(X|Y) represents the uncertainty of X given Y. According to the summation and product rules of probabilities, the relationship between mutual information and conditional entropy is as follows:

[0088] (6);

[0089] As can be seen from the above formula, To the extent that the introduction of Y reduces the uncertainty of X, the amount of reduction is... Therefore, the closer the relationship between X and Y, the better. The larger it is, the better. The maximum value is ,at this time A value of 0 indicates that X and Y are perfectly correlated. That is, with Y involved, X is a constant, and there is no probability of other uncertainties arising. Therefore, It is 0. The minimum value is 0, which means that X and Y are independent. The introduction of Y has no effect on X.

[0090] Currently, the objective function for extracting feature variables using an autoencoder (AE) is to minimize the loss function, which can be expressed by Equation 3-26 as follows:

[0091] (7);

[0092] in, and Here, T represents the encoder's connection parameters, N represents the objective function, and M represents the number of input variable parameters. This represents the transfer function of the decoder. This represents the connection weight of the j-th input node of the decoder. This represents the connection parameter of the i-th input of the encoder. Let 'i' represent the i-th input variable of the encoder, and 'b' represent the encoder's connection parameters. This represents the connection threshold of the j-th input to the decoder. This represents the encoder's transfer function.

[0093] To ensure the correlation between the feature variables extracted by the AE and the target parameters, the feature variable H (here referring to the hidden layer output of the AE) is extracted. The greater the mutual information between the target parameter Y and the target parameter Y, the better. Therefore, the above formula can be modified as follows: Taking a single target parameter as an example, we have... :

[0094] (8);

[0095] (9);

[0096] (10);

[0097] in, This indicates the hidden layer output of AE. This represents the joint probability distribution of the hidden layer output and y. Let represent the joint probability distribution of the output of the i-th hidden layer and y. This represents the marginal probability distribution of the output of the i-th hidden layer. Let y represent the marginal probability distribution.

[0098] Therefore, combining the above relationships, the objective function of the improved stack auto-encoding can be obtained as follows:

[0099] (11);

[0100] Where T represents the objective function, N represents the number of input variable parameters, and M represents the number of hidden layers in the neural network. This represents the transfer function of the decoder. This represents the connection weight of the j-th input node of the decoder. This represents the connection parameter of the i-th input of the encoder. Let 'i' represent the i-th input variable of the encoder, and 'b' represent the encoder's connection parameters. This represents the connection threshold of the j-th input to the decoder. This represents the encoder's transfer function. This represents the joint probability distribution of the encoder's output and y. This represents the marginal probability distribution of the encoder's output. Let y represent the marginal probability distribution.

[0101] As can be seen from the above formula, the autoencoder based on mutual information can extract the feature variables that are most correlated with the target parameters.

[0102] S230: Uses an autoencoder to extract features from each temperature field data to obtain historical temperature feature data corresponding to each historical operating condition information.

[0103] Specifically, in this embodiment of the invention, after obtaining the trained autoencoder, the trained autoencoder can be further used to extract features from the temperature field data corresponding to each historical working condition, thereby obtaining the historical temperature feature data corresponding to each historical working condition.

[0104] S240: A shallow neural network is trained using multiple sets of historical operating condition information and corresponding historical temperature feature data to obtain a temperature feature prediction model.

[0105] It should be noted that after obtaining the historical temperature characteristic data corresponding to each historical operating condition, such as... Figure 5 As shown, historical operating condition information can be further used as input to a shallow neural network, which includes an input layer, a hidden layer, and an output layer. Error feedback is performed by combining historical temperature feature data (i.e., actual feature data) with predicted temperature feature data to update the parameters of the neural network. Training stops when the accuracy of the trained model meets the preset conditions, and a well-trained temperature feature prediction model is obtained.

[0106] S250: A rotary kiln temperature prediction model is formed based on the temperature characteristic prediction model and the self-encoder.

[0107] In this embodiment of the invention, a temperature feature prediction model is combined with an autoencoder (specifically, a decoding model of a mutual information-based autoencoder MISAE) to form a rotary kiln temperature prediction model. For the specific structure, please refer to [reference needed]. Figure 6 .

[0108] It is understandable that the process of using this rotary kiln temperature prediction model to analyze the current operating condition information and obtain the current temperature field data of the rotary kiln may include:

[0109] The temperature feature prediction model in the rotary kiln temperature prediction model is used to process the current operating condition information to obtain various current temperature features corresponding to the current operating condition information.

