A method for optimizing control parameters of a cement rotary kiln

Abstract

The application relates to a cement rotary kiln control parameter optimization method, which comprises the following steps: firstly, obtaining each target bit number data type related to the cement f-CaO content under the target cement rotary kiln working state; then, obtaining each to-be-referenced optimal control parameter vector based on a vector composed of each target bit number data type data value at each to-be-analyzed time point; then, obtaining the to-be-referenced optimal control parameter vector with the highest similarity to the current vector in each to-be-referenced optimal control parameter vector based on a current vector composed of the data values of each target bit number data type of the target cement rotary kiln under the current working state, and forming a target optimal control parameter vector; finally, adjusting each target bit number data type corresponding to the current working state of the target cement rotary kiln with the target optimal control parameter vector as the target; in this way, the working condition gradually approaches or reaches the historical best level, so that the output and quality of the clinker are effectively improved.

Description

TECHNICAL FIELD

[0001] The application relates to a cement rotary kiln control parameter optimization method and belongs to the technical field of cement rotary kiln control. BACKGROUND

[0002] The cement industry is an important industry for social infrastructure construction, and its production efficiency directly affects the stability of social infrastructure construction and the safety of human life and property, and the automation level shows the overall ability of a country. China is a large cement consumer, and with the development of production and construction, the cement industry in China maintains a rapid and stable development level and has developed into a large cement producer with an annual average output of about 2.5 billion tons, accounting for about half of the world's cement output. However, the energy consumption and environmental pollution problems in the cement production process in China have gradually become the focus of social attention. The production level of the cement industry has an important influence on the effective planning of social infrastructure construction and the normal operation of the economy. Therefore, under the premise of ensuring the stable operation of the cement rotary kiln, how to improve the production line working condition, improve the clinker quality, reduce the energy consumption and reduce the environmental pollution has become a problem to be solved in the cement industry.

[0003] The new dry process cement production mainly consists of three links of raw material preparation, clinker sintering and cement product preparation. The preheating and decomposition of the cement raw material are mainly carried out in the pre-decomposition furnace, and the preheated and decomposed raw material enters the rotary kiln rotating at a certain speed under the action of inertia, moves in the rotary kiln from the kiln tail to the kiln head, and completes the key link of cement clinker sintering. Therefore, the cement rotary kiln is a key equipment for cement production, and its running state directly determines the yield, quality and energy consumption of the cement clinker, and the optimization of the key parameters of the rotary kiln is the key to improving the quality, saving energy and reducing consumption of the cement enterprise. The rotary kiln is a complex thermal system, and extremely complex physical and chemical reactions occur in the rotary kiln. The process flow has the characteristics of multivariable, large time delay and nonlinearity, and it is difficult to establish a system model by using the traditional method, and it is difficult to realize the optimization adjustment of the working condition, so that the traditional control method is difficult to be applied in the actual production process. As a key equipment in the cement production process, the rotary kiln. At present, the parameter optimization setting of the rotary kiln in most cement plants is basically in the form of manual setting, which is greatly affected by human factors. Even experienced operators also have difficulty in maintaining the thermal stability of the kiln system for a long time SUMMARY

[0004] The technical problem to be solved by the application is to provide a cement rotary kiln control parameter optimization method, which adopts a new design strategy and can realize the optimization of the key parameters of the rotary kiln, and has important significance for improving the clinker production capacity of the production line, improving the clinker quality and promoting energy saving and emission reduction.

[0005] The application adopts the following technical scheme to solve the above technical problems: the application designs a cement rotary kiln control parameter optimization method for optimizing the control parameters of a target cement rotary kiln under a working state, comprising the following steps:

[0006] Step A. Based on the time series data values of all bit number data types corresponding to the preset historical time period at each collection time point under the normal working state of the target cement rotary kiln, the cleaning and screening method is used to obtain each analysis time point, and further obtain each target bit number data type related to the cement f-CaO content under the working state of the target cement rotary kiln, and then enter step B;

[0007] Step B. Based on the vector composed of the data values of each target bit number data type at each analysis time point, the clustering analysis method is used to cluster the vectors at each analysis time point, and obtain each reference optimal control parameter vector, and then enter step C;

[0008] Step C. Based on the current vector composed of the data values of each target bit number data type corresponding to the current working state of the target cement rotary kiln, the reference optimal control parameter vector with the highest similarity to the current vector in each reference optimal control parameter vector is obtained, and the target optimal control parameter vector is formed, and then enter step D;

[0009] Step D. Taking the target optimal control parameter vector as the target, the target cement rotary kiln is adjusted according to the current working state of each target bit number data type.

