Soft measurement method and equipment for free calcium content of cement clinker and storage medium

Abstract

The embodiment of the invention provides a soft measurement method and device for the free calcium content of cement clinker and a storage medium. And carrying out free calcium content soft measurement on the cement clinker output by the grate cooler at the current time by utilizing a free calcium content soft measurement model based on the variable data sequence, the kiln temperature information of the rotary kiln, the kiln head observation image sequence and other multi-mode data. Therefore, a soft measurement mode of the free calcium content of the cement clinker based on the multi-modal data is provided, the information source of the soft measurement model of the free calcium content is enriched by the multi-modal data, the soft measurement of the free calcium content of the cement clinker can be automatically, quickly and accurately performed by the mode, and the effect of the soft measurement of the free calcium content is improved.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to a soft measurement method, device and storage medium for free calcium content of cement clinker. Background Art

[0002] In the cement industry, free calcium in cement clinker is an important indicator for judging the quality of cement clinker. In the traditional cement production model, the free calcium content of cement clinker is tested by manual offline testing or automatic testing machine offline testing, that is, the cement clinker is sampled and tested every hour or so to detect the free calcium content of cement clinker. However, the offline testing method has the disadvantages of long detection cycle, poor real-time performance, low efficiency and accuracy that is easily affected by sampling and testing methods, resulting in the test results of the free calcium content of cement clinker being unable to guide the real-time control and optimization of the cement clinker burning process. Therefore, it is very necessary to quickly and accurately detect the free calcium content of cement clinker. Summary of the invention

[0003] Multiple aspects of the present application provide a soft measurement method, device and storage medium for free calcium content of cement clinker, which are used to quickly and accurately detect the free calcium content of cement clinker.

[0004] The embodiment of the present application provides a soft measurement method for free calcium content of cement clinker, comprising: if a grate cooler outputs cement clinker at the current time, obtaining a variable data sequence, wherein the variable data sequence includes variable data of multiple production process variables related to the grate cooler within a first time period starting from a first time and variable data of multiple production process variables related to a designated device within a first time period starting from a second time, the first time is the time when the cement clinker enters the grate cooler, the second time is the time when the cement raw material enters the rotary kiln, and the designated device performs the cooling process performed by the grate cooler. The preceding process of the sequence, the designated equipment at least includes a rotary kiln; obtaining the kiln temperature information of the rotary kiln and a kiln head fire viewing image sequence within a first time period starting from a second time, wherein the kiln head fire viewing image sequence includes a plurality of kiln head fire viewing images, and the kiln head fire viewing image is obtained by collecting images of the internal combustion flame state of the rotary kiln; inputting the variable data sequence, the kiln temperature information of the rotary kiln and the kiln head fire viewing image sequence into a free calcium content soft measurement model, so as to perform a soft measurement of the free calcium content of the cement clinker output by the grate cooler at the current time through the free calcium content soft measurement model.

[0005] The embodiment of the present application also provides a model training method, including: obtaining a training data set, wherein the training data set includes multiple training data, and the training data include: a variable data sequence corresponding to the cement clinker output by the grate cooler at a historical time, the kiln temperature information of the rotary kiln, a kiln head fire viewing image sequence and a free calcium content test result; using the training data set to perform model training on a free calcium content soft measurement model to be trained, and obtaining a trained free calcium content soft measurement model, wherein the free calcium content soft measurement model is used to perform free calcium content soft measurement on the cement clinker output by the grate cooler at the current time.

[0006] An embodiment of the present application also provides an electronic device, comprising: a memory and a processor; the memory is used to store a computer program; the processor is coupled to the memory, and is used to execute the computer program to execute the steps of a soft measurement method or a medium model training method for free calcium content of cement clinker.

[0007] An embodiment of the present application also provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the processor is enabled to implement the steps in the soft measurement method or the model training method for the free calcium content of cement clinker.

[0008] The embodiment of the present application provides a method, device and storage medium for soft measurement of free calcium content of cement clinker. In the embodiment of the present application, when it is necessary to measure the free calcium content of cement clinker outputted by the grate cooler at the current time, firstly obtain a variable data sequence, kiln temperature information of the rotary kiln within a specified time period from the time when the cement raw material enters the rotary kiln, and a kiln head fire image sequence, wherein the variable data sequence includes variable data of multiple production process variables related to the grate cooler from the time when the cement clinker enters the grate cooler and variable data of multiple production process variables related to the specified equipment from the time when the cement raw material enters the rotary kiln. Then, the free calcium content soft measurement model is used to perform soft measurement of the free calcium content of cement clinker outputted by the grate cooler at the current time based on multimodal data such as the variable data sequence, kiln temperature information of the rotary kiln and the kiln head fire image sequence. Therefore, a soft measurement method for the free calcium content of cement clinker based on multimodal data is provided. The multimodal data enriches the information source of the soft measurement model of the free calcium content. The above method can automatically, quickly and accurately perform soft measurement of the free calcium content of cement clinker, thereby improving the effect of the soft measurement of the free calcium content, and can better provide guidance for the optimization of the equipment operating parameters of the subsequent cement clinker burning system. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0010] Figure 1 is a schematic structural diagram of an exemplary cement clinker burning system;

[0011] Figure 2 A flow chart of a soft measurement method for free calcium content of cement clinker provided in an embodiment of the present application;

[0012] Figure 3 It is the side view of rotary kiln;

[0013] Figure 4 This is a cross-sectional view of a rotary kiln;

[0014] Figure 5 A model structure diagram of an exemplary free calcium content soft-sensing model provided in an embodiment of the present application;

[0015] Figure 6 A flowchart of a model training method provided in an embodiment of the present application;

[0016] Figure 7 A schematic diagram of the structure of a soft measurement device for free calcium content in cement clinker provided in an embodiment of the present application;

[0017] Figure 8 A schematic diagram of the structure of a model training device provided in an embodiment of the present application;

[0018] Fig. 9 A schematic diagram of the structure of an electronic device provided in this application. DETAILED DESCRIPTION

[0019] In order to make the purpose, technical solution and advantages of the present application clearer, the technical solution of the present application will be clearly and completely described below in combination with the specific embodiments of the present application and the corresponding drawings. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present application.

[0020] In the embodiments of the present application, "at least one" refers to one or more, and "more than one" refers to two or more. "And / or" describes the access relationship of associated objects, indicating that three relationships may exist. For example, A and / or B may represent: A exists alone, A and B exist at the same time, and B exists alone, where A and B may be singular or plural. In the text description of the present application, the character " / " generally indicates that the previous and next associated objects are in an "or" relationship. In addition, in the embodiments of the present application, "first", "second", "third", etc. are only used to distinguish the contents of different objects and have no other special meanings.

[0021] The following is an introduction to some terms provided in the embodiments of the present application:

[0022] Soft measurement: It is to organically combine the knowledge of the production process, apply computer technology to important variables that are difficult to measure or temporarily unmeasurable, select other easily measurable variables, infer or estimate them by forming certain mathematical relationships, and replace the functions of hardware with software.

[0023] Cement clinker: A semi-finished product obtained by mixing limestone, clay and iron raw materials into raw materials in appropriate proportions, burning them until they are partially or completely melted, and then cooling them.

[0024] Cement raw material: The material which is mixed in proportion and ground to a certain fineness by lime raw materials, clay raw materials and a small amount of correction raw materials (sometimes mineralizers, crystal seeds, etc. are also added, and coal is also added in vertical kiln production) is called cement raw material.

[0025] Production process variables: refers to the variables that need to be paid attention to during the production process. For example, production process variables include but are not limited to: raw material testing parameters, outlet temperature of the decomposition furnace, secondary air temperature, temperature of the kiln tail smoke chamber, raw material feed rate, coal feed rate of the decomposition furnace, coal feed rate of the rotary kiln, and air intake of the rotary kiln.

