Temperature field detection method and device of rotary kiln, electronic equipment and storage medium

By training neural network models, combining point cloud and temperature field network models, the steady-state temperature field of industrial rotary kilns is predicted, which solves the efficiency and accuracy of temperature field detection in the existing technology, and achieves efficient and accurate temperature field acquisition .

CN118912900APending Publication Date: 2024-11-08ZHONGYE-CHANGTIAN INT ENG CO LTD
View PDF 0 Cites 1 Cited by

Patent Information

Application Number
CN202411181321.6
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-27
Publication Date
2024-11-08

AI Technical Summary

Technical Problem

The prior art is difficult to efficiently and accurately obtain the real-time temperature field of industrial rotary kilns. The traditional methods have problems such as sensor vulnerability, limited temperature measurement points, poor real-time performance and long simulation time.

Method used

By pre-training the rotary kiln temperature field prediction model based on multiple sets of historical working conditions and temperature point cloud data, the neural network model is used to combine the point cloud and temperature field network model to predict the steady-state temperature field under the current working conditions.

Benefits of technology

It realizes efficient and accurate acquisition of the rotary kiln temperature field, improves the real-time and accuracy of temperature field detection, and reduces the time-consuming of the simulation process.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118912900A_ABST
    Figure CN118912900A_ABST
Patent Text Reader

Abstract

The embodiment of the invention discloses a temperature field detection method and device of a rotary kiln, electronic equipment and a storage medium, is applied to the technical field of rotary kilns, and aims to solve the problem that the real-time temperature field of the rotary kiln cannot be efficiently obtained in real time in the prior art. The method comprises the following steps: training a rotary kiln temperature field prediction model in advance based on multiple groups of historical working conditions of the rotary kiln and temperature point cloud data corresponding to each group of historical working conditions; acquiring the current working condition of the rotary kiln; inputting the current working condition and the preset point cloud data into a rotary kiln temperature field prediction model, and analyzing the current working condition and the preset point cloud data through the rotary kiln temperature field prediction model to obtain a steady-state temperature field corresponding to the current working condition; the method can obtain the temperature field of the rotary kiln, and is high in efficiency and accuracy.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of rotary kilns, and in particular to a temperature field detection method, device, electronic equipment and computer-readable storage medium of a rotary kiln. Background Art

[0002] Industrial rotary kiln (also known as rotary kiln) is a large-scale high-temperature thermal equipment that rotates continuously. It is widely used in pillar industries of the national economy such as smelting, power generation, and building materials. Accurate acquisition of the temperature field inside the kiln is a key process parameter that determines the stable operation of high-temperature industrial process production and pollutant emissions. It is currently a bottleneck problem for traditional industries to improve quality and reduce consumption and achieve green development.

[0003] There are three main methods for constructing temperature fields: (1) Temperature sensor array. This method is to place a sensor array in the measured space to obtain temperature information at different spatial positions. However, the temperature inside the industrial rotary kiln is relatively high, and the sensor is easily damaged. Only a limited number of temperature measurement points can be arranged, and the overall three-dimensional temperature field cannot be constructed; (2) Combining a binocular camera with a thermal imaging camera. The binocular camera obtains the point cloud information of the object, and the thermal imaging camera collects the thermal imaging image, and then the two are mapped and superimposed. However, the measurement range of the industrial rotary kiln mouth is limited, and only a local temperature field can be obtained, and it is impossible to achieve measurement inside the rotary kiln; (3) ANSYS finite element three-dimensional simulation. This method establishes a material operation mechanism model in the space, determines the boundary conditions of the space, and simulates the temperature field in the kiln. This method can completely simulate the three-dimensional temperature field in the kiln, but it belongs to mechanism modeling and numerical simulation. The simulation takes a long time, has poor real-time performance, and cannot guarantee accuracy.

[0004] In view of this, how to obtain the real-time temperature field of the industrial rotary kiln more efficiently and accurately is a problem that technical personnel in this field need to solve. Summary of the invention

[0005] The purpose of the embodiments of the present invention is to provide a method, device, electronic device and computer-readable storage medium for detecting the temperature field of a rotary kiln, which can obtain the temperature field of the rotary kiln during use with high efficiency and high accuracy.

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

[0007] In one aspect, the present invention provides a method for detecting a temperature field of a rotary kiln, comprising:

[0008] Pre-training a rotary kiln temperature field prediction model based on multiple groups of historical operating conditions and temperature point cloud data corresponding to each group of historical operating conditions;

[0009] Get the current working condition of the rotary kiln;

[0010] The current working condition and the preset point cloud data are input into the rotary kiln temperature field prediction model, and the rotary kiln temperature field prediction model analyzes the current working condition and the preset point cloud data to obtain a steady-state temperature field corresponding to the current working condition.