[0010] As a preferred technical scheme of the application: the step A comprises the following steps A1 to A3:

[0011] Step A1. Collect the time series data values of all bit number data types corresponding to the preset historical time period at each collection time point under the normal working state of the target cement rotary kiln, i.e. all bit number data types corresponding to each collection time point of the target cement rotary kiln, and then enter step A2;

[0012] Step A2. According to the normal working value range corresponding to each bit number data type, the data values of each bit number data type corresponding to each collection time point of the target cement rotary kiln are deleted, and further screening is performed to obtain each collection time point with a clinker yield greater than a preset clinker yield threshold in the remaining collection time points, and each analysis time point is formed, and then enter step A3;

[0013] Step A3. Based on all the data values of each bit number data type corresponding to each analysis time point and each preset analysis bit number data type in all bit number data types at each analysis time point of the target cement rotary kiln without considering the composition of the cement raw material, each analysis bit number data type related to the cement f-CaO content in the preset analysis bit number data type is obtained by correlation analysis, and each target bit number data type is formed.

[0014] As a preferred technical solution of the present application: in step A3, the following steps A3-1 to A3-3 are used to obtain each analysis bit number data type related to the cement f-CaO content in the preset analysis bit number data type by grey correlation analysis, and each target bit number data type is formed.

[0015] Step A3-1. For each analysis bit number data type, the following formula is used:

[0016]

[0017] The grey correlation coefficient ξ i (l) of each analysis bit number data type corresponding to each analysis time point is obtained, wherein i=1,...,I, I represents the number of preset analysis bit number data types, l=1,...,L, L represents the number of analysis time points, Δ i (l) = |t(l)-x i (l) |, t(l) represents the cement f-CaO content at the lth analysis time point, x i (l) represents the data value of the ith analysis bit number data type corresponding to the lth analysis time point, ρ represents a preset resolution coefficient, and 0<ρ<1, represents the minimum value of each Δ i (l), represents the maximum value of each Δ i (l), and ξ i (l) represents the grey correlation coefficient of the ith analysis bit number data type corresponding to the lth analysis time point, and then step A3-2 is entered.

[0018] Step A3-2. For each analysis bit number data type, the following formula is used:

[0019]

[0020] The grey correlation degree γ i of each analysis bit number data type related to the cement f-CaO content is obtained, wherein γ i represents the grey correlation degree of the ith analysis bit number data type related to the cement f-CaO content, and then step A3-3 is entered.

[0021] Step A3-3. Select each target bit number data type with the grey correlation degree greater than the preset grey correlation degree threshold, i.e. to constitute each target bit number data type.

[0022] As a preferred technical solution of the present application: the preset each target bit number data type includes secondary air temperature, flue gas NOx content, high-temperature fan speed, kiln tail temperature, flue gas oxygen content, kiln main motor current, decomposing furnace outlet temperature, tertiary air temperature, kiln head negative pressure, kiln tail pressure, decomposing furnace outlet temperature, preheater outlet temperature parameter, preheater outlet pressure parameter, preheater cone temperature parameter, preheater cone pressure parameter, high-temperature fan inlet temperature, high-temperature fan inlet pressure, grate cooler pressure and flue gas chamber pressure.

[0023] As a preferred technical solution of the present application: the step B includes the following steps B1 to B3.

[0024] Step B1. According to the number L of the time points to be analyzed, the following formula is used:

[0025]

[0026] The number K of clusters is obtained, and then step B2 is entered.

[0027] Step B2. Based on the vector composed of the data values of each target bit number data type at each time point to be analyzed, K-Means clustering processing is performed on the vector at each time point to be analyzed according to the number K of clusters, and K clusters are obtained, and then step B3 is entered.

[0028] Step B3. The vector closest to the cluster center in each cluster is obtained, and each reference optimal control parameter vector is constituted.

[0029] As a preferred technical solution of the present application: in the K-Means clustering processing in the step B, the distance between vectors adopts cosine distance.