[0026] Mechanism refers to the reason and logic of the change of things. From the perspective of the concept of mechanism, the mechanism includes two aspects: the formation elements and the relationship between the formation elements. Its principle refers to the internal working mode of each element in a certain system structure to achieve a specific function, as well as the operating rules and principles of the elements' mutual connection and interaction under certain environmental conditions.

[0027] Mechanism model (white box): An accurate mathematical model established based on the internal mechanism of an object or production process or the transfer mechanism of a material flow. It is a mathematical model of an object or process based on mass balance equations, energy balance equations, momentum balance equations, phase balance equations, and certain physical property equations, chemical reaction laws, and basic circuit laws. The advantage of the mechanism model is that the parameters have very clear physical meanings. This type of model exists in all walks of life and requires sufficient input conditions. The output is obtained through the model and the entire process can be simulated.

[0028] Feature Engineering is the process of converting raw data into feature vectors that better express the essence of the problem. The purpose is to extract feature vectors from the raw data to the maximum extent possible for use by algorithms and models.

[0029] Figure 1 FIG. 1 is a schematic diagram of an exemplary cement clinker burning system. Figure 1, the cement clinker burning system includes a raw material bin, a preheater, a decomposition furnace, a pulverized coal bin, a rotary kiln and a grate cooler, etc. The cement clinker burning process is roughly as follows: the raw material powder (i.e., cement raw material) falls from the raw material bin into the preheater for preheating, and the preheated raw material powder falls from the preheater into the decomposition furnace for decomposition. The decomposed cement raw material falls from the decomposition furnace into the rotary kiln for calcination to calcine the cement raw material into cement clinker, wherein the rotary kiln is turned circumferentially under the drive of its motor. The cement clinker falls from the rotary kiln into the grate cooler for cooling, and the cooled cement clinker is transported to the clinker storage for storage. In addition, pulverized coal is respectively input from the pulverized coal bin into the decomposition furnace and the rotary kiln to be burned in the decomposition furnace and the rotary kiln.

[0030] In the traditional cement production mode, the free calcium content of cement clinker is tested by manual offline testing or automatic testing machine offline testing, that is, the free calcium content of cement clinker is tested by sampling and testing cement clinker every hour or so. However, the offline testing method has the disadvantages of long testing cycle, poor real-time performance, low efficiency and accuracy that is easily affected by sampling and testing methods, resulting in the test results of free calcium content of cement clinker being unable to guide the real-time control and optimization of cement clinker burning process.

[0031] To this end, the embodiment of the present application provides a method, device and storage medium for soft measurement of free calcium content of cement clinker. In the embodiment of the present application, when it is necessary to measure the free calcium content of cement clinker output by the grate cooler at the current time, first obtain a variable data sequence, kiln temperature information of the rotary kiln within a specified time period from the time when the cement raw material enters the rotary kiln, and a kiln head fire image sequence, wherein the variable data sequence includes variable data of multiple production process variables related to the grate cooler from the time when the cement clinker enters the grate cooler and variable data of multiple production process variables related to the specified equipment from the time when the cement raw material enters the rotary kiln. Then, the free calcium content soft measurement model is used to perform soft measurement of the free calcium content of cement clinker output by the grate cooler at the current time based on multimodal data such as the variable data sequence, the kiln temperature information of the rotary kiln and the kiln head fire image sequence. Therefore, a soft measurement method for the free calcium content of cement clinker based on multimodal data is provided. The multimodal data enriches the information source of the soft measurement model of the free calcium content. The above method can automatically, quickly and accurately perform soft measurement of the free calcium content of cement clinker, thereby improving the effect of the soft measurement of the free calcium content, and can better provide guidance for the optimization of the equipment operating parameters of the subsequent cement clinker burning system.

[0032] In addition, the soft measurement of free calcium content in cement clinker faces many challenges, including multiple time scales, dynamic time lags and strong nonlinearity. The traditional soft measurement of free calcium content in cement clinker is difficult to promote and apply. Specifically, first, in the traditional cement production mode, the sampling frequency of free calcium content in cement clinker is at the hourly level, that is, sampling and testing once an hour, while the sampling frequency of production process variables in the cement clinker burning system can reach the minute level or even higher. Therefore, the soft measurement modeling of free calcium content in cement clinker is a multi-time scale problem; secondly, there are dynamic time lags between production process variables in the cement clinker burning system. Among them, the time lag from the entrance of the rotary kiln to the exit of the rotary kiln (about 30 minutes) and the time lag from the exit of the rotary kiln to the exit of the grate cooler (about 20 minutes) are the main ones; therefore, it is necessary to match the test value of free calcium content and the production process variables in time series. Finally, the physical and chemical reactions in the cement clinker burning process are coupled with each other, the thermal state is complex, and nonlinearity is prominent.

[0033] The soft measurement method for free calcium content based on the free calcium content soft measurement model provided in the embodiment of the present application can effectively deal with the problems of multiple time scales, dynamic time lag and strong nonlinearity in the soft measurement of free calcium content, and has good promotion value.

[0034] The technical solutions provided by various embodiments of the present application are described in detail below in conjunction with the accompanying drawings.

[0035] Figure 2 This is a flow chart of a soft measurement method for free calcium content in cement clinker provided in an embodiment of the present application. Figure 2 , the method may include the following steps:

[0036] 201. If the grate cooler outputs cement clinker at the current time, obtain the variable data sequence.

[0037] In this embodiment, the free calcium content soft measurement of the cement clinker output by the grate cooler at the current time can be triggered periodically, regularly or when the set trigger condition is met, but it is not limited to this. When it is necessary to perform soft measurement of the free calcium content of the cement clinker output by the grate cooler at the current time, a variable data sequence is obtained. Among them, the variable data sequence includes variable data of multiple production process variables related to the grate cooler within a first time period from a first time and variable data of multiple production process variables related to a specified device within a first time period from a second time. The first time is the time when the cement clinker enters the grate cooler, and the second time is the time when the cement raw material enters the rotary kiln.

[0038] It can be understood that when it is determined that it is necessary to perform soft measurement on the free calcium content of the cement clinker output by the grate cooler at the current time, by finding the variable data of each production process variable after the cement clinker enters the grate cooler and the variable data of each production process variable after the cement raw material for preparing the cement clinker enters the rotary kiln, the variable data of each production process variable with time series matching is used to perform soft measurement of the free calcium content, which effectively solves the dynamic time lag problem existing in soft measurement.

[0039] In this embodiment, multiple production process variables related to the grate cooler include, for example, but are not limited to: the current, air volume and air pressure of the fan equipped with the grate cooler, the grate pressure under the grate cooler, etc.

[0040] In this embodiment, the designated device performs the preceding process of the cooling process performed by the grate cooler, and the designated device includes at least a rotary kiln. Further optionally, in order to improve the soft measurement effect, the designated device may also include, but is not limited to: a preheater and a decomposition furnace. The preheating process performed by the preheater, the decomposition process performed by the decomposition furnace, and the calcination process performed by the rotary kiln are all preceding processes of the cooling process performed by the grate cooler.

[0041] In this embodiment, multiple production process variables related to the rotary kiln include, but are not limited to: secondary air temperature of the rotary kiln, temperature of the kiln tail smoke chamber of the rotary kiln, raw material feed amount of the rotary kiln, coal feed amount of the rotary kiln, air intake amount of the rotary kiln, kiln head hood pressure of the rotary kiln, etc. Multiple production process variables related to the preheater include, but are not limited to: preheater temperature at each level, preheater outlet temperature. Multiple production process variables related to the decomposition furnace include, but are not limited to: outlet temperature of the decomposition furnace and coal feed amount of the decomposition furnace.

[0042] In practical applications, more variable data of production process variables can be added to the variable data sequence, and the addition can be flexibly done as needed. For example, the variable data sequence can also add cement raw material testing parameters, etc.