[0011] In an exemplary embodiment, the method of training the rotary kiln temperature field prediction model based on multiple groups of historical operating conditions and temperature point cloud data corresponding to each group of historical operating conditions includes:

[0012] Obtain various historical working conditions of the rotary kiln;

[0013] A numerical simulation model is used to simulate the steady-state temperature field of the rotary kiln under each of the historical operating conditions to obtain a historical steady-state temperature field corresponding to each of the historical operating conditions;

[0014] For each of the historical steady-state temperature fields, extracting corresponding historical point cloud data and historical temperature data according to the temperature point cloud data of the steady-state temperature field;

[0015] A neural network model is trained according to each group of historical operating conditions and corresponding historical point cloud data and historical temperature data to obtain a rotary kiln temperature field prediction model; wherein the neural network model includes a point cloud network model and a temperature field network model.

[0016] In an exemplary embodiment, the training of the neural network model according to each group of historical operating conditions and the corresponding historical point cloud data and historical temperature data to obtain a rotary kiln temperature field prediction model includes:

[0017] For each group of the historical working conditions and the corresponding historical point cloud data and historical temperature data, the historical working conditions and the corresponding historical point cloud data are used as a group of historical input data, and the corresponding historical temperature data are used as a group of historical output training target data;

[0018] Inputting historical point cloud data in the historical input data into the point cloud network model, and processing the historical point cloud data through the point cloud network model to obtain corresponding point cloud features;

[0019] The point cloud features are integrated with the historical operating conditions in the historical input data and then input into a temperature network model, and the point cloud features and the historical operating conditions are processed by the temperature network model to obtain a predicted temperature field;

[0020] Extracting predicted temperature data of the predicted temperature field, and calculating a loss value according to the predicted temperature data and the historical temperature data;

[0021] The network parameters of the point cloud network model and the network parameters of the temperature network model are updated according to the loss value, and the next round of training is carried out until the preset requirements are met to obtain a trained rotary kiln temperature field prediction model.

[0022] In an exemplary embodiment, the point cloud network model includes two T-net layers, five shared MLP layers and a maximum pooling layer group.

[0023] In an exemplary embodiment, the temperature network model includes three shared MLP layers and a one-dimensional convolutional layer.

[0024] In an exemplary embodiment, the preset point cloud data is a historical point cloud data selected from each of the historical point cloud data.

[0025] In an exemplary embodiment, the historical operating conditions include a combination of one or more of the gas volume, total air volume, secondary air volume ratio, rotary kiln speed, feed volume and total coal ratio.

[0026] Another aspect of the present invention provides a temperature field detection device for a rotary kiln, comprising:

[0027] Establishing a module for pre-training a rotary kiln temperature field prediction model based on multiple groups of historical operating conditions and temperature point cloud data corresponding to each group of historical operating conditions;

[0028] An acquisition module is used to obtain the current working condition of the rotary kiln;

[0029] The prediction module is used to input the current working condition and the preset point cloud data into the rotary kiln temperature field prediction model, and the rotary kiln temperature field prediction model analyzes the current working condition and the preset point cloud data to obtain a steady-state temperature field corresponding to the current working condition.

[0030] Another aspect of the present invention provides an electronic device, comprising:

[0031] Memory for storing computer programs;

[0032] A processor is used to implement the steps of the temperature field detection method of the rotary kiln as described above when executing the computer program.

[0033] Another aspect of the present invention provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the temperature field detection method of the rotary kiln as described above are implemented.

[0034] It can be seen from the above technical solutions that the embodiments of the present invention have the following advantages:

[0035] A temperature field detection method for a rotary kiln is provided in an embodiment of the present invention, comprising: pre-training a rotary kiln temperature field prediction model based on multiple groups of historical operating conditions and temperature point cloud data corresponding to each group of historical operating conditions; obtaining the current operating condition of the rotary kiln; inputting the current operating condition and the preset point cloud data into the rotary kiln temperature field prediction model, and analyzing the current operating condition and the preset point cloud data through the rotary kiln temperature field prediction model to obtain a steady-state temperature field corresponding to the current operating condition.

[0036] It can be seen that in the embodiment of the present invention, the rotary kiln temperature field prediction model is trained in advance according to multiple groups of historical operating conditions of the rotary kiln and the temperature point cloud data corresponding to the historical operating conditions. Then, when the temperature field of the rotary kiln needs to be detected, the current operating condition of the rotary kiln is collected, and the current operating condition and the preset point cloud data are input into the rotary kiln temperature field prediction model. The steady-state temperature field of the rotary kiln under the current operating condition is obtained through the prediction of the rotary kiln temperature field prediction model. The present invention can obtain the temperature field of the rotary kiln with high efficiency and accuracy.