[0030] As a preferred technical solution of the present application: in the step C, the similarity between each reference optimal control parameter vector and the current vector is obtained by the following operation.

[0031] Operation: for each reference optimal control parameter vector y, the current vector q composed of the data values of each target bit number data type of the target cement rotary kiln under the current working state is combined, and the following formula is used:

[0032]

[0033] obtaining similarity S(q, y) between the optimal control parameter vector y to be referred to and the current vector q, that is, obtaining similarity between each optimal control parameter vector to be referred to and the current vector, wherein n = 1,..., N, N represents the number of target bit number data types, q n represents the data value of the nth target bit number data type in the current vector q, y n represents the data value of the nth target bit number data type in the optimal control parameter vector y to be referred to, max(q n ,y n represents the maximum value of the nth target bit number data type between the optimal control parameter vector y to be referred to and the current vector q, min(q n ,y n represents the minimum value of the nth target bit number data type between the optimal control parameter vector y to be referred to and the current vector q, |q n -y n | represents the difference vector between the optimal control parameter vector y to be referred to and the current vector q.

[0034] The cement rotary kiln control parameter optimization method provided by the application has the following technical effects compared with the prior art by adopting the above technical scheme:

[0035] The cement rotary kiln control parameter optimization method designed by the application first obtains each analysis time point and each target bit number data type related to the cement f-CaO content under the target cement rotary kiln working state, then performs clustering processing on the vectors composed of the data values of each target bit number data type under each analysis time point, obtains each optimal control parameter vector to be referred to, then obtains the optimal control parameter vector to be referred to with the highest similarity to the current vector based on the current vector composed of the data values of each target bit number data type of the target cement rotary kiln under the current working state, and forms the target optimal control parameter vector, finally adjusts each target bit number data type of the target cement rotary kiln under the current working state with the target optimal control parameter vector as the target; in this way, the historical best level parameters under similar working conditions are compared, the working conditions gradually approach or reach the historical best level, and thus the yield and quality of the clinker are effectively improved. BRIEF DESCRIPTION OF DRAWINGS

[0036] Figure 1 is a flowchart of the cement rotary kiln control parameter optimization method designed by the application;

[0037] Figure 2 is a flowchart of the clustering analysis method involved in the application;

[0038] Figure 3This is a schematic diagram of a cement rotary kiln. DETAILED DESCRIPTION

[0039] The specific embodiments of the present invention will be further described in detail below with reference to the accompanying drawings.

[0040] The present invention designs a cement rotary kiln control parameter optimization method for Figure 3 The control parameters of the target cement rotary kiln working state are optimized. In actual application, according to Figure 1 As shown, specifically perform the following steps A to D.

[0041] Step A. Based on the time series data values ​​corresponding to each acquisition time point in the preset historical time period for all bit number data types under the normal working state of the target cement rotary kiln, each time point to be analyzed is obtained through cleaning and screening, and further each target bit number data type related to the cement f-CaO content under the working state of the target cement rotary kiln is obtained, and then step B is entered.

[0042] In actual application, the above step A specifically performs the following steps A1 to A3.

[0043] Step A1. Read the DCS configuration through OPC, collect the time series data values ​​corresponding to all the data types of the target cement rotary kiln at each collection time point in the preset historical time period under the normal working state of the target cement rotary kiln, that is, the data values ​​of all the data types of the target cement rotary kiln corresponding to each collection time point, and then proceed to step A2.

[0044] Step A2. Based on the preset normal working value range corresponding to each data type, for each data value of each data type at each acquisition time point corresponding to the target cement rotary kiln, delete the acquisition time points to which each abnormal data value belongs and the data values ​​of each data type at each acquisition time point, and further screen the remaining acquisition time points where the clinker production is greater than the preset clinker production threshold to form the time points to be analyzed, and then proceed to step A3.

[0045] Step A3. Without considering the composition of cement raw materials, based on the data values ​​of all bit number data types corresponding to each time point to be analyzed in the target cement rotary kiln, and the preset bit number data types to be analyzed in all bit number data types, the bit number data types to be analyzed related to the cement f-CaO content in the preset bit number data types to be analyzed are obtained through the correlation analysis method to constitute each target bit number data type.