[0043] In this embodiment, the time is sorted from early to late, which is: the second time when the cement raw material enters the rotary kiln, the first time when the cement clinker enters the grate cooler, and the current time when the cement clinker is currently output by the grate cooler.

[0044] In this embodiment, data mining can be performed on the time when a large amount of cement clinker enters the grate cooler and the time when the cement clinker is discharged from the grate cooler to obtain the corresponding relationship between the time when the cement clinker enters the grate cooler and the time when the cement clinker is discharged from the grate cooler, and the first time corresponding to the current time is determined based on the corresponding relationship. Alternatively, data mining can be performed on the time when a large amount of cement clinker enters the grate cooler and the time when the cement clinker is discharged from the grate cooler to determine the average residence time of the cement clinker in the grate cooler. The first time is calculated based on the current time and the average residence time. In some optional implementations, in order to more accurately determine the first time, the first time can be determined using a mechanism analysis method. Exemplarily, the third time spent by the grate cooler to cool the cement clinker can be determined based on the average value of the grate speed of the grate cooler in the second time period before the current time; the first time is determined based on the current time and the third time period. Among them, the second time period is flexibly set as needed.

[0045] Specifically, the correlation between the average value of the grate speed of the grate cooler and the third duration is determined based on the results of the mechanism analysis. Assume that the correlation between the two is expressed by a functional relationship. The functional relationship is recorded as y=f(x), where x refers to the independent variable and y refers to the dependent variable. Here, the average value of the grate speed of the grate cooler is used as the independent variable, and the third duration is used as the dependent variable. Specifically, taking the inverse proportional relationship as an example, after the average value of the grate speed of the grate cooler is determined, it is not difficult to determine the third duration based on the inverse proportional relationship.

[0046] In this embodiment, data mining can be performed on the time when a large amount of cement raw materials enter the rotary kiln and the time when cement clinker is discharged from the rotary kiln to obtain the corresponding relationship between the time when the cement raw materials enter the rotary kiln and the time when the cement clinker is discharged from the rotary kiln, and the second time corresponding to the first time (which can also be regarded as the time when the cement clinker is discharged from the rotary kiln) is determined based on the corresponding relationship. Alternatively, data mining can be performed on the time when a large amount of cement raw materials enter the rotary kiln and the time when the cement clinker is discharged from the rotary kiln to determine the time it takes for the cement raw materials to be prepared into cement clinker. The second time is calculated based on the first time and the time it takes for the cement raw materials to be prepared into cement clinker.

[0047] In some optional implementations, in order to more accurately determine the second time, the second time may be determined using a mechanism analysis method. Exemplarily, the fifth time taken by the rotary kiln to prepare cement raw materials into cement clinker is determined based on the average value of the rotation speed of the rotary kiln in the fourth time period before the first time; the second time is determined based on the first time and the fifth time period. The fourth time period is flexibly set as needed, and the rotation speed of the rotary kiln specifically refers to the rotation speed of the drum of the rotary kiln.

[0048] Specifically, the correlation between the average value of the rotation speed of the rotary kiln and the fifth time length is determined based on the results of the mechanism analysis. It is assumed that the correlation between the two is expressed by a functional relationship. The functional relationship is recorded as y=f(x), where x refers to the independent variable and y refers to the dependent variable. Here, the average value of the rotation speed of the rotary kiln is used as the independent variable and the fifth time length is used as the dependent variable. After the fifth time length is obtained, the fifth time length is calculated forward based on the first time, and the second time can be determined.

[0049] As an example, in order to accurately determine the fifth time period, an optional implementation method is to determine the fifth time period consumed by the rotary kiln to prepare cement raw materials into cement clinker based on the average value of the rotation speed of the rotary kiln in the fourth time period before the first time: determine the fifth time period consumed by the rotary kiln to prepare cement raw materials into cement clinker based on the structural parameters of the rotary kiln, the average value of the rotation speed of the rotary kiln in the fourth time period before the first time, and the average value of the cement raw material feed amount.

[0050] See also Figure 3 and Figure 4 , the structural parameters of the rotary kiln include, for example, but are not limited to: the kiln length of the rotary kiln, the inner radius of the rotary kiln, the kiln inclination angle of the rotary kiln, the material bed height of the rotary kiln (also referred to as the material layer thickness), the material bed center angle of the material bed height of the rotary kiln, and the natural angle of repose of the material. Among them, the material layer thickness can be understood as the thickness of the free inclined surface of the material accumulation layer. Assuming that the kiln length of the rotary kiln is recorded as L, the kiln length of the rotary kiln refers to the effective length of the barrel of the rotary kiln. Assuming that the inner radius of the rotary kiln is recorded as R, the inner radius of the rotary kiln refers to the radius of the kiln barrel of the rotary kiln. Assuming that the kiln inclination angle of the rotary kiln is recorded as β, the kiln inclination angle of the rotary kiln refers to the inclination of the rotary kiln, that is, the angle between the axis of the barrel of the rotary kiln and the horizontal line. Assuming that the natural angle of repose of the material is recorded as γ, the natural angle of repose of the material refers to the maximum angle formed by the free inclined surface of the material accumulation layer and the horizontal plane. Among them, the kiln length L of the rotary kiln, the inner radius R of the rotary kiln, the kiln inclination angle β of the rotary kiln, and the material repose angle γ are all known quantities (that is, known constants). Assume that the material bed height of the rotary kiln is recorded as h, and the center angle of the material bed of the rotary kiln is recorded as θ. The material bed height h of the rotary kiln is calculated by the cement raw material feed amount m and the rotary kiln speed n. The center angle θ of the material bed of the rotary kiln is calculated by the material bed height h of the rotary kiln. The cement raw material feed amount m and the rotary kiln speed n are input parameters. See Figure 4 The center angle θ of the material bed refers to the center angle of the material bed corresponding to the circumference of the material bed with the center point of the rotary kiln cross section as the center.

[0051] In some optional embodiments, when determining the time lag in the rotary kiln, that is, determining the time taken by the rotary kiln to prepare cement raw materials into cement clinker, first, it is necessary to calculate the material bed height h by the cement raw material feed amount m and the rotary kiln speed n, that is:

[0052]

[0053] Among them, ρ s Refers to the density of the material.

[0054] By transforming formula (1), we can get the following formula:

[0055]

[0056] After calculating the height of the material bed, we can further calculate the central angle θ of the material bed and the chord length L of the material bed. c ;

[0057] θ=2acos((Rh) / R)……(3)

[0058]

[0059] Under the assumption that the average residence time of the material in the rotary kiln (denoted as MRT) is equal to the ratio between the total mass M of the material in the rotary kiln and the cement raw material feed rate per unit time (denoted as m), based on the above, the average residence time is:

[0060]

[0061] It is worth noting that the differential and integral forms of formulas (1) to (4) are intended to describe the bed morphology when the bed height is unevenly distributed along the kiln length. In practical applications, the differential and integral forms can be simplified to directly obtain the average parameters of the bed morphology.

[0062] In summary, based on the cement raw material feed amount m, the kiln speed n of the rotary kiln and various structural size parameters of the rotary kiln, the average residence time of the material in the rotary kiln can be calculated, that is, the time it takes for the rotary kiln to prepare cement raw material into cement clinker.

[0063] In this embodiment, the first duration is flexibly set as needed, for example, 1 hour. The time when the high-temperature cement clinker falls into the grate cooler is used as the starting time, and variable data of multiple production process variables related to the grate cooler within the first duration are collected. The time when the cement raw material enters the rotary kiln is used as the starting time, and variable data of multiple production process variables related to the rotary kiln within the first duration are collected. Further, optionally, variable data of multiple production process variables related to the decomposition furnace or preheater within the first duration can also be collected.