[0110] The autoencoder in the rotary kiln temperature prediction model is used to reverse process each current temperature feature to generate current temperature field data corresponding to the current operating condition information.

[0111] In other words, please refer to Figure 6 In this embodiment of the invention, the current operating condition information is input into the temperature feature prediction model in the rotary kiln temperature prediction model. The temperature feature prediction model analyzes the current operating condition information and outputs various current temperature features corresponding to the current operating condition information. These current temperature features are then input into the autoencoder. The autoencoder's decoding model performs reverse decoding on the current temperature features to generate corresponding 1D temperature data, that is, the temperature features are converted into current temperature field data.

[0112] Therefore, in this embodiment of the invention, mutual information can be used to modify the stack self-encoding objective function in advance to obtain the modified stack self-encoding objective function. Then, based on the historical operating information of the rotary kiln and the historical temperature field data corresponding to the historical operating information, a rotary kiln temperature prediction model is established in combination with the modified stack self-encoding objective function. When predicting the rotary kiln temperature, the current operating information of the rotary kiln is analyzed through the rotary kiln temperature prediction model, and the corresponding current temperature field data can be quickly predicted, thereby improving the real-time performance of the rotary kiln temperature prediction and facilitating better monitoring of the rotary kiln's operating status.

[0113] In other words, the embodiments of the present invention can solve the problem of lack of labeled data for important and difficult-to-measure variables in industrial processes; it has good real-time performance and low requirements for equipment computing power; it can help realize industrial digital twin modeling and guide the control process in real time.

[0114] This invention also provides a corresponding apparatus for the rotary kiln temperature prediction method, further enhancing the practicality of the method. The apparatus can be described from both a functional module perspective and a hardware perspective. The rotary kiln temperature prediction apparatus provided by this invention is described below. This apparatus is used to implement the rotary kiln temperature prediction method provided by this invention. In this embodiment, the rotary kiln temperature prediction apparatus may include or be divided into one or more program modules. These one or more program modules are stored in a storage medium and executed by one or more processors to complete the rotary kiln temperature prediction method disclosed in the above embodiments. The program module referred to in this invention is a series of computer program instruction segments capable of performing specific functions, which is more suitable than the program itself for describing the execution process of the rotary kiln temperature prediction apparatus in the storage medium. The following description will specifically introduce the functions of each program module in this embodiment. The rotary kiln temperature prediction apparatus described below corresponds to the rotary kiln-based temperature prediction method described above.

[0115] From the perspective of functional modules, see Figure 7 , Figure 7 A structural diagram of the rotary kiln temperature prediction device provided by the present invention in a specific embodiment, the device may include:

[0116] The acquisition module 11 is used to acquire the current operating condition information of the rotary kiln;

[0117] Prediction module 12 is used to analyze the current operating condition information using a pre-established rotary kiln temperature prediction model to obtain the current temperature field data of the rotary kiln; wherein, the establishment module 20 is used for:

[0118] Based on the historical operating information of the rotary kiln and the corresponding historical temperature field data, a rotary kiln temperature prediction model is established by combining the objective function of the stack autoencoder with mutual information correction.

[0119] In one exemplary embodiment, the establishment module 20 includes:

[0120] The acquisition unit is used to acquire multiple sets of historical operating condition information of the rotary kiln and historical temperature field data corresponding to the historical operating condition information.

[0121] The first training unit is used to train an autoencoder by using multiple sets of historical operating condition information and corresponding historical temperature field data, combined with a stacked autoencoder objective function modified by mutual information.

[0122] The extraction unit is used to extract features from each temperature field data using an autoencoder to obtain historical temperature feature data corresponding to each historical operating condition information.

[0123] The second training unit is used to train a shallow neural network using multiple sets of historical operating condition information and corresponding historical temperature feature data to obtain a temperature feature prediction model.

[0124] A unit is established to form a rotary kiln temperature prediction model based on temperature characteristic prediction models and autoencoders.

[0125] It should be noted that the rotary kiln temperature prediction device provided in the embodiments of the present invention has the same beneficial effects as the rotary kiln temperature prediction method provided in the above embodiments, and for a detailed description of the rotary kiln temperature prediction method involved in the embodiments of the present invention, please refer to the above embodiments, which will not be repeated here.