[0037] In addition, the present invention also provides a corresponding implementation device, electronic device and computer-readable storage medium for the temperature field detection method of the rotary kiln, which further makes the method more practical, and the device, electronic device and computer-readable storage medium have corresponding advantages.

[0038] It is to be understood that the foregoing general description and the following detailed description are exemplary only and are not restrictive of the invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the prior art and the drawings required for use in the embodiments are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0040] Figure 1 A schematic flow chart of a temperature field detection method of a rotary kiln provided in an embodiment of the present invention;

[0041] Figure 2 A schematic diagram of the architecture of temperature field detection of a rotary kiln provided by an embodiment of the present invention;

[0042] Figure 3 A schematic diagram of temperature point cloud data provided by an embodiment of the present invention;

[0043] Figure 4 A table of temperature point cloud data provided by an embodiment of the present invention;

[0044] Figure 5An overall architecture diagram of a point cloud neural network model and a temperature neural network model provided by an embodiment of the present invention;

[0045] Figure 6 A specific structural diagram of a point cloud neural network model and a temperature neural network model provided by an embodiment of the present invention;

[0046] Figure 7 A network structure diagram of an MLP provided by an embodiment of the present invention;

[0047] Figure 8 A network structure diagram of a shared MLP provided by an embodiment of the present invention;

[0048] Fig. 9 An architecture diagram of a shared MLP provided by an embodiment of the present invention;

[0049] Fig.10 An architecture diagram of a T-net module provided in an embodiment of the present invention;

[0050] Fig.11 A schematic structural diagram of a temperature field detection device for a rotary kiln provided in an embodiment of the present invention;

[0051] Fig.12 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention;

[0052] Fig.13 A schematic diagram of the structure of a computer-readable storage medium provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0053] The embodiments of the present invention provide a method, device, electronic device and computer-readable storage medium for detecting the temperature field of a rotary kiln, which can obtain the temperature field of the rotary kiln with high efficiency and accuracy during use.

[0054] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0055] Please refer to Figure 1 , Figure 1 A schematic flow chart of a temperature field detection method of a rotary kiln provided in an embodiment of the present invention. The temperature field detection method of a rotary kiln comprises:

[0056] S110: Pre-training a rotary kiln temperature field prediction model based on multiple groups of historical operating conditions of the rotary kiln and temperature point cloud data corresponding to each group of historical operating conditions;

[0057] It should be noted that in actual applications, multiple groups of historical operating conditions of the rotary kiln can be obtained, as well as the temperature point cloud data corresponding to each group of historical operating conditions, wherein the temperature point cloud data includes point cloud data and corresponding temperature data, wherein the point cloud data is the two-dimensional coordinates of each point, and the temperature data includes the temperature value corresponding to each point. Specifically, the neural network can be trained according to the multiple groups of historical operating conditions of the rotary kiln and the temperature point cloud data corresponding to each group of historical operating conditions, to obtain a trained rotary kiln temperature field prediction model. Among them, S110 can be executed once in advance, and there is no need to repeat it in the subsequent detection process. Of course, the rotary kiln temperature field prediction model can also be corrected with new sample data at intervals of a certain period of time to improve the prediction accuracy.

[0058] S120: Acquire the current working condition of the rotary kiln;

[0059] Specifically, when the temperature field of the rotary kiln needs to be predicted, the current working condition of the rotary kiln can be obtained.

[0060] S130: inputting the current working condition and the preset point cloud data into the rotary kiln temperature field prediction model, and analyzing the current working condition and the preset point cloud data by the rotary kiln temperature field prediction model to obtain a steady-state temperature field corresponding to the current working condition.

[0061] It should be noted that, since the shape and area of ​​the cross-section of the rotary kiln are stable, point cloud data corresponding to the cross-section of the rotary kiln can be pre-set. The point cloud data includes the two-dimensional coordinates of each point on the cross-section of the rotary kiln. After obtaining the current working condition of the rotary kiln, the current working condition and the preset point cloud data can be input into the rotary kiln temperature field prediction model. The current working condition and the preset point cloud data are analyzed by the rotary kiln temperature field prediction model to predict the steady-state temperature field of the rotary kiln under the current working condition.

[0062] In one embodiment, the process of pre-training the rotary kiln temperature field prediction model based on multiple groups of historical operating conditions and the temperature point cloud data corresponding to each group of historical operating conditions in S110 may include:

[0063] Obtain various historical working conditions of the rotary kiln;

[0064] The numerical simulation model is used to simulate the steady-state temperature field of the rotary kiln under various historical working conditions, and the historical steady-state temperature field corresponding to each historical working condition is obtained;

[0065] For each historical steady-state temperature field, extract the corresponding historical point cloud data and historical temperature data according to the temperature point cloud data of the steady-state temperature field;

[0066] The neural network model is trained according to each group of historical working conditions and the corresponding historical point cloud data and historical temperature data to obtain a rotary kiln temperature field prediction model; wherein the neural network model includes a point cloud network model and a temperature field network model.