[0064] In practical applications, the variable data of some production process variables can be collected and acquired through corresponding sensors. Sensors include, but are not limited to: temperature sensors, air volume sensors, pressure sensors, coal volume sensors, flow sensors, current sensors, displacement sensors, and quality sensors. The variable data of some production process variables rely on manual input or automatic testing machine upload, for example, raw material testing parameters.

[0065] Further optionally, in order to improve the efficiency of soft measurement, multiple sensors set in the cement clinker production system can be periodically controlled to collect data, so as to obtain and save the variable data of the production process variables collected by the multiple sensors at each sampling time. In this way, when obtaining the variable data sequence, the variable data of multiple production process variables related to the grate cooler within the first time period from the first time and the variable data of multiple production process variables related to the specified equipment within the first time period from the second time can be obtained from the variable data of the production process variables collected by the multiple sensors at each sampling time.

[0066] In some optional application scenarios, some variable data of production process variables that are not collected by sensors are also required, for example, variable data of production process variables manually entered or uploaded by automatic inspection machines. In practical applications, the variable data of these production process variables that are not collected by sensors can be maintained and managed in the database. In this way, when soft measurement needs to add variable data of such production process variables to the variable data sequence, it can be obtained from the database and added to the variable data sequence.

[0067] In practical applications, for the variable data of multiple production process variables collected in each cycle, before saving the variable data of multiple production process variables collected in the cycle, the variable data of multiple production process variables collected in the cycle can be cleaned. Data cleaning includes, but is not limited to: data alignment, missing value processing, abnormal value processing, duplicate value removal and data conversion, etc. It can be understood that data cleaning can remove some dirty data, delete duplicate information, correct existing errors and ensure data consistency.

[0068] 202. Acquire kiln temperature information of the rotary kiln and a sequence of kiln head fire viewing images within a first time period starting from a second time.

[0069] In this embodiment, the kiln temperature information of the rotary kiln within the first time period is collected with the time when the cement raw material enters the rotary kiln as the starting time. The kiln temperature information of the rotary kiln includes, but is not limited to: the cement clinker temperature inside the rotary kiln, the flue gas temperature on the kiln head side of the rotary kiln, the inner wall temperature of the rotary kiln, and the outer surface temperature of the barrel of the rotary kiln. In some scenarios, the kiln temperature information of the rotary kiln within the first time period from the second time may refer to the average value of the kiln temperature information of the rotary kiln within the first time period, or the kiln temperature information of the rotary kiln of each time unit within the first time period, and the time unit is, for example, milliseconds, seconds, or minutes.

[0070] In practical applications, the temperature sensor can be used to periodically collect the kiln temperature information of the rotary kiln. In this way, the kiln temperature information of the rotary kiln within the first time period starting from the second time can be obtained from the kiln temperature information of the rotary kiln stored at each collection time.

[0071] In some optional embodiments, in order to efficiently and accurately obtain the kiln temperature information of the rotary kiln within the first time period starting from the second time, the average value of the variable data of multiple production process variables related to the designated equipment within the first time period starting from the second time can also be input into the rotary kiln temperature mechanism model to obtain the kiln temperature information of the rotary kiln output by the rotary kiln temperature mechanism model.

[0072] Specifically, for each of the multiple production process variables related to the designated equipment, the variable data of the production process variable related to the designated equipment within the first time period starting from the second time is averaged to obtain the average value of the variable data of the production process variable related to the designated equipment within the first time period starting from the second time.

[0073] The rotary kiln temperature mechanism model is a mechanism model for detecting the kiln temperature information of the rotary kiln. The model is constructed by comprehensively considering factors such as the material movement state, heat transfer state, and physicochemical reaction inside the rotary kiln. The model uses the average value of the variable data of multiple production process variables as input parameters and the kiln temperature information of the rotary kiln as output results. Multiple production process variables include, but are not limited to: raw material test parameters, outlet temperature of the decomposition furnace, secondary air temperature, temperature of the kiln tail smoke chamber, raw material feed amount, coal feed amount of the decomposition furnace, coal feed amount of the rotary kiln, and air intake of the rotary kiln.

[0074] In this embodiment, on the one hand, the rotary kiln temperature mechanism model performs feature engineering on the variable data of the original production process variables, thereby generating new features with high information content (kiln temperature information of the rotary kiln) for free calcium soft measurement. On the other hand, since the rotary kiln temperature mechanism model is highly sensitive to the input parameters, the output result (i.e., the kiln temperature information of the rotary kiln) can also be used as a reference for abnormal value detection and processing.

[0075] In this embodiment, the time when the cement raw material enters the rotary kiln is taken as the starting time, and the kiln head fire image sequence within the first time length is collected. The kiln head fire image sequence includes a plurality of kiln head fire images, which are obtained by collecting images of the internal combustion flame state of the rotary kiln. Figure 1 In practical applications, the image acquisition device disposed at the kiln head side of the rotary kiln can be used to acquire images of the internal combustion flame state of the rotary kiln to obtain the kiln head fire image. The image acquisition device can be any device with an image acquisition function. For example, the image acquisition device can be divided into a surface array camera and a line array camera according to the structural characteristics of the sensor. For another example, the image acquisition device can be divided into a standard definition camera and a high definition camera according to the different picture resolutions. For another example, the image acquisition device can be an analog camera and a digital camera according to the signal classification.

[0076] Further optionally, in order to improve the efficiency of soft measurement, images of the internal combustion flame state of the rotary kiln can be periodically collected to obtain and save the kiln head fire viewing images collected at each sampling time; in this way, when it is necessary to obtain a sequence of kiln head fire viewing images within the first time period starting from the second time, the kiln head fire viewing image sequence within the first time period starting from the second time can be obtained from the saved kiln head fire viewing images collected at each sampling time.

[0077] In practical applications, for the kiln head fire-watching images collected in each period, before saving the kiln head fire-watching images collected in the period, various image processing such as cropping, scaling, Gamma transformation and linear transformation processing can be performed on the kiln head fire-watching images to convert them into W*H*C RGB format images, thereby minimizing data redundancy. Among them, Gamma transformation is a nonlinear operation on the grayscale value of the input image, so that the grayscale value of the output image is exponentially related to the grayscale value of the input image. W is the image width, H is the image height, C is the number of channels, W*H*C is 20*20*3, for example, RGB represents the colors of the three channels of red, green and blue.

[0078] 203. Input the variable data sequence, the kiln temperature information of the rotary kiln and the kiln head fire viewing image sequence into the free calcium content soft measurement model, so as to perform soft measurement of the free calcium content of the cement clinker output by the grate cooler at the current time through the free calcium content soft measurement model.

[0079] In the present embodiment, a free calcium content soft measurement model with a soft measurement function of free calcium content of cement clinker is pre-trained. Free calcium content soft measurement is an end-to-end free calcium soft measurement framework of cement clinker, which can effectively solve the problems of multiple time scales, dynamic time lag and strong nonlinearity in the soft measurement of free calcium content of cement clinker. The variable data sequence, the kiln temperature information of the rotary kiln and the kiln head fire image sequence are processed by the free calcium content soft measurement model to obtain the free calcium content soft measurement result of the cement clinker output by the grate cooler at the current time. For example, the variable data sequence includes the variable data of each production process variable in each of 60 minutes within 1 hour, the kiln temperature information of the rotary kiln includes the kiln temperature information of the rotary kiln in each of 60 minutes within 1 hour, and the kiln head fire image sequence includes 20*20*3 kiln head fire images in each of 60 minutes within 1 hour.

[0080] In this embodiment, the network structure of the free calcium content soft-sensing model includes, for example, but is not limited to: Convolutional Neural Networks (CNN), Recurrent Neural Networks (RNN) and Long Short-Term Memory (LSTM).