[0046] The types of bit numbers to be analyzed are analyzed in the specific design implementation, including secondary air temperature, nitrogen oxide content in the flue gas chamber, high-temperature fan speed, kiln tail temperature, oxygen content in the flue gas chamber, kiln main motor current, decomposing furnace outlet temperature, tertiary air temperature, kiln head negative pressure, kiln tail pressure, decomposing furnace outlet temperature, preheater outlet temperature parameter, preheater outlet pressure parameter, preheater cone temperature parameter, preheater cone pressure parameter, high-temperature fan inlet temperature, high-temperature fan inlet pressure, grate cooler pressure and flue gas chamber pressure.

[0047] Step A3, in the design implementation, specifically as follows Step A3-1 to Step A3-3, through grey correlation degree analysis method, obtains each target bit number data type in each type of bit number data to be analyzed related to cement f-CaO content, and constitutes each target bit number data type;

[0048] Step A3-1, respectively for each type of bit number data to be analyzed, as follows:

[0049]

[0050] Obtain the grey correlation coefficient ξ i (l) of each type of bit number data to be analyzed corresponding to each analysis time point, wherein, i = 1,...,I, I represents the number of preset bit number data to be analyzed, l = 1,...,L, L represents the number of analysis time points, Δ i (l) = |t(l)-x i (l)|, t(l) represents the cement f-CaO content at the lth analysis time point, x i (l) represents the data value corresponding to the lth analysis time point of the ith type of bit number data to be analyzed, ρ represents the preset resolution coefficient, and 0<ρ<1, represents the minimum value of each Δ i (l), represents the maximum value of each Δ i (l), ξ i (l) represents the grey correlation coefficient of the ith type of bit number data to be analyzed corresponding to the lth analysis time point, and then enters Step A3-2.

[0051] Step A3-2, respectively for each type of bit number data to be analyzed, as follows:

[0052]

[0053] Obtain the grey correlation degree γ i of each type of bit number data to be analyzed related to cement f-CaO content, wherein, γ iThe grey correlation degree of the i-th bit number data type to be analyzed related to the cement f-CaO content is represented, and then step A3-3 is entered.

[0054] Step A3-3. Select each bit number data type to be analyzed with a grey correlation degree greater than a preset grey correlation degree threshold value, that is, to constitute each target bit number data type.

[0055] In actual implementation and application, 1000 groups of historical data of the process parameters and the cement f-CaO content are preliminarily selected, and the grey correlation degree values between the process parameters and the cement f-CaO content are calculated in the above manner. Since the correlation degrees of different process parameters to the cement f-CaO content have certain differences, in order to better establish the functional relationship between the process parameters and the cement f-CaO content, the grey correlation degree threshold value is set to 0.75. Then, each target bit number data type includes the feeding amount, the secondary air temperature, the flue gas nitrogen oxide content, the kiln tail temperature, the flue gas oxygen content, the kiln main motor current, the decomposing furnace outlet temperature, and the kiln head coal.

[0056] Step B. Based on the vector composed of the data values of each target bit number data type at each time point to be analyzed, the clustering analysis method as shown in Figure 2 is used to perform clustering processing on the vectors at each time point to be analyzed, to obtain each reference optimal control parameter vector, and then step C is entered.

[0057] Step B1. According to the number L of the time points to be analyzed, the following formula is used:

[0058]

[0059] The number K of clusters is obtained, and then step B2 is entered.

[0060] Step B2. Based on the vector composed of the data values of each target bit number data type at each time point to be analyzed, according to the number K of clusters, the cosine distance is used as the distance between vectors, and the K-Means clustering processing is performed on the vectors at each time point to be analyzed, to obtain K clusters, and then step B3 is entered.

[0061] Step B3. The vector closest to the cluster center in each cluster is obtained, to constitute each reference optimal control parameter vector.

[0062] Step C. Based on the current vector composed of the data values of each target bit number data type of the target cement rotary kiln corresponding to the current working state, the similarity between each reference optimal control parameter vector and the current vector is obtained by the following operation, and then the reference optimal control parameter vector with the highest similarity to the current vector in each reference optimal control parameter vector is obtained, to constitute the target optimal control parameter vector, and then step D is entered.