[0067] It should be noted that the historical operating conditions in the embodiment of the present invention may include one or more combinations of gas volume, total air volume, secondary air volume ratio, rotary kiln speed, feed amount and total coal ratio. In practical applications, a group of 6 variable factors can be designed according to the orthogonal experimental method, which are gas volume, total air volume, secondary air volume ratio, rotary kiln speed, feed amount and total coal ratio, and a CDF simulation test with 3 or 5 levels. The level number corresponds to the specific value of the variable, and the number of experimental groups is determined according to the specific experiment. 18 groups or 25 groups, respectively, and the experimental table is shown in Table 1 below (not limited to these data in practical applications), and each group of experiments obtains a group of initial steady-state temperature fields according to the specific values ​​of the variable factors.

[0068] Table 1 Experimental table

[0069]

[0070] Specifically, the CFD calculation domain grid vertex can be regarded as a point cloud, and the spatial position of the vertex corresponds to its temperature value one by one, thereby obtaining the temperature point cloud data. In the temperature point cloud data, each grid node of the point cloud data reflects the temperature value of the node, so that each sample contains all the simulation data of the temperature field in the rotary kiln, avoiding the loss of temperature information and being more conducive to the accurate prediction of the temperature field distribution.

[0071] It should also be noted that in order to further improve the simulation accuracy in actual applications, corresponding tests can be carried out in the test kiln according to the set level. Please refer to Figure 2 The schematic diagram of the real-time temperature field construction architecture of the rotary kiln can calibrate the ANSYS numerical simulation model through the temperature data collected by the kiln head thermal imager and thermocouple to obtain the final numerical simulation model. Then, the calibrated numerical simulation model is sampled to simulate the temperature field under different historical working conditions to obtain the steady-state temperature field corresponding to each historical working condition, and then the historical point cloud data and historical temperature data of each steady-state temperature field are extracted, and each historical working condition and the corresponding historical point cloud data and historical temperature data are input into the neural network model (including the point cloud network model and the temperature field network model) to train the rotary kiln temperature field prediction model, where the model outputs the temperature field.

[0072] It is understandable that when extracting the historical point cloud data and historical temperature data of each steady-state temperature field, the extract function of the Tecplot software can be used to extract the temperature value and its coordinates on the central longitudinal section of the steady-state temperature field. The cross-sectional data and its structured grid node distribution visualization are obtained as follows: Figure 3 and Figure 4 As shown, each node is a 1*3 array, and a temperature field contains 24522 nodes.

[0073] In one embodiment, the process of training a neural network model according to each group of historical operating conditions and the corresponding historical point cloud data and historical temperature data to obtain a rotary kiln temperature field prediction model may include:

[0074] For each group of historical working conditions and the corresponding historical point cloud data and historical temperature data, the historical working conditions and the corresponding historical point cloud data are used as a group of historical input data, and the corresponding historical temperature data are used as a group of historical output training target data;

[0075] Inputting historical point cloud data in the historical input data into the point cloud network model, processing the historical point cloud data through the point cloud network model, and obtaining corresponding point cloud features;

[0076] The point cloud features are fused with the historical working conditions in the historical input data and then input into the temperature network model. The point cloud features and the historical working conditions are processed by the temperature network model to obtain the predicted temperature field.

[0077] Extract predicted temperature data of the predicted temperature field, and calculate the loss value according to the predicted temperature data and historical temperature data;

[0078] The network parameters of the point cloud network model and the temperature network model are updated according to the loss value, and the next round of training is carried out until the preset requirements are met to obtain a trained rotary kiln temperature field prediction model.

[0079] It should be noted that in practical applications, multiple groups of steady-state temperature field data can be divided into training data and test data in a ratio of 7:3. For each steady-state temperature field data in each group, the point cloud data and temperature data of the temperature field are separated, so that the historical point cloud data (wherein the historical point cloud data includes the coordinate values ​​of each point) and historical temperature data (that is, historical temperature point cloud data, which is also in the form of point cloud and includes the temperature value of each point) corresponding to each historical working condition can be obtained, and then the temperature value of each point in the historical temperature data is normalized to obtain the processed historical temperature data. Then, for each group of historical working conditions and the corresponding historical point cloud data and historical temperature data, the historical working conditions, the corresponding historical point cloud data and the corresponding historical temperature data are used as a set of historical input data.