[0081] The technical solution provided in the embodiment of the present application uses a free calcium content soft measurement model to perform soft measurement of the free calcium content of cement clinker output by the grate cooler at the current time based on multimodal data such as variable data sequence, kiln temperature information of the rotary kiln, and kiln head fire image sequence. Thus, a soft measurement method for the free calcium content of cement clinker based on multimodal data is provided, and multimodal data enriches the information source of the free calcium content soft measurement model. The above method can automatically, quickly and accurately perform soft measurement of the free calcium content of cement clinker, improve the effect of the free calcium content soft measurement, and can provide good guidance for the optimization of equipment operating parameters of the subsequent cement clinker burning system.

[0082] In this embodiment, there is no restriction on the network structure of the free calcium content soft measurement model. Figure 5, the free calcium content soft measurement model may include an image feature extraction network, a dynamic feature extraction network and a regression network connected in sequence. Wherein, the image feature extraction network may be any feature extraction network with an image feature extraction function. The dynamic feature extraction network may be any feature extraction network with a dynamic feature extraction function. Preferably, the dynamic feature extraction network includes an encoder with an encoding function and a decoder with a decoding function connected in sequence. The regression network may be any neural network with a regression analysis function. The regression network may include, for example, but is not limited to, a multilayer perceptron (MLP).

[0083] Based on the above, the variable data sequence, the kiln temperature information of the rotary kiln and the kiln head fire viewing image sequence are input into the free calcium content soft measurement model, and an optional implementation method for performing soft measurement of the free calcium content of the cement clinker output by the grate cooler at the current time through the free calcium content soft measurement model is: input the kiln head fire viewing image sequence into the image feature extraction network to obtain the image feature sequence corresponding to the kiln head fire viewing image sequence; input the variable data sequence and the image feature sequence into the dynamic feature extraction network to obtain the dynamic features; input the average value of the variable data of multiple production process variables, the dynamic features and the kiln temperature information of the rotary kiln into the regression network to obtain the soft measurement result of the free calcium content of the cement clinker output by the grate cooler at the current time.

[0084] See also Figure 5 The image feature extraction network extracts image features from the kiln head fire watching images in the kiln head fire watching image sequence, obtains the image features corresponding to each kiln head fire watching image in the kiln head fire watching image sequence, and then obtains the image feature sequence according to the image features corresponding to each kiln head fire watching image.

[0085] Next, the variable data sequence and the image feature sequence are concatenated, and the concatenated features are input into the encoder in the dynamic feature extraction network for encoding processing to obtain dynamic features.

[0086] Next, the variable data sequence includes the variable data of each production process variable in each time unit within the first time period, where the time unit is, for example, milliseconds, seconds, or minutes, etc. For any production process variable, the variable data of each time unit within the first time period is averaged (i.e., averaged) to obtain the average value of the variable data of the production process variable.

[0087] Finally, the kiln temperature information, dynamic characteristics and the average value of the variable data of multiple production process variables of the rotary kiln within the first time period from the second time are input into the regression network for prediction processing, and the soft measurement result of the free calcium content of the cement clinker output by the grate cooler at the current time is obtained and output.

[0088] The above content is an introduction to the soft measurement of free calcium content using the free calcium content soft measurement model. The following is an introduction to the training process of the free calcium content soft measurement model.

[0089] Figure 6 A flowchart of a model training method provided in an embodiment of the present application. Figure 6 , the method may include the following steps:

[0090] 601. Obtain a training data set, wherein the training data set includes a plurality of training data, and the training data includes: a variable data sequence corresponding to cement clinker output by a grate cooler in historical time, kiln temperature information of a rotary kiln, a kiln head fire viewing image sequence, and a free calcium content test result.

[0091] 602. Perform model training on the free calcium content soft measurement model to be trained using the training data set to obtain a trained free calcium content soft measurement model, wherein the free calcium content soft measurement model is used to perform soft measurement of the free calcium content of cement clinker output by the grate cooler at the current time.

[0092] In this embodiment, there is no restriction on the network structure of the free calcium content soft measurement model. The network structure of the free calcium content soft measurement model includes, for example, but is not limited to: Convolutional Neural Networks (CNN), Recurrent Neural Networks (RNN) and Long Short-Term Memory (LSTM).

[0093] In the model training stage, a large amount of training data is collected. Among them, each training data includes: the variable data sequence corresponding to the cement clinker output by the grate cooler at historical time, the kiln temperature information of the rotary kiln, the kiln head fire image sequence and the free calcium content test results. Among them, the historical time refers to the time before the current time. For the cement clinker output by the grate cooler at historical time, determine the time when the cement clinker enters the grate cooler and the time when the cement raw material for preparing the cement clinker enters the rotary kiln. In addition, obtain the variable data of multiple production process variables related to the grate cooler within the first time period from the time when the cement clinker enters the grate cooler, and obtain the variable data of multiple production process variables related to the specified equipment within the first time period from the time when the cement raw material for preparing the cement clinker enters the rotary kiln, and generate the variable data sequence corresponding to the cement clinker output by the grate cooler at historical time according to the variable data of the multiple production process variables obtained. In addition, the kiln temperature information of the rotary kiln within the first time period from the time when the raw cement material for preparing the cement clinker enters the rotary kiln is also obtained as the kiln temperature information of the rotary kiln corresponding to the cement clinker output by the grate cooler in the historical time; the kiln head fire image sequence within the first time period from the time when the raw cement material for preparing the cement clinker enters the rotary kiln is also obtained as the kiln head fire image sequence corresponding to the cement clinker output by the grate cooler in the historical time. In addition, for the cement clinker output by the grate cooler in the historical time, the cement clinker output by the grate cooler in the historical time can be tested by manual offline testing to obtain the free calcium content test result corresponding to the cement clinker output by the grate cooler in the historical time.

[0094] After preparing the training data set, multiple training data in the training data set are used to iteratively perform multiple model trainings. After each model training, the free calcium content soft measurement model after the model parameters are adjusted is used as the next free calcium content soft measurement model to be trained, and the next model training is performed using new training samples until the end training condition is met, and the free calcium content soft measurement model obtained by the last training is used as the final free calcium content soft measurement model. The end training condition may be that the number of model trainings reaches a specified number, or that the free calcium content soft measurement model converges, which is not specifically limited here.

[0095] In practical applications, each time the model is trained, the variable data sequence corresponding to the cement clinker output by the grate cooler in the historical time in the training data, the kiln temperature information of the rotary kiln, and the kiln head fire image sequence are input into the free calcium content soft measurement model to obtain the predicted free calcium content soft measurement results output by the free calcium content soft measurement model. The loss value of the training data is determined based on the free calcium content test results and the predicted free calcium content soft measurement results in the training data. The model parameters of the free calcium content soft measurement model are adjusted according to the loss value of the training data. It is worth noting that the smaller the loss value, the better the robustness of the model. Therefore, in the iterative model training process, the model parameters of the free calcium content soft measurement model are continuously adjusted with the goal of minimizing the loss value until the loss value is minimized or the model training times reach the specified times. For more information about model training, please refer to the relevant technology.

[0096] In this embodiment, any loss function can be used to determine the loss value of the training data. Any loss function includes, but is not limited to, mean square error loss function, L2 loss function (also known as Euclidean distance), L1 loss function (Manhattan distance), cross entropy loss function, softmax loss function, and Focal loss function, which is mainly used to solve the problem of imbalance between difficult and easy samples.

[0097] It is worth noting that when using the loss function to determine the loss value of the training data, the free calcium content test result in the training data is input into the loss function as the true value, and the predicted free calcium content soft measurement result is input into the loss function as the predicted value, and the loss function calculates the loss value of the training data based on the input true value and predicted value. For more information about the loss function to calculate the loss value, please refer to the related technology.