[0063] Operation: For each optimal control parameter vector y to be referenced, combined with the current vector q composed of the data values ​​of each target bit number data type of the target cement rotary kiln corresponding to the current working state, the following formula is used:

[0064]

[0065] Obtain the similarity S(q,y) between the optimal control parameter vector y to be referenced and the current vector q, that is, obtain the similarity between each optimal control parameter vector to be referenced and the current vector, where n=1,...,N, N represents the number of target bit data types, q n Indicates the data value of the nth target bit data type in the current vector q, y n Indicates the data value of the nth target bit data type in the optimal control parameter vector y to be referenced, max(q n ,y n ) represents the maximum value of the nth target bit data type between the optimal control parameter vector y to be referenced and the current vector q, min(q n ,y n ) represents the minimum value of the nth target bit data type between the optimal control parameter vector y to be referenced and the current vector q, |q n -y n | represents the difference vector between the reference optimal control parameter vector y and the current vector q.

[0066] Step D: Taking the target optimal control parameter vector as the target, adjusting the data types of each target bit number corresponding to the current working state of the target cement rotary kiln.

[0067] The control parameter optimization method of a cement rotary kiln designed by the present invention first obtains each time point to be analyzed and each target bit number data type related to the cement f-CaO content under the working state of the target cement rotary kiln; then, based on the vector composed of the data values ​​of each target bit number data type at each time point to be analyzed, clustering processing is performed on the vectors at each time point to be analyzed to obtain each optimal control parameter vector to be referenced; then, based on the current vector composed of the data values ​​of each target bit number data type corresponding to the current working state of the target cement rotary kiln, the optimal control parameter vector to be referenced with the highest similarity to the current vector is obtained among the optimal control parameter vectors to be referenced, and a target optimal control parameter vector is formed; finally, with the target optimal control parameter vector as the target, adjustments are made to each target bit number data type corresponding to the current working state of the target cement rotary kiln; thus, a comparison of historical best level parameters under similar working conditions is provided, so that the working conditions gradually approach or reach the historical best level, thereby effectively improving the output and quality of clinker.

[0068] The embodiments of the present application are described in detail above with reference to the accompanying drawings, but the present application is not limited to the above-described embodiments, and various changes can be made within the knowledge of those skilled in the art without departing from the spirit of the present application.

Claims

1. A method for optimizing control parameters of a cement rotary kiln, characterized in that, The application relates to a method for optimizing control parameters under a normal working state of a target cement rotary kiln. Step A. Based on time sequence data values of all bit number data types corresponding to preset historical time periods at each collection time point under the normal working state of the target cement rotary kiln, each analysis time point is obtained through cleaning and screening, and each target bit number data type related to the cement f-CaO content under the working state of the target cement rotary kiln is further obtained, and then step B is entered. The above step A includes the following steps A1 to A3: Step A1. Collecting time sequence data values of all bit number data types corresponding to preset historical time periods at each collection time point under the normal working state of the target cement rotary kiln, i.e. all bit number data type data values of the target cement rotary kiln corresponding to each collection time point, and then entering step A2; Step A2. According to the normal working value range corresponding to each bit number data type, the abnormal data values of each collection time point and the bit number data type data values at the collection time point are deleted from all bit number data type data values of the target cement rotary kiln corresponding to each collection time point, and further screening is performed to obtain each collection time point with a clinker yield greater than a preset clinker yield threshold in the remaining collection time points, thereby forming each analysis time point, and then entering step A3; Step A3. Without considering the cement raw material composition, based on all bit number data type data values of the target cement rotary kiln corresponding to each analysis time point and a preset analysis bit number data type in all bit number data types, each analysis bit number data type related to the cement f-CaO content in the preset analysis bit number data type is obtained through correlation analysis, thereby forming each target bit number data type; Step B. Based on the vector composed of the target bit number data type data values at each analysis time point, the vectors at each analysis time point are clustered through cluster analysis, and each reference optimal control parameter vector is obtained, and then step C is entered; Step C. Based on the current vector composed of the data values of the target bit number data types corresponding to the current working state of the target cement rotary kiln, the reference optimal control parameter vector with the highest similarity to the current vector in each reference optimal control parameter vector is obtained, thereby forming a target optimal control parameter vector, and then step D is entered; In the above step C, the similarity between each reference optimal control parameter vector and the current vector is obtained by the following operation: for each reference optimal control parameter vector y, the current vector q composed of the data values of the target bit number data types corresponding to the current working state of the target cement rotary kiln is combined, and the following formula is used: obtaining a similarity S(q, y) between the to-be-referenced optimal control parameter vector y and the current vector q, that is, obtaining a similarity between each to-be-referenced optimal control parameter vector and the current vector, where n = 1,..., N, N represents a quantity of target bit number data types, q n represents a data value of an nth target bit number data type in the current vector q, y n represents a data value of an nth target bit number data type in the to-be-referenced optimal control parameter vector y, max(q n ,y n ) represents a maximum value of the nth target bit number data type between the to-be-referenced optimal control parameter vector y and the current vector q, min(q n ,y n ) represents a minimum value of the nth target bit number data type between the to-be-referenced optimal control parameter vector y and the current vector q, |q n -y n | represents a difference value vector between the to-be-referenced optimal control parameter vector y and the current vector q; Step D. The target optimal control parameter vector is taken as the target, and the target bit number data types corresponding to the current working state of the target cement rotary kiln are adjusted.