[0080] Specifically, each group of historical input data in the training data is used to train the point cloud neural network model and the temperature neural network model to obtain the trained rotary kiln temperature field prediction model. The overall architecture of the cloud neural network model and the temperature neural network model is as follows: Figure 5 shown.

[0081] In the early stage of training, the model parameters of the point cloud network model and the model parameters of the temperature network model can be initialized, and the entire model can be iteratively trained. Specifically, a set of historical point cloud data in the historical input data can be input into the point cloud network model (i.e., the point cloud encoder model), the historical point cloud data is processed by the point cloud network model, the point cloud features are extracted, and the point cloud features corresponding to the historical point cloud data are obtained, and then the point cloud features are fused with the historical working conditions and historical temperature data in the historical input data, and the fused data is input into the temperature network model (temperature decoder and generator model), the point cloud features and the historical working conditions are processed by the temperature network model, and the training is performed with the temperature data as the training target to obtain the predicted temperature field (composed of the point cloud features and the preset temperature data), and the corresponding predicted temperature data is extracted from the predicted temperature field, and then the loss value of this round is calculated based on the predicted temperature data and the historical temperature data, and the network parameters of the point cloud network model and the network parameters of the temperature network model are updated based on the loss value, and the next round of training is performed until the preset requirements are met, and a trained rotary kiln temperature field prediction model is obtained. In practical applications, after calculating the loss value, we can first determine whether the preset requirements are met (such as whether the training rounds reach the maximum number of rounds or the loss value is less than the preset loss value). If so, we can stop updating the model parameters and obtain a trained rotary kiln temperature field prediction model.

[0082] Of course, after the training is completed, the trained rotary kiln temperature field prediction model can also be tested using test data. If the test accuracy meets the preset accuracy, the trained rotary kiln temperature field prediction model can be used as the final trained rotary kiln temperature field prediction model. If the test accuracy does not meet the preset accuracy, the training is continued until the final trained rotary kiln temperature field prediction model is obtained, and the final rotary kiln temperature field prediction model is used to predict the temperature field of the rotary kiln.

[0083] It should also be noted that since the cross-sectional shape and area of ​​the rotary kiln are stable, the various historical point cloud data are basically the same. Therefore, one historical point cloud data can be selected from the various historical point cloud data as the preset point cloud data. When predicting the temperature field, the predicted point cloud data and the current working conditions can be input into the rotary kiln temperature field prediction model. The point cloud data is output through the trained point cloud network model to obtain the corresponding point cloud features. At the same time, the point cloud features are fused with the current working conditions and input into the trained temperature network model. After prediction by the temperature network model, the corresponding steady-state temperature field is obtained.

[0084] In one embodiment, if Figure 6 As shown, the point cloud network model in the embodiment of the present invention includes two T-net layers, five shared MLP (Multilayer Perceptron) layers and a maximum pooling layer group; the temperature network model includes three shared MLP layers and a one-dimensional convolution layer Conv 1D, the convolution kernel size is 128*1, and the step length is 1. Among them, Figure 6 The shared MLP (64, 64) in the point cloud model in includes two shared MLP layers. The convolution kernel layer of each shared MLP layer has 64 convolution kernels. The shared MLP (64, 128, 1024) represents three shared MLP layers. Figure 6 The shared MLP (512, 256, 128) in the temperature field model in includes three shared MLP layers.

[0085] Specifically, Figure 6 The data processing process of the point cloud model in is as follows: the input point cloud passes through T-net to output a 2*2 transformation matrix, the transformation matrix is ​​matrix multiplied with the input point cloud to obtain the transformed n*2-dimensional data, the n*2-dimensional data passes through two shared MLPs with 64 convolution kernels to output an n*64-dimensional matrix, the n*64-dimensional matrix passes through the second T-net to obtain a 64*64 transformation matrix, the n*64-dimensional matrix is ​​multiplied with the 64*64 transformation matrix to obtain a new changed n*64-dimensional matrix, and then passes through 3 convolution kernel layers with 64, 128, and 1024 shared MLPs to obtain n*1024-dimensional matrix data, and then passes through the maximum pooling layer to reduce the dimension to 1*1024, and finally passes through 3 hidden layer nodes of 512, 256, and 128 MLPs to further extract point cloud features.

[0086] Figure 6The data processing process of the temperature field model in is as follows: the working condition is spliced ​​with the n*64-dimensional local features and n*1024-dimensional global features of the point cloud, the working condition is completed to n*6 dimensions by copying, and then the spliced ​​n*1094-dimensional matrix data is passed through three convolution kernel layers with shared MLPs of 512, 256, and 128 respectively to obtain an n*128 data matrix, and then passed through a 1D convolution layer with a convolution kernel size of 128*1 and a step number of 1 to obtain n*1-dimensional data, and finally spliced ​​with the original point cloud data to obtain the predicted temperature point cloud.