[0098] The model training method provided in the embodiment of the present application can train a free calcium content soft measurement model based on multimodal data such as variable data sequence, kiln temperature information of the rotary kiln, and kiln head fire image sequence to perform soft measurement of free calcium content on the cement clinker output by the grate cooler at the current time. Multimodal data enriches the information source of the free calcium content soft measurement model, so that the free calcium content soft measurement model can automatically, quickly and accurately perform soft measurement of free calcium content on cement clinker, thereby improving the effect of free calcium content soft measurement.

[0099] Further optionally, in order to improve the soft measurement effect, the free calcium content soft measurement model may include an image feature extraction network, a dynamic feature extraction network and a regression network connected in sequence. Based on this, a training data set is used to train a soft measurement model of free calcium content to be trained, and one implementation method of obtaining a trained soft measurement model of free calcium content is as follows: for the training data in the training data set, a kiln head fire viewing image sequence corresponding to the cement clinker output by the grate cooler in the historical time in the training data is input into an image feature extraction network to obtain an image feature sequence corresponding to the training data; a variable data sequence and an image feature sequence corresponding to the training data are spliced ​​to obtain a spliced ​​feature corresponding to the training data, and the spliced ​​feature is input into an encoder in a dynamic feature extraction network to obtain a dynamic feature corresponding to the training data; the average value of the variable data of multiple production process variables corresponding to the training data, the dynamic feature and the kiln temperature information of the rotary kiln are input into a regression network to obtain a predicted soft measurement result of free calcium content corresponding to the training data; the regression loss of the training data is determined according to the predicted soft measurement result of free calcium content of the training data and the test result of free calcium content; the model parameters of the image feature extraction network, the dynamic feature extraction network and the regression network are adjusted according to the regression loss of the training data to obtain a trained soft measurement model of free calcium content.

[0100] In this embodiment, for any production process variable in the variable data sequence of the training data, the variable data of each time unit of the production process variable within the first time period is averaged (i.e., averaged) to obtain the average value of the variable data of the production process variable. The average value of the variable data of each production process variable in the multiple production process variables corresponding to the training data is obtained, and then the average value of the variable data of the multiple production process variables corresponding to the training data is obtained.

[0101] In this embodiment, the regression loss refers to the loss value determined according to the predicted free calcium content soft measurement result and the free calcium content test result of the training data. When the regression loss of the training data is determined using the loss function, the free calcium content test result in the training data is input into the loss function as the true value, and the predicted free calcium content soft measurement result is input into the loss function as the predicted value, and the loss function calculates the loss value of the training data based on the input true value and predicted value.

[0102] It can be understood that in the process of model training, a large amount of training data is used to train the model, and the process of using each training data to train the model is similar.

[0103] Further optionally, in order to improve the model performance of the free calcium content soft sensor model, in the model training stage, both regression loss and reconstruction loss are concerned, wherein the reconstruction loss is a loss value determined according to the splicing features and reconstruction features of the training data. When the loss function is used to determine the reconstruction loss of the training data, the splicing features of the training data are input into the loss function as true values, and the reconstruction features are input into the loss function as predicted values, and the loss function calculates the loss value of the training data based on the input true values ​​and predicted values.

[0104] Based on the above, the model parameters of the image feature extraction network, the dynamic feature extraction network and the regression network are adjusted according to the regression loss of the training data to obtain a trained soft measurement model for free calcium content. An optional implementation method is: inputting the dynamic features corresponding to the training data into the decoder in the dynamic feature extraction network to obtain the reconstruction features corresponding to the training data; determining the reconstruction loss of the training data according to the splicing features and the reconstruction features of the training data; adjusting the model parameters of the image feature extraction network, the dynamic feature extraction network and the regression network according to the regression loss and the reconstruction loss of the training data to obtain a trained soft measurement model for free calcium content.

[0105] In practical applications, the regression loss and reconstruction loss of the training data can be weighted summed, averaged, or squared to obtain a total loss value, and the model parameters of the image feature extraction network, dynamic feature extraction network, and regression network can be adjusted according to the total loss value. Further, in order to prevent the model from overfitting, a regularization term can be added to the loss function.

[0106] Figure 7 This is a schematic diagram of the structure of a soft measurement device for free calcium content in cement clinker provided in an embodiment of the present application. Figure 7 , the device may include:

[0107] an acquisition module 71, for acquiring a variable data sequence if the grate cooler outputs cement clinker at the current time, wherein the variable data sequence includes variable data of a plurality of production process variables related to the grate cooler within a first time period starting from a first time and variable data of a plurality of production process variables related to a designated device within a first time period starting from a second time, the first time being the time when the cement clinker enters the grate cooler, the second time being the time when the cement raw material enters the rotary kiln, the designated device performs a preceding process of the cooling process performed by the grate cooler, and the designated device at least includes a rotary kiln;

[0108] The acquisition module 71 is further used to acquire kiln temperature information of the rotary kiln and a kiln head fire observation image sequence within a first time period starting from the second time, wherein the kiln head fire observation image sequence includes a plurality of kiln head fire observation images, and the kiln head fire observation image is obtained by collecting images of the internal combustion flame state of the rotary kiln;

[0109] The soft measurement module 72 is used to input the variable data sequence, the kiln temperature information of the rotary kiln and the kiln head fire viewing image sequence into the free calcium content soft measurement model, so as to perform soft measurement of the free calcium content of the cement clinker output by the grate cooler at the current time through the free calcium content soft measurement model.

[0110] Further optionally, the free calcium content soft-sensing model includes an image feature extraction network, a dynamic feature extraction network and a regression network connected in sequence;

[0111] Correspondingly, the soft measurement module 72 is specifically used for: inputting the kiln head fire viewing image sequence into the image feature extraction network to obtain the image feature sequence corresponding to the kiln head fire viewing image sequence; inputting the variable data sequence and the image feature sequence into the dynamic feature extraction network to obtain the dynamic features; inputting the average value of the variable data of multiple production process variables, the dynamic features and the kiln temperature information of the rotary kiln into the regression network to obtain the soft measurement result of the free calcium content of the cement clinker output by the grate cooler at the current time.

[0112] Further optionally, the above-mentioned device also includes: a determination module, used to determine the third time spent by the grate cooler in cooling the cement clinker based on the average value of the grate speed of the grate cooler in the second time before the current time; and determine the first time based on the current time and the third time.

[0113] Further optionally, the determination module is also used to determine the fifth time period consumed by the rotary kiln to prepare the cement raw material into cement clinker based on the average value of the rotation speed of the rotary kiln within the fourth time period before the first time; and determine the second time based on the first time and the fifth time period.

[0114] Further optionally, when the determination module determines the fifth time period consumed by the rotary kiln to prepare cement raw materials into cement clinker, it is specifically used to: determine the fifth time period consumed by the rotary kiln to prepare cement raw materials into cement clinker based on the structural parameters of the rotary kiln, the average value of the rotation speed of the rotary kiln in the fourth time period before the first time, and the average value of the cement raw material feeding amount.

[0115] Further optionally, when the acquisition module 71 acquires the kiln temperature information of the rotary kiln within the first time period starting from the second time, it is specifically used to: input the average value of the variable data of multiple production process variables related to the designated equipment within the first time period starting from the second time into the rotary kiln temperature mechanism model, and acquire the kiln temperature information of the rotary kiln output by the rotary kiln temperature mechanism model.

[0116] Further optionally, the above device further comprises: a collection module, which is used to periodically control a plurality of sensors arranged in the cement clinker production system to collect data, so as to obtain and save variable data of production process variables collected by the plurality of sensors at each sampling time;

[0117] Accordingly, when the acquisition module 71 acquires the variable data sequence, it is specifically used to:

[0118] From the variable data of the production process variables collected by multiple sensors at each sampling time, variable data of multiple production process variables related to the grate cooler within a first time period starting from a first time and variable data of multiple production process variables related to the specified equipment within a first time period starting from a second time are obtained.