2. The method for optimizing the control parameters of a cement rotary kiln according to claim 1, characterized in that: In the step A3, the preset each to be analyzed bit number data type related to the cement f-CaO content is obtained by the grey correlation degree analysis method through the following steps A3-1 to A3-3, and each target bit number data type is constituted; Step A3-1. For each to be analyzed bit number data type, the following formula is used: obtaining the grey correlation coefficients ξ corresponding to each to-be-analyzed bit number data type at each to-be-analyzed time point i (l), wherein i = 1,..., I, I represents a preset number of to-be-analyzed bit number data types, l = 1,..., L, L represents a number of to-be-analyzed time points, Δ i (l) = |t(l) - x i (l)|, t(l) represents a cement f-CaO content at the lth to-be-analyzed time point, x i (l) represents a data value of the ith to-be-analyzed bit number data type corresponding to the lth to-be-analyzed time point, and ρ represents a preset resolution coefficient, and 0 < ρ < 1, representing a minimum value in each Δ i (l), representing a maximum value in each Δ i (l), and ξ i (l) represents a grey correlation coefficient of the ith to-be-analyzed bit number data type corresponding to the lth to-be-analyzed time point, and then step A3-2 is entered. Step A3-2. For each to be analyzed bit number data type, the following formula is used: obtain the grey correlation degree γ of each to-be-analyzed bit number data type with respect to the cement f-CaO content i wherein γ i represents the grey correlation degree of the i-th to-be-analyzed bit number data type with respect to the cement f-CaO content, and then entering step A3-3; Step A3-3. Select each to be analyzed bit number data type with the grey correlation degree greater than the preset grey correlation degree threshold, that is, constitute each target bit number data type.

3. The method for optimizing control parameters of a cement rotary kiln according to claim 1 or 2, characterized in that: The preset each to be analyzed bit number data type includes the secondary air temperature, the flue gas NOx content, the high-temperature fan speed, the kiln tail temperature, the flue gas oxygen content, the kiln main machine current, the decomposing furnace outlet temperature, the tertiary air temperature, the kiln head negative pressure, the kiln tail pressure, the decomposing furnace outlet temperature, the preheater outlet temperature parameter, the preheater outlet pressure parameter, the preheater cone temperature parameter, the preheater cone pressure parameter, the high-temperature fan inlet temperature, the high-temperature fan inlet pressure, the grate cooler pressure and the flue gas chamber pressure.

4. The method for optimizing control parameters of a cement rotary kiln according to claim 1, characterized in that: The step B includes the following steps B1 to B3; Step B1. According to the number L of to be analyzed time points, the following formula is used: The number K of clusters is obtained, and then the step B2 is entered; Step B2. Based on the vector composed of each target bit number data type data value at each to be analyzed time point, the K-Means clustering processing is performed on the vector at each to be analyzed time point according to the number K of clusters, and K clusters are obtained, and then the step B3 is entered; Step B3. The vector closest to the cluster center in each cluster is obtained, and each to be referred optimal control parameter vector is constituted.

5. The method for optimizing control parameters of a cement rotary kiln according to claim 4, characterized in that: In the K-Means clustering processing in the step B, the distance between vectors adopts the cosine distance.

Images