[0087] It should be noted that the common multi-layer fully connected neural network MLP is used to process one-dimensional data. The point cloud data in the embodiment of the present invention is two-dimensional data, and the temperature point cloud data is three-dimensional data. Therefore, the shared MLP is used in the embodiment of the present invention. The network structure of MLP is as follows: Figure 7 As shown, the network structure of the shared MLP is as follows Figure 8 As shown, the architecture diagram of the shared MLP is as follows Fig. 9 As shown in Figure 2, for shared MLP, some point cloud data passes through the convolution kernel and outputs one-dimensional features with the same number of convolution kernels.

[0088] In addition, point cloud data above two dimensions must be normalized by affine transformation for later processing convenience. In the embodiment of the present invention, T-net is used to predict the transformation matrix. The structure of the T-net module is shown in 10, which can include a maximum pooling layer for data dimensionality reduction and a fully connected layer for predicting the change matrix. The point cloud obtained by inputting features through the transformation matrix is ​​not affected by geometric transformation, which is more beneficial to temperature field prediction.

[0089] It can be seen that in the embodiment of the present invention, the rotary kiln temperature field prediction model is trained in advance according to multiple groups of historical operating conditions of the rotary kiln and the temperature point cloud data corresponding to the historical operating conditions. Then, when the temperature field of the rotary kiln needs to be detected, the current operating condition of the rotary kiln is collected, and the current operating condition and the preset point cloud data are input into the rotary kiln temperature field prediction model. The steady-state temperature field of the rotary kiln under the current operating condition is obtained through the prediction of the rotary kiln temperature field prediction model. The present invention can obtain the real-time temperature field of the rotary kiln with high efficiency and accuracy.

[0090] The present invention also provides a corresponding device for the temperature field detection method of the rotary kiln, which further makes the method more practical. Among them, the device can be described from the perspective of functional modules and hardware. The temperature field detection device of the rotary kiln provided by the present invention is introduced below. The device is used to implement the temperature field detection method of the rotary kiln provided by the present invention. In this embodiment, the temperature field detection device of the rotary kiln may include or be divided into one or more program modules, and the one or more program modules are stored in a storage medium and executed by one or more processors to complete the temperature field detection method of the rotary kiln disclosed in the above embodiment. The program module referred to in the present invention refers to a series of computer program instruction segments that can complete specific functions, which is more suitable for describing the execution process of the temperature field detection device of the rotary kiln in the storage medium than the program itself. The following description will specifically introduce the functions of each program module of this embodiment. The temperature field detection device of the rotary kiln described below and the temperature field detection method based on the rotary kiln described above can be referred to each other.

[0091] From the perspective of functional modules, see Fig.11 , Fig.11 The structure diagram of the temperature field detection device of the rotary kiln provided by the present invention in a specific embodiment, the device may include:

[0092] Another aspect of the present invention provides a temperature field detection device for a rotary kiln, comprising:

[0093] Establishing module 11, for training the rotary kiln temperature field prediction model based on multiple groups of historical working conditions and temperature point cloud data corresponding to each group of historical working conditions in advance;

[0094] An acquisition module 12 is used to acquire the current working condition of the rotary kiln;

[0095] The prediction module 13 is used to input the current working condition and the preset point cloud data into the pre-trained rotary kiln temperature field prediction model, and the rotary kiln temperature field prediction model analyzes the current working condition and the preset point cloud data to obtain a steady-state temperature field corresponding to the current working condition.

[0096] In one embodiment, the establishment module 11 includes:

[0097] An acquisition unit, used to acquire various historical working conditions of the rotary kiln;

[0098] A simulation unit is used to simulate the steady-state temperature field of the rotary kiln under various historical working conditions by using a numerical simulation model to obtain the historical steady-state temperature field corresponding to each historical working condition;

[0099] An extraction unit, for extracting corresponding historical point cloud data and historical temperature data according to the temperature point cloud data of the steady-state temperature field for each historical steady-state temperature field;

[0100] The training unit is used to train the neural network model according to each group of historical working conditions and the corresponding historical point cloud data and historical temperature data to obtain a rotary kiln temperature field prediction model; wherein the neural network model includes a point cloud network model and a temperature field network model.

[0101] In one embodiment, the training unit includes:

[0102] Determine a subunit for taking the historical working conditions and the corresponding historical point cloud data and the historical temperature data as a set of historical input data and taking the corresponding historical temperature data as a set of historical output training target data for each set of historical working conditions and the corresponding historical point cloud data and the historical temperature data;

[0103] A first processing subunit is used to input historical point cloud data in the historical input data into the point cloud network model, and process the historical point cloud data through the point cloud network model to obtain corresponding point cloud features;

[0104] The second processing subunit is used to fuse the point cloud features with the historical working conditions in the historical input data and input them into the temperature network model. The temperature network model processes the point cloud features and the historical working conditions to obtain a predicted temperature field.

[0105] A calculation subunit, used for extracting predicted temperature data of the predicted temperature field, and calculating the loss value according to the predicted temperature data and the historical temperature data;

[0106] The updating subunit is used to update the network parameters of the point cloud network model and the temperature network model according to the loss value, and to carry out the next round of training until the preset requirements are met to obtain a trained rotary kiln temperature field prediction model.

[0107] In one embodiment, the point cloud network model includes two T-net layers, five shared MLP layers, and a maximum pooling layer group;

[0108] In one embodiment, the temperature network model includes three shared MLP layers and a one-dimensional convolutional layer.

[0109] In one implementation, the preset point cloud data is a historical point cloud data selected from various historical point cloud data.

[0110] In one embodiment, the historical operating conditions include a combination of one or more of the gas volume, total air volume, secondary air volume ratio, rotary kiln speed, feed volume and total coal ratio.

[0111] It should be noted that the temperature field detection device of the rotary kiln provided in the embodiment of the present invention has the same beneficial effects as the temperature field detection method of the rotary kiln provided in the above embodiment, and for the specific introduction of the temperature field detection method of the rotary kiln involved in the embodiment of the present invention, please refer to the above embodiment, and this application will not repeat it here.

[0112] The temperature field detection device of the rotary kiln 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. Fig.12 A structural diagram of an electronic device provided in an embodiment of the present application, such as Fig.12 As shown, the electronic device includes: a memory 20 for storing a computer program;

[0113] The processor 21 is used to implement the steps of the temperature field detection method of the rotary kiln in the above embodiment when executing the computer program.

[0114] The electronic device provided in this embodiment may include but is not limited to a smart phone, a tablet computer, a laptop computer or a desktop computer.

[0115] Among them, the processor 21 may include one or more processing cores, such as a 4-core processor, an 8-core processor, etc. The processor 21 can be implemented in at least one hardware form of 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 is a processor for processing data in the awake state, also known as a CPU (Central Processing Unit); the coprocessor is a low-power processor for processing data in the standby state. In some embodiments, the processor 21 may be integrated with a GPU (Graphics Processing Unit), which is responsible for rendering and drawing the content to be displayed on the display screen. In some embodiments, the processor 21 may also include an AI (Artificial Intelligence) processor, which is used to process computing operations related to machine learning.

[0116] The memory 20 may include one or more computer-readable storage media, which may be non-transitory. The memory 20 may also include a high-speed random access memory, and a non-volatile memory, such as one or more disk storage devices, flash memory storage devices. In some embodiments, the memory 20 may be an internal storage unit of an electronic device, such as a hard disk of a server. In other embodiments, the memory 20 may also be an external storage device of an electronic device, such as a plug-in hard disk equipped on a server, a smart memory card (Smart Media Card, SMC), a secure digital (SecureDigital, SD) card, a flash card (Flash Card), etc. Further, the memory 20 may also include both an internal storage unit of an electronic device and an external storage device. The memory 20 may not only be used to store application software and various types of data installed in the electronic device, such as: the code of the program in the process of executing the temperature field detection method of the rotary kiln, etc., but also be used to temporarily store data that has been output or is to be output. In this embodiment, the memory 20 is at least used to store the following computer program 201, wherein, after the computer program is loaded and executed by the processor 21, the relevant steps of the temperature field detection method of the rotary kiln disclosed in any of the aforementioned embodiments can be implemented. In addition, the resources stored in the memory 20 may also include an operating system 202 and data 203, etc., 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 field detection result of the rotary kiln) and the like.

[0117] 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. Among them, the display screen 22 and the input / output interface 23, such as a keyboard, belong to the user interface, and the optional user interface may also include a standard wired interface, a wireless interface, and the like. Optionally, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, and an OLED (Organic Light-Emitting Diode) touch device, and the like. The display may also be appropriately referred to as a display screen or a display unit, which is 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, and the like, which are generally used to establish a communication connection 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, and the like. The bus may be divided into an address bus, a data bus, a control bus, and the like. For ease of representation, Fig.12 Only one thick line is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0118] Those skilled in the art will understand that Fig.12 The structure shown in the figure does not constitute a limitation on the electronic device, and may include more or fewer components than shown in the figure.