[0119] Further optionally, the acquisition module is also used to periodically acquire images of the internal combustion flame state of the rotary kiln, so as to obtain and save the kiln head fire image acquired at each sampling time;

[0120] Correspondingly, when the acquisition module 71 acquires the sequence of kiln head fire viewing images within the first time period starting from the second time, it is specifically used to: acquire the sequence of kiln head fire viewing images within the first time period starting from the second time from the saved kiln head fire viewing images collected at each sampling time.

[0121] Figure 7 The device shown can perform Figure 2 The method shown in the embodiment shown, its implementation principle and technical effect are not described in detail. Figure 7 The specific manner in which each module and unit of the device shown performs operations has been described in detail in the embodiment of the method, and will not be elaborated here.

[0122] Figure 8 This is a schematic diagram of the structure of a model training device provided in an embodiment of the present application. Figure 8 , the device may include:

[0123] An acquisition module 81 is used to acquire a training data set, wherein the training data set includes a plurality of training data, and the training data includes: a variable data sequence corresponding to cement clinker output by a grate cooler in historical time, kiln temperature information of a rotary kiln, a kiln head fire viewing image sequence, and a free calcium content test result;

[0124] The model training module 82 is used to train the free calcium content soft measurement model to be trained using the training data set to obtain a trained free calcium content soft measurement model. The free calcium content soft measurement model is used to perform soft measurement of the free calcium content of the cement clinker output by the grate cooler at the current time.

[0125] Further optionally, the free calcium content soft measurement model includes an image feature extraction network, a dynamic feature extraction network and a regression network connected in sequence; the model training module 82 is specifically used for: for the training data in the training data set, inputting the kiln head fire viewing image sequence corresponding to the cement clinker output by the grate cooler in the historical time in the training data into the image feature extraction network to obtain the image feature sequence corresponding to the training data; splicing the variable data sequence and the image feature sequence corresponding to the training data to obtain the splicing features corresponding to the training data, and inputting the splicing features into the encoder in the dynamic feature extraction network to obtain the dynamic features corresponding to the training data;

[0126] The average values ​​of the variable data of multiple production process variables corresponding to the training data, the dynamic characteristics and the kiln temperature information of the rotary kiln are input into the regression network to obtain the predicted free calcium content soft measurement results corresponding to the training data; the regression loss of the training data is determined according to the predicted free calcium content soft measurement results of the training data and the free calcium content test results; the model parameters of the image feature extraction network, the dynamic feature extraction network and the regression network are adjusted according to the regression loss of the training data to obtain a trained free calcium content soft measurement model.

[0127] Further optionally, the model training module 82 adjusts the model parameters of the image feature extraction network, the dynamic feature extraction network and the regression network according to the regression loss of the training data to obtain a trained soft measurement model for free calcium content, and is specifically used to: input the dynamic features corresponding to the training data into the decoder in the dynamic feature extraction network to obtain the reconstruction features corresponding to the training data; determine the reconstruction loss of the training data according to the splicing features and reconstruction features of the training data; adjust the model parameters of the image feature extraction network, the dynamic feature extraction network and the regression network according to the regression loss and the reconstruction loss of the training data to obtain a trained soft measurement model for free calcium content.

[0128] Figure 8 The device shown can perform Figure 6 The method shown in the embodiment shown, its implementation principle and technical effect are not described in detail. Figure 8 The specific manner in which each module and unit of the device shown performs operations has been described in detail in the embodiment of the method, and will not be elaborated here.

[0129] It should be noted that the execution subject of each step of the method provided in the above embodiment can be the same device, or the method can be executed by different devices. For example, the execution subject of steps 201 to 203 can be device A; for another example, the execution subject of steps 201 and 202 can be device A, and the execution subject of step 203 can be device B; and so on.

[0130] In addition, in some of the processes described in the above embodiments and the accompanying drawings, multiple operations that appear in a specific order are included, but it should be clearly understood that these operations may not be executed in the order in which they appear in this article or executed in parallel, and the sequence numbers of the operations, such as 401, 402, etc., are only used to distinguish between different operations, and the sequence numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions of "first", "second", etc. in this article are used to distinguish different messages, devices, modules, etc., do not represent the order of precedence, and do not limit the "first" and "second" to be different types.

[0131] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0132] Fig. 9 This is a schematic diagram of the structure of an electronic device provided in this application. Fig. 9 As shown, the electronic device includes: a memory 91 and a processor 92;

[0133] The memory 91 is used to store computer programs and can be configured to store various other data to support operations on the computing platform. Examples of such data include instructions for any application or method operating on the computing platform, contact data, phone book data, messages, pictures, videos, etc.

[0134] The memory 91 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read only memory (EEPROM), erasable programmable read only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, disk or optical disk.

[0135] The processor 92 is coupled to the memory 91 and is used to execute the computer program in the memory 91 to perform the steps in the soft measurement method or model training method of the free calcium content of cement clinker.

[0136] Further, if Fig. 9 As shown, the electronic device also includes: a communication component 93, a display 94, a power component 95, an audio component 96 and other components. Fig. 9 Only some components are shown schematically, which does not mean that the electronic device only includes Fig. 9 In addition, Fig. 9 The components in the dashed box are optional components, not mandatory components, and the specific components depend on the product form of the electronic device. The electronic device of this embodiment can be implemented as a terminal device such as a desktop computer, a laptop computer, a smart phone, or an IOT (Internet of Things) device, or a server device such as a conventional server, a cloud server, or a server array. If the electronic device of this embodiment is implemented as a terminal device such as a desktop computer, a laptop computer, a smart phone, etc., it can include Fig. 9 If the electronic device of this embodiment is implemented as a server device such as a conventional server, a cloud server or a server array, it may not include Fig. 9 Components within the dashed box.

[0137] For the detailed implementation process of the processor executing each action, please refer to the relevant description in the aforementioned method embodiment or device embodiment, which will not be repeated here.

[0138] Accordingly, an embodiment of the present application further provides a computer-readable storage medium storing a computer program, which, when executed, can implement each step that can be executed by an electronic device in the above method embodiment.

[0139] Accordingly, an embodiment of the present application also provides a computer program product, including a computer program / instruction. When the computer program / instruction is executed by a processor, the processor is enabled to implement each step in the above method embodiment that can be executed by an electronic device.

[0140] The above-mentioned communication component is configured to facilitate wired or wireless communication between the device where the communication component is located and other devices. The device where the communication component is located can access a wireless network based on a communication standard, such as WiFi (Wireless Fidelity), 2G (2 Generation, 2 Generation), 3G (3 Generation, 3 Generation), 4G (4 Generation, 4 Generation) / LTE (long Term Evolution), 5G (5 Generation, 5 Generation) and other mobile communication networks, or a combination thereof. In an exemplary embodiment, the communication component receives a broadcast signal or broadcast-related information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communication component also includes a near field communication (Near Field Communication, NFC) module to facilitate short-range communication. For example, the NFC module can be based on Radio Frequency Identification (RFID) technology, Infrared Data Association (IrDA) technology, Ultra Wide Band (UWB) technology, Bluetooth (BT) technology and other technologies.

[0141] The above-mentioned display includes a screen, and the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive input signals from a user. The touch panel includes one or more touch sensors to sense touch, slide, and gestures on the touch panel. The touch sensor may not only sense the boundary of a touch or slide action, but also detect the duration and pressure associated with the touch or slide operation.

[0142] The power supply assembly provides power to various components of the device where the power supply assembly is located. The power supply assembly may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power to the device where the power supply assembly is located.

[0143] The above-mentioned audio component can be configured to output and / or input audio signals. For example, the audio component includes a microphone (MIC), and when the device where the audio component is located is in an operating mode, such as a call mode, a recording mode, and a speech recognition mode, the microphone is configured to receive an external audio signal. The received audio signal can be further stored in a memory or sent via a communication component. In some embodiments, the audio component also includes a speaker for outputting an audio signal.