[0126] The rotary kiln temperature prediction device mentioned above is described from the perspective of functional modules. Furthermore, the present invention also provides an electronic device, which is described from the perspective of hardware. Figure 8 A structural diagram of an electronic device provided in an embodiment of this application, such as... Figure 8 As shown, the electronic device includes: a memory 20 for storing computer programs;

[0127] The processor 21 is used to execute a computer program to implement the steps of the rotary kiln temperature prediction method as described in the above embodiment.

[0128] The electronic devices provided in this embodiment may include, but are not limited to, smartphones, tablets, laptops, or desktop computers.

[0129] The processor 21 may include one or more processing cores, such as a quad-core processor or an octa-core processor. The processor 21 may be implemented using at least one hardware form selected from DSP (Digital Signal Processing), FPGA (Field-Programmable Gate Array), and PLA (Programmable Logic Array). The processor 21 may also include a main processor and a coprocessor. The main processor, also known as a CPU (Central Processing Unit), is used to process data in the wake-up state; the coprocessor is a low-power processor used to process data in the standby state. In some embodiments, the processor 21 may integrate a GPU (Graphics Processing Unit), which is responsible for rendering and drawing the content to be displayed on the screen. In some embodiments, the processor 21 may also include an AI (Artificial Intelligence) processor, which is used to handle computational operations related to machine learning.

[0130] The memory 20 may include one or more computer-readable storage media, which may be non-transitory. The memory 20 may also include high-speed random access memory and non-volatile memory, such as one or more disk storage devices or flash memory devices. In some embodiments, the memory 20 may be an internal storage unit of an electronic device, such as a server hard drive. In other embodiments, the memory 20 may be an external storage device of an electronic device, such as a plug-in hard drive on a server, a smart media card (SMC), a secure digital card (SD), a flash card, etc. Furthermore, the memory 20 may include both internal and external storage units of the electronic device. The memory 20 can be used not only to store application software and various types of data installed on the electronic device, such as code in the process of executing the rotary kiln temperature prediction method, but also to temporarily store data that has been output or will be output. In this embodiment, the memory 20 is used to store at least the following computer program 201, which, after being loaded and executed by the processor 21, is capable of implementing the relevant steps of the rotary kiln temperature prediction method disclosed in any of the foregoing embodiments. In addition, the resources stored in the memory 20 may also include an operating system 202 and data 203, and the storage method may be temporary storage or permanent storage. The operating system 202 may include Windows, Unix, Linux, etc. The data 203 may include, but is not limited to, data corresponding to the temperature prediction results of the rotary kiln.

[0131] In some embodiments, the electronic device may further include a display screen 22, an input / output interface 23, a communication interface 24, a power supply 25, and a communication bus 26. The display screen 22 and input / output interface 23, such as a keyboard, are user interfaces; optional user interfaces may also include standard wired interfaces, wireless interfaces, etc. Optionally, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, or an OLED (Organic Light-Emitting Diode) touchscreen, etc. The display may also be appropriately referred to as a screen or display unit, used to display information processed in the electronic device and to display a visual user interface. The communication interface 24 may optionally include a wired interface and / or a wireless interface, such as a Wi-Fi interface, a Bluetooth interface, etc., typically used to establish communication connections between the electronic device and other electronic devices. The communication bus 26 may be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. This bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 8 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0132] Those skilled in the art will understand that Figure 8 The structures shown do not constitute a limitation on electronic devices and may include more or fewer components than those shown.

[0133] It is understood that if the rotary kiln temperature prediction method in the above embodiments 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, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and executes all or part of the steps of the methods in the various embodiments of this application. The aforementioned storage medium includes: USB flash drive, mobile hard drive, read-only memory (ROM), random access memory (RAM), electrically erasable programmable ROM, register, hard disk, removable disk, CD-ROM, magnetic disk, or optical disk, and other media capable of storing program code.

[0134] Based on this, such as Figure 9As shown, this embodiment of the invention also provides a computer-readable storage medium 30, on which a computer program 31 is stored. When the computer program 31 is executed by a processor, it implements the steps of the rotary kiln temperature prediction method described above.

[0135] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to the method section.