[0119] It is understandable that if the temperature field detection method of the rotary kiln in the above embodiment is implemented in the form of 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 the present application is essentially 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, and the computer software product is stored in a storage medium to execute all or part of the steps of the various embodiments of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, 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 that can store program codes.

[0120] Based on this, Fig.13As shown, the embodiment of the present invention further provides a computer-readable storage medium, on which a computer program 31 is stored. When the computer program 31 is executed by a processor, the steps of the temperature field detection method of the rotary kiln are implemented.

[0121] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0122] It should also be noted that, in this specification, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "comprise", "include" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the statement "comprises a ..." does not exclude the presence of other identical elements in the process, method, article or device including the element.

[0123] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present invention. Various modifications to these embodiments will be 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 present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for detecting the temperature field of a rotary kiln, characterized in that: include: Pre-training a rotary kiln temperature field prediction model based on multiple groups of historical operating conditions and temperature point cloud data corresponding to each group of historical operating conditions; Get the current working condition of the rotary kiln; The current working condition and the preset point cloud data are input into the rotary kiln temperature field prediction model, and the rotary kiln temperature field prediction model analyzes the current working condition and the preset point cloud data to obtain a steady-state temperature field corresponding to the current working condition.

2. The temperature field detection method of a rotary kiln according to claim 1, characterized in that: The method of training the rotary kiln temperature field prediction model based on multiple groups of historical operating conditions and temperature point cloud data corresponding to each group of historical operating conditions comprises: Obtain various historical operating conditions of the rotary kiln; A numerical simulation model is used to simulate the steady-state temperature field of the rotary kiln under each of the historical operating conditions to obtain a historical steady-state temperature field corresponding to each of the historical operating conditions; For each of the historical steady-state temperature fields, extracting corresponding historical point cloud data and historical temperature data according to the temperature point cloud data of the steady-state temperature field; A neural network model is trained according to each group of historical operating conditions and corresponding historical point cloud data and historical temperature data to obtain a rotary kiln temperature field prediction model; wherein the neural network model includes a point cloud network model and a temperature field network model.

3. The temperature field detection method of a rotary kiln according to claim 2, characterized in that: The training of the neural network model according to each group of historical working conditions and the corresponding historical point cloud data and historical temperature data to obtain the rotary kiln temperature field prediction model includes: For each group of the historical working conditions and the corresponding historical point cloud data and historical temperature data, the historical working conditions and the corresponding historical point cloud data are used as a group of historical input data, and the corresponding historical temperature data are used as a group of historical output training target data; Inputting historical point cloud data in the historical input data into the point cloud network model, and processing the historical point cloud data through the point cloud network model to obtain corresponding point cloud features; The point cloud features are integrated with the historical operating conditions in the historical input data and then input into a temperature network model, and the point cloud features and the historical operating conditions are processed by the temperature network model to obtain a predicted temperature field; Extracting predicted temperature data of the predicted temperature field, and calculating a loss value according to the predicted temperature data and the historical temperature data; The network parameters of the point cloud network model and the network parameters of the temperature network model are updated according to the loss value, and the next round of training is carried out until the preset requirements are met to obtain a trained rotary kiln temperature field prediction model.

4. The temperature field detection method of a rotary kiln according to claim 3, characterized in that: The point cloud network model includes two T-net layers, five shared MLP layers and a maximum pooling layer group.

5. The temperature field detection method of a rotary kiln according to claim 3, characterized in that: The temperature network model includes three shared MLP layers and a one-dimensional convolutional layer.

6. The temperature field detection method of a rotary kiln according to any one of claims 2 to 5, characterized in that: The preset point cloud data is a historical point cloud data selected from each of the historical point cloud data.

7. The temperature field detection method of a rotary kiln according to claim 6, characterized in that: The historical operating conditions include one or more combinations of gas volume, total air volume, secondary air volume ratio, rotary kiln speed, feed volume and total coal ratio.

8. A temperature field detection device for a rotary kiln, characterized in that: include: Establishing a module for pre-training a rotary kiln temperature field prediction model based on multiple groups of historical operating conditions and temperature point cloud data corresponding to each group of historical operating conditions; An acquisition module is used to obtain the current working condition of the rotary kiln; The prediction module is used to input the current working condition and the preset point cloud data into the rotary kiln temperature field prediction model, and the rotary kiln temperature field prediction model analyzes the current working condition and the preset point cloud data to obtain a steady-state temperature field corresponding to the current working condition.

9. An electronic device, characterized in that: include: Memory for storing computer programs; A processor is used to implement the steps of the temperature field detection method of the rotary kiln as claimed in any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the temperature field detection method of the rotary kiln as claimed in any one of claims 1 to 7 are implemented.

Citation Information

Cited By

  • Fuel adding control method and system for aluminum slag solid waste rotary kiln

    CN120740306A