[0144] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-readable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0145] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0146] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0147] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0148] In a typical configuration, a computing device includes one or more processors (Central Processing Unit, CPU), input / output interface, network interface and memory.

[0149] Memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. Memory is an example of a computer-readable medium.

[0150] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, Phase Change RAM (PRAM), Static Random-Access Memory (SRAM), Dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), Read Only Memory (ROM), Electrically-Erasable Programmable Read-Only Memory (EEPROM), flash memory or other memory technology, CD-ROM, Digital Versatile Disc (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. According to the definition in this article, computer-readable media does not include temporary computer-readable media (transitory media), such as modulated data signals and carrier waves.

[0151] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity 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, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0152] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included within the scope of the claims of the present application.

Claims

1. A soft measurement method for free calcium content of cement clinker, characterized in that: include: If the grate cooler outputs cement clinker at the current time, a variable data sequence is obtained, wherein the variable data sequence includes variable data of a plurality of production process variables related to the grate cooler within a first time period starting from a first time and variable data of a plurality of production process variables related to a designated device within a first time period starting from a second time, the first time is the time when the cement clinker enters the grate cooler, the second time is the time when the cement raw material enters the rotary kiln, the designated device performs a preceding process of the cooling process performed by the grate cooler, and the designated device at least includes a rotary kiln; Acquire kiln temperature information of the rotary kiln and a kiln head fire observation image sequence within a first time period starting from the second time, wherein the kiln head fire observation image sequence includes a plurality of kiln head fire observation images, and the kiln head fire observation images are obtained by collecting images of the internal combustion flame state of the rotary kiln; The variable data sequence, the kiln temperature information of the rotary kiln and the kiln head fire viewing image sequence are input into a free calcium content soft measurement model, so as to perform soft measurement of the free calcium content of the cement clinker output by the grate cooler at the current time through the free calcium content soft measurement model.

2. The method according to claim 1, characterized in that The free calcium content soft-sensing model comprises an image feature extraction network, a dynamic feature extraction network and a regression network connected in sequence; The variable data sequence, the kiln temperature information of the rotary kiln and the kiln head fire viewing image sequence are input into a free calcium content soft measurement model, so as to perform a soft measurement of the free calcium content of the cement clinker output by the grate cooler at the current time through the free calcium content soft measurement model, including: Inputting the kiln head fire watching image sequence into the image feature extraction network to obtain an image feature sequence corresponding to the kiln head fire watching image sequence; Inputting the variable data sequence and the image feature sequence into a dynamic feature extraction network to obtain dynamic features; The average value of the variable data of multiple production process variables, the dynamic characteristics and the kiln temperature information of the rotary kiln are input into the regression network to obtain the soft measurement result of the free calcium content of the cement clinker output by the grate cooler at the current time.

3. The method according to claim 1, characterized in that Before obtaining the variable data sequence, the method further includes: Determining a third time duration consumed by the grate cooler to cool the cement clinker according to an average value of the grate speed of the grate cooler within a second time duration before the current time; The first time is determined according to the current time and the third duration.

4. The method according to claim 1, characterized in that: Before obtaining the variable data sequence, the method further includes: Determining a fifth time duration consumed by the rotary kiln to prepare the cement raw material into the cement clinker according to an average value of the rotation speed of the rotary kiln within a fourth time duration before the first time; The second time is determined according to the first time and the fifth time duration.

5. The method according to claim 4, characterized in that Determining a fifth time duration consumed by the rotary kiln to prepare the cement raw material into the cement clinker according to an average value of the rotation speed of the rotary kiln within a fourth time duration before the first time, comprising: The fifth time period consumed by the rotary kiln to prepare the cement raw material into the cement clinker is determined based on the structural parameters of the rotary kiln, the average value of the rotation speed of the rotary kiln in the fourth time period before the first time, and the average value of the cement raw material feeding amount.

6. The method according to any one of claims 1 to 5, characterized in that: Acquiring kiln temperature information of the rotary kiln within a first time period starting from the second time, comprising: The average value of the variable data of multiple production process variables related to the designated equipment within the first time period starting from the second time is input into the rotary kiln temperature mechanism model to obtain the rotary kiln temperature information of the rotary kiln output by the rotary kiln temperature mechanism model.

7. The method according to any one of claims 1 to 5, characterized in that: Before obtaining the variable data sequence, the method further includes: Periodically controlling multiple sensors arranged in the cement clinker production system to collect data, so as to obtain and save variable data of production process variables collected by the multiple sensors at each sampling time; Acquiring the variable data sequence includes: From the variable data of the production process variables collected by multiple sensors at each sampling time, the variable data of multiple production process variables related to the grate cooler within a first time period starting from a first time and the variable data of multiple production process variables related to the designated equipment within a first time period starting from a second time are obtained.

8. The method according to any one of claims 1 to 5, characterized in that: Before obtaining the kiln head fire viewing image sequence within the first time period starting from the second time, the method further includes: Periodically collecting images of the internal combustion flame state of the rotary kiln to obtain and save kiln head fire images collected at each sampling time; Accordingly, obtaining a sequence of kiln head fire viewing images within a first time period starting from the second time includes: From the stored kiln head fire viewing images collected at various sampling times, a kiln head fire viewing image sequence within a first time period starting from the second time is obtained.

9. A model training method, characterized in that: include: Acquire a training data set, wherein the training data set includes a plurality of training data, and the training data includes: a variable data sequence corresponding to cement clinker output by a grate cooler in historical time, kiln temperature information of a rotary kiln, a kiln head fire viewing image sequence, and a free calcium content test result; The training data set is used to perform model training on a free calcium content soft measurement model to be trained, thereby obtaining a trained free calcium content soft measurement model, which is used to perform soft measurement of free calcium content on cement clinker output by a grate cooler at the current time.

10. The method according to claim 9, characterized in that The free calcium content soft-sensing model comprises an image feature extraction network, a dynamic feature extraction network and a regression network connected in sequence; The training data set is used to perform model training on the free calcium content soft measurement model to be trained, so as to obtain a trained free calcium content soft measurement model, including: For the training data in the training data set, a kiln head fire viewing image sequence corresponding to the cement clinker output by the grate cooler at a historical time in the training data is input into the image feature extraction network to obtain an image feature sequence corresponding to the training data; splicing the variable data sequence and the image feature sequence corresponding to the training data to obtain splicing features corresponding to the training data, and inputting the splicing features into the encoder in the dynamic feature extraction network to obtain dynamic features corresponding to the training data; Inputting the average value, dynamic characteristics and kiln temperature information of the variable data of multiple production process variables corresponding to the training data into the regression network to obtain the predicted free calcium content soft measurement result corresponding to the training data; Determine the regression loss of the training data according to the predicted free calcium content soft measurement result and the free calcium content test result of the training data; The model parameters of the image feature extraction network, the dynamic feature extraction network and the regression network are adjusted according to the regression loss of the training data to obtain the trained free calcium content soft measurement model.

11. The method according to claim 10, characterized in that The model parameters of the image feature extraction network, the dynamic feature extraction network and the regression network are adjusted according to the regression loss of the training data to obtain the trained free calcium content soft measurement model, including: Inputting the dynamic features corresponding to the training data into a decoder in the dynamic feature extraction network to obtain reconstructed features corresponding to the training data; Determine the reconstruction loss of the training data according to the splicing features and reconstruction features of the training data; The model parameters of the image feature extraction network, the dynamic feature extraction network and the regression network are adjusted according to the regression loss and the reconstruction loss of the training data to obtain the trained free calcium content soft measurement model.

12. An electronic device, characterized in that: include: Memory and processor; The memory is used to store computer programs; The processor is coupled to the memory and configured to execute the computer program to perform the steps of the method according to any one of claims 1 to 11.

13. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the processor is enabled to implement the steps in the method according to any one of claims 1 to 11.

Images