[0136] It should also be noted that, in this specification, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0137] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A temperature prediction method of a rotary kiln, characterized by, The method comprises the following steps: obtaining current working condition information of a rotary kiln; analyzing the current working condition information by using a pre-established rotary kiln temperature prediction model to obtain current temperature field data of the rotary kiln; wherein the establishment process of the rotary kiln temperature prediction model is as follows: based on historical working condition information and corresponding historical temperature field data of the rotary kiln, and in combination with a stack auto-encoding objective function modified by mutual information, a rotary kiln temperature prediction model is established; wherein: the rotary kiln temperature prediction model is established based on the historical working condition information and the corresponding historical temperature field data of the rotary kiln, and in combination with the stack auto-encoding objective function modified by mutual information, and comprises the following steps: obtaining multiple sets of historical working condition information of the rotary kiln and historical temperature field data corresponding to the historical working condition information; training, by using multiple sets of the historical working condition information and the corresponding historical temperature field data, in combination with the stack auto-encoding objective function modified by mutual information, to obtain an auto-encoder; performing feature extraction on each of the temperature field data by using the auto-encoder to obtain historical temperature feature data corresponding to each of the historical working condition information; performing shallow neural network training on multiple sets of the historical working condition information and the corresponding historical temperature feature data to obtain a temperature feature prediction model; forming a rotary kiln temperature prediction model according to the temperature feature prediction model and the auto-encoder.

2. The temperature prediction method of a rotary kiln according to claim 1, characterized by, the stack auto-encoding objective function modified by mutual information is as follows: ； wherein T represents a target function, N represents a number of input variable parameters, M represents a number of hidden layer neural networks, represents a transfer function of a decoder, represents a connection weight of a jth input node of a decoder, represents a connection parameter of a ith input of an encoder, represents an ith input variable of an encoder, and b represents a connection parameter of an encoder, represents a connection threshold of a jth input of a decoder, represents a transfer function of an encoder, represents a joint probability distribution of an output of an encoder and y, represents an edge probability distribution of an output of an encoder, represents an edge probability distribution of y.

3. The temperature prediction method of a rotary kiln according to claim 1, characterized by, the rotary kiln temperature prediction model is established based on the historical working condition information and the corresponding historical temperature field data of the rotary kiln, and in combination with the stack auto-encoding objective function modified by mutual information, and comprises the following steps: processing the current working condition information by using a temperature feature prediction model in the rotary kiln temperature prediction model to obtain each current temperature feature corresponding to the current working condition information; performing reverse processing on each of the current temperature features by using an auto-encoder in the rotary kiln temperature prediction model to generate current temperature field data corresponding to the current working condition information.

4. The temperature prediction method of a rotary kiln according to any one of claims 1 to 3, characterized in that, The working condition information includes: gas quantity, total air quantity, secondary air port air quantity proportion, rotary kiln speed, feeding quantity, and total coal proportion.

5. The temperature prediction method of a rotary kiln according to claim 1, characterized by, the multiple sets of historical working condition information of the rotary kiln and the historical temperature field data corresponding to the historical working condition information are obtained by comprising the following steps: obtaining multiple sets of historical working condition information; obtaining historical temperature field data corresponding to each set of the historical working condition information by using CFD numerical simulation.

6. A temperature prediction device for a rotary kiln, characterized by The method comprises the following steps: an obtaining module is configured to obtain current working condition information of a rotary kiln; a prediction module is configured to analyze the current working condition information by using a pre-established rotary kiln temperature prediction model to obtain current temperature field data of the rotary kiln; an establishment module is configured to establish a rotary kiln temperature prediction model based on historical working condition information and corresponding historical temperature field data of the rotary kiln, and in combination with a stack auto-encoding objective function modified by mutual information; wherein: the establishment module comprises: an obtaining unit is configured to obtain multiple sets of historical working condition information of the rotary kiln and historical temperature field data corresponding to the historical working condition information; The first training unit is configured to train a self-encoder by using a plurality of sets of historical working condition information and corresponding historical temperature field data in combination with a stack self-encoding objective function modified by mutual information. The extraction unit is configured to extract features of each of the temperature field data by using the self-encoder to obtain historical temperature feature data corresponding to each of the historical working condition information. The second training unit is configured to train a temperature feature prediction model by using a plurality of sets of historical working condition information and corresponding historical temperature feature data. The establishment unit is configured to form a rotary kiln temperature prediction model according to the temperature feature prediction model and the self-encoder.

7. An electronic device, comprising: The computer program is stored in the computer readable storage medium and is executed by the processor to implement the steps of the rotary kiln temperature prediction method according to any one of claims 1 to 5. The computer program is stored in the computer readable storage medium and is executed by the processor to implement the steps of the rotary kiln temperature prediction method according to any one of claims 1 to 5. ​ 8. A computer-readable storage medium, characterized in that, ​

Images