A method for adjusting operation plan of a steel continuous heat treatment production line

Abstract

The present application relates to a kind of steel continuous heat treatment production line operation plan adjustment method, method includes: S1, the energy medium flow data of continuous heat treatment operation line is collected;S2, according to the grouping and alignment of energy medium flow data with the production material basic attribute and unit production state;S3, according to the grouping of S2, based on the heat value level of each energy medium, construct unit energy medium consumption theme table;S4, based on unit energy medium consumption theme table, the energy consumption prediction model corresponding to different groups is constructed, and energy consumption prediction model is updated regularly;S5, the prediction result of energy consumption prediction model obtained based on S4, give the optimal adjustment strategy of production line operation plan.Compared with prior art, the present application has the advantages of improving the accuracy of steel energy consumption prediction, meeting the long cycle control needs of energy consumption in actual production process and the like.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of steel production, in particular to a steel continuous heat treatment production line operation plan adjustment method. BACKGROUND

[0002] According to the actual cost estimation of a large steel plant, the energy consumption cost accounts for 25-30% of the total steel manufacturing cost; and for the continuous heat treatment production line, the energy consumption cost accounts for about 50%. Therefore, for steel enterprises, especially for the continuous heat treatment production line, effectively controlling energy consumption is the key to cost control. Considering the complex working conditions of actual steel production, there are many uncontrollable factors in the short term, but the overall energy consumption level in the long cycle (more than one week, which can be adjusted according to the actual control cycle) can be effectively controlled. Therefore, it is necessary to accurately predict the energy consumption level in the long cycle to provide a reference for production operation adjustment and product structure strategy adjustment, and support effective control of manufacturing cost.

[0003] There are some patents and related literature on steel energy consumption prediction methods at present, which mainly focus on the prediction method of steel energy consumption, and its main application is the tracking of dynamic prediction results and the use of energy scheduling, and it does not involve specific prediction model elements, which cannot meet the long-cycle control needs of energy consumption in the actual production process of steel enterprises. SUMMARY

[0004] The purpose of the present application is to improve the accuracy of steel energy consumption prediction and meet the long-cycle control needs of energy consumption in the actual production process, and provide a steel continuous heat treatment production line operation plan adjustment method. The present application constructs an energy consumption prediction model under different unit operating states, and iteratively updates the evaluation according to the specified cycle to ensure the accuracy of the prediction results, and adjusts the production line operation plan in time according to the control cycle to meet the long-cycle control needs.

[0005] The purpose of the present application can be realized by the following technical solutions:

[0006] A steel continuous heat treatment production line operation plan adjustment method, the method comprising:

[0007] S1, collecting energy medium flow data of the continuous heat treatment operation line;

[0008] S2, grouping and aligning the energy medium flow data according to the production material basic properties and unit production state to obtain energy medium flow data of different groups, wherein the unit state and material group of the same group data are the same, and the same material group represents the same basic properties and process properties of the material;

[0009] S3, constructing a unit energy medium consumption theme table based on the heat value level of each energy medium according to the grouping of S2;

[0010] S4, constructing energy consumption prediction models corresponding to different groups based on the unit energy medium consumption theme table, and performing timing update on the energy consumption prediction models;

[0011] S5, giving an optimal adjustment strategy of the production line operation plan based on the prediction result of the energy consumption prediction model obtained in S4.

[0012] Further, the specific steps of S3 are:

[0013] S31, calculating the medium consumption q of medium j of the i-th coil j，i ;

[0014] S32, calculating the total consumption of medium j in a single period, the comprehensive standard energy consumption in a single period, the medium consumption per ton of steel in a single period, and the comprehensive standard energy consumption per ton of steel in a single period corresponding to each group of data;

[0015] S33, setting a single period as month, quarter, half year and year respectively, executing S32 for each period, and rolling calculating the total consumption of medium j, the comprehensive standard energy consumption, the medium consumption per ton of steel and the comprehensive standard energy consumption per ton of steel to form a unit energy medium consumption theme table with month, quarter, half year and year as the period respectively.

[0016] Further, the medium consumption q of medium j of the i-th coil j，i is:

[0017] q j，i = L j，i -L j，i-1

[0018] wherein q j，i represents the medium j consumption of the i-th coil, L j，i and L j，i-1 represent the medium j metering data at the time when the i-th coil and the i-1-th coil are off-line respectively.

[0019] Further, the total consumption of medium j in a single period of a group of data is:

[0020]

[0021] wherein k represents the number of coils.

[0022] Further, the comprehensive standard energy consumption in a single period of a group of data is:

[0023]

[0024] wherein c j is the standard energy consumption conversion coefficient of medium j.

[0025] Further, the specific steps of constructing the energy consumption prediction model corresponding to different groups based on the unit energy medium consumption theme table are as follows:

[0026] The variables of the energy consumption prediction model are weight w, temperature T, process speed s, and time t, and the output of the mth prediction model is the consumption Q j，m ′ of medium j, and the comprehensive standard energy consumption level SQ m ′; the energy consumption prediction model is divided into a unit medium consumption prediction model and a unit standard energy consumption prediction model, the output of the mth unit medium consumption prediction model is the consumption Q j，m ′ of medium j, and the output of the mth unit standard energy consumption prediction model is the comprehensive standard energy consumption level SQ m ′;

[0027] Based on the mth group of data in the unit energy medium consumption theme table, a plurality of algorithms are used to generate the mth unit medium consumption initial prediction model and the mth unit standard energy consumption initial prediction model, the optimal medium consumption initial prediction model and the optimal standard energy consumption initial prediction model are selected from the initial prediction model, the optimal medium consumption initial prediction model is taken as the mth medium consumption prediction model corresponding to the mth group of data, and the optimal standard energy consumption initial prediction model is taken as the mth standard energy consumption prediction model corresponding to the mth group of data.

[0028] Further, the specific steps of periodically updating the energy consumption prediction model are as follows:

[0029] It is detected whether the data of all groups of the unit energy medium consumption theme table all exist corresponding medium consumption prediction models and standard energy consumption prediction models, if all exist, it is judged whether the evaluation period is reached,

[0030] if reached, the energy consumption prediction model corresponding to different groups is constructed again, if not reached, it is continuously judged whether the evaluation period is reached;

[0031] If the data of a certain group does not exist the corresponding medium consumption prediction model or the standard energy consumption prediction model, the energy consumption prediction model corresponding to the data of the group is constructed.

[0032] Further, the specific steps of the prediction results of the energy consumption prediction model obtained based on S4 are as follows:

[0033] The energy consumption prediction model is periodically updated, and then based on the energy consumption prediction model obtained based on S4, the consumption Q j，m ′ of medium j predicted by each model and the comprehensive standard energy consumption level SQ m ′ are obtained, the consumption Q j，m ′ of medium j predicted by each model is added to obtain the total consumption Q j, the comprehensive standard energy consumption levels SQ predicted by each model are added to obtain a total comprehensive standard energy consumption level SQ' m , the total consumption Q of the medium j is calculated based on the total comprehensive standard energy consumption level SQ' j , the total consumption Q of the medium j is calculated based on the total comprehensive standard energy consumption level SQ' j , the total consumption Q of the medium j is calculated based on the total comprehensive standard energy consumption level SQ', the total comprehensive standard energy consumption level SQ', the total consumption Q of the medium j and the total comprehensive standard energy consumption level SQ_std' are the prediction results.

[0034] Further, based on the prediction results of the energy consumption prediction model obtained in S4, the specific steps for optimizing the adjustment strategy of the production line operation plan are as follows:

[0035] It is judged whether any value in the prediction results is out of standard, if yes, the maintenance plan is adjusted or the product structure is adjusted, otherwise, the process returns to S1.

[0036] Further, the specific steps for adjusting the maintenance plan are adjusting the maintenance time t, and the specific steps for adjusting the product structure are adjusting the weight w of the single group of materials.

[0037] Compared with the prior art, the present application has the following beneficial effects:

[0038] (1) The present application constructs energy consumption prediction models corresponding to different groups based on the unit energy medium consumption theme table, the unit energy medium consumption theme table is constructed based on the calorific value level of each energy medium according to grouping, the basic properties and production state of the data in each group are different, which provides a basis support for accurate modeling of material energy consumption level prediction, and a more accurate prediction model can be provided.

[0039] (2) The present application is evaluated iteratively updated according to the specified period, and the period is adjusted according to the management requirements to ensure the accuracy of the prediction results.

[0040] (3) The present application supports recommending the optimization strategy of the operation plan according to the control period under the condition of quantitative medium consumption and / or comprehensive standard energy consumption, and can realize long-period control and optimization of the steel continuous heat treatment production line. BRIEF DESCRIPTION OF DRAWINGS

[0041] Figure 1 is a flowchart of the present application;

[0042] Figure 2 is a flowchart of S1-S3 of the present application;

[0043] Figure 3 is a flowchart of S4 of the present application;

[0044] Figure 4A flow chart for generating a prediction result of the present application;

[0045] Figure 5 A flow chart for giving an optimization adjustment strategy of a production line operation plan of the present application;

[0046] Figure 6 An architecture diagram for collecting data of the present application. DETAILED DESCRIPTION

[0047] The present application will be described in detail below in combination with the drawings and specific embodiments. The embodiments are implemented on the premise of the technical solutions of the present application, and detailed implementation modes and specific operation processes are given, but the protection scope of the present application is not limited to the following embodiments.

[0048] The present application provides a steel continuous heat treatment production line operation plan adjustment method aiming at the actual situation of a steel continuous heat treatment operation line, and a flow chart of the method is shown as Figure 1 The flow chart of S1-S3 of the present application is shown as Figure 2

[0049] Based on the collection of energy medium flow data of the continuous heat treatment operation line, the present application firstly groups materials according to the basic properties of production materials (group), and secondly realizes the alignment of medium consumption data and materials according to the time sequence of the unit production state. Based on the calorific value level of each energy medium, the consumption value of each energy medium is converted into standard energy consumption, and the sum of the standard energy consumption of each medium is the comprehensive standard energy consumption level. On this basis, the unit energy medium consumption theme table is constructed, which provides a basic support for the accurate modeling of material energy consumption level prediction.

[0050] Based on the medium consumption data theme table, a big data analysis modeling tool is used, a machine learning method is used, algorithms such as clustering, regression, fitting, decision tree, random forest, neural network, etc. are used to construct energy consumption prediction models under different unit operating states, and evaluation iteration update is carried out according to the specified period (which can be adjusted according to management requirements) to ensure the accuracy of the prediction result. Among them, the fault time is according to the average level of the last year, and the key process (key temperature, process speed) is according to the average level of the same material group distance in the last year as the calculation basis.

[0051] Based on the energy consumption prediction model, an energy consumption prediction application module and an operation optimization adjustment module are constructed, which support the optimization strategy (including product structure arrangement strategy, maintenance plan strategy, etc.) of the operation plan recommended according to the control period under the condition of medium consumption and / or comprehensive standard energy consumption quantification.

[0052] The present application comprises the following steps:

[0053] S1, collecting energy medium flow data of the continuous heat treatment operation line;​

[0054] S2. Group and align the energy medium flow data according to the basic properties of the production materials and the production status of the unit to obtain energy medium flow data of different groups. Among them, the unit status and material group are the same for the same group of data. The same material group means that the basic properties and process properties of the materials are the same.

[0055] S3. Based on the grouping in S2 and the calorific value of each energy medium, construct a table of energy medium consumption for the unit.

[0056] S4. Based on the unit energy medium consumption theme table, construct energy consumption prediction models corresponding to different groups, and update the energy consumption prediction models regularly.

[0057] S5. Based on the prediction results of the energy consumption prediction model obtained in S4, provide the optimal adjustment strategy for the production line operation plan.

[0058] This invention collects energy medium flow data from continuous heat treatment lines and groups and aligns the data. It primarily employs a three-layer information system architecture: "control layer - edge layer - platform layer," as shown below. Figure 6 As shown.

[0059] The control layer is responsible for collecting basic energy medium flow data, while the edge layer is responsible for standardizing and organizing energy medium consumption data, aligning material performance, unit energy medium consumption, and other data by timestamp.

[0060] The system adopts a three-tier information system architecture: "control layer - edge layer - platform layer". Figure 6 The system collects and organizes energy medium consumption data, and constructs an energy medium consumption topic table. The control layer is responsible for collecting basic energy medium flow data, the edge layer is responsible for organizing and aligning energy medium consumption data, and the platform layer supports the construction of the energy medium consumption topic table and the deployment of energy consumption prediction applications.

[0061] The control layer is responsible for collecting raw data on the energy medium flow (flow meter) of the unit and auxiliary facilities, and uploading it to the edge layer, which is to execute step S1.

[0062] In S2, the edge layer is responsible for the standardization of energy medium consumption data, aligning material performance, unit energy medium consumption, etc. data according to timestamp, and realizing the calculation of energy medium consumption: 1) realizing the association of unit energy medium consumption and material (steel coil), i.e. realizing the calculation of energy medium consumption according to material (steel coil), synchronizing with the basic attributes of material (specification group distance, steel grade, etc.), process attributes (process speed, representative temperature, etc. average), and 2) realizing the association of unit state and energy medium consumption data, i.e. dividing the unit state into running, maintenance and fault, defining the model group according to the combination condition of "unit state + material group"

see Table 1

[0063] Table 1 Configuration table of unit energy consumption prediction model group

[0064]

[0065] In S3, the platform layer builds energy medium consumption theme tables based on the data arrangement of the edge layer, including: 1) unit material energy medium consumption theme table; 2) unit material energy medium weekly consumption theme table.

[0066] The specific steps of S3 are:

[0067] S31: Medium consumption calculation:

[0068] q j，i = L j，i - L j，i-1

[0069] In the formula, q j，i represents the medium j consumption of the i th steel coil, L j，i and L j，i-1 represent the medium j metering data at the time when the i th steel coil and the i-1 th steel coil are offline. For a continuous heat treatment unit, the time when the i-1 th steel coil is offline is also the starting production time of the i th steel coil.

[0070] S32: Calculation of various comprehensive standard energy consumption and medium data:

[0071]

[0072]

[0073]

[0074]

[0075]

[0076] Qj，T _std = Q j，T / SUM(w)

[0077] SQ T _std = SQ T / SUM(w)

[0078] In the formula, SQ i is the comprehensive standard energy consumption level of the i-th coil, c j is the standard energy consumption conversion coefficient of each medium, J is the total number of media, Q j，m is the consumption of medium j under the m-th model group, k represents k steel coils under the m-th model group, SQ m is the comprehensive standard energy consumption level under the m-th model group, Q j，T is the total consumption of medium j in a single period, M represents M models in a single period, SQ T is the comprehensive standard energy consumption level in each period, Q j，T _std represents the medium consumption level per ton of steel, SQ T _std represents the comprehensive standard energy consumption level per ton of steel.

[0079] S33: Construct an energy medium consumption theme table, and respectively calculate the medium consumption (Q j，T ), the comprehensive standard energy consumption (SQ r ), the medium consumption per ton of steel (Q j，T _std), and the comprehensive standard energy consumption per ton of steel (SQ T _std) data according to the monthly / quarterly / half-yearly / yearly period.

[0080] According to actual production experience, the heat generated by the energy consumption of the steel continuous annealing furnace cannot instantaneously or in a short time (minute level) affect the change of the furnace condition or even the strip temperature. Considering the unstable factors of the energy consumption value in a short time (minute level), the present application takes the change of the energy medium consumption from the start to the end of a steel coil as the energy consumption value. The medium consumption q j of each material is calculated (Table 2), and the difference between the medium consumption at the moment when the material is offline (i.e. the moment when production is finished) and the moment when the previous material is offline is calculated. That is, q j，i = L j，i -L j，i-1 , in which q j，i represents the medium j consumption of the i-th steel coil, L j，i and L j，i-1 respectively represent the medium j metering data at the moment when the i-th steel coil and the i-1-th steel coil are offline. For a continuous heat treatment unit, the moment when the i-1-th steel coil is offline is also the moment when the i-th steel coil starts production. Note: the transition material is uniformly classified as a 1st material group.

[0081] On the basis of medium consumption calculation, according to the calorific value of each medium, the standard energy consumption conversion is carried out by comparing with the calorific value level of standard coal, and the standard energy consumption conversion coefficient of each medium is c j The comprehensive standard energy consumption level Sq i of the i-th roll is:

[0082]

[0083] On the basis of the medium consumption subject table, the model group defined according to the combination condition of "unit state + material group" is combined

see Table 1

[0084] Assuming that there are k steel rolls under each model group, the consumption Q j，m of medium j under the m-th model group is:

[0085]

[0086] According to the calorific value of each medium, the standard energy consumption conversion is carried out by comparing with the calorific value level of standard coal, and the comprehensive standard energy consumption level SQ m of the m-th model group is

[0087]

[0088] Assuming that there are M models in each period, the total consumption of medium j is:

[0089]

[0090] And the comprehensive standard energy consumption level in each period is:

[0091]

[0092] The ton steel consumption level is calculated as follows (Note: SUM(w) in the formula does not include transition material):

[0093] Q j，T _std=Q j，T / SUM(w)

[0094] SQ T _std=SQ T / SUM(w)

[0095] To support the production optimization adjustment strategy recommendation function, on the basis of Table 2, Table 4 (including Table 4_1 and Table 4_2) is constructed, and the medium consumption (Q j，T ), the comprehensive standard energy consumption (SQ T ) and the medium consumption per ton of steel (Q j，T _std), the comprehensive standard energy consumption per ton of steel (SQ T _std) data are calculated rolling per month / quarter / half year / year cycle respectively.

[0096] Table 2: Unit energy medium consumption subject table (example)

[0097]

[0098] Table 3: Unit energy medium weekly consumption subject table (example)

[0099]

[0100] Table 4_1: Unit energy medium consumption subject table rolling per cycle (month / quarter / half year / year) per week (example)

[0101]

[0102] Table 4_2: Unit energy medium consumption subject table rolling per cycle (month / quarter / half year / year) per week (example)

[0103]

[0104] Then the energy medium consumption prediction model is constructed.

[0105] Based on the historical data of the medium consumption subject table (Table 2) (generally, the data of the last year is taken as a sample. If the data cycle is less than one year, the whole sample is taken as the historical data.), grouped according to “unit state (running / maintenance / failure) + material group”, the machine learning method is used, and the clustering, regression, fitting, decision tree, random forest, neural network and other algorithms are used to construct the energy medium consumption prediction model, taking weight w, process attribute (representing temperature T, process speed s), time t and other variables as variables. For the same model group, the optimal algorithm is selected according to the principle of minimum difference between the predicted value and the actual consumption value, and the final energy consumption prediction model is determined, and the corresponding model is stored in the algorithm server of the platform layer. Considering the changes of objective conditions such as climate and equipment state, which will affect the actual energy consumption level, in order to ensure the availability of the prediction model, the effect of the model needs to be evaluated regularly (the cycle is preferably monthly), and the appropriate algorithm needs to be selected to update the original algorithm. The specific process is shown in Figure 3 .

[0106] Under the running state of the unit, the model variables of the production material include weight w, temperature T, and process speed s, and the model variables of the transition material include weight w and time t. Under the shutdown (maintenance or failure) state of the unit, the model variable is time t.

[0107] Unit medium consumption prediction model f m，j : Energy consumption prediction model of medium j under the mth model group. Taking weight w, process attribute (representing temperature T and process speed s), and time t as variables, the consumption level q of medium j is predicted. m，j .

[0108] Unit standard energy consumption prediction model F m : Unit standard energy consumption prediction model under the mth model group. Taking weight w, process attribute (representing temperature T and process speed s), and time t as variables, the standard energy consumption level Sq is predicted. m .

[0109] After the construction is completed, S5 is performed for prediction. Based on the unit energy consumption model, the product variety structure and maintenance schedule (Table 5) within a certain period (usually more than one week) are configured, taking material grouping weight (w) and unit state (running, maintenance, and failure) time (t) as variable combinations, to predict the consumption Q j and the comprehensive standard energy consumption SQ of medium j.

[0110] Among them, the process attribute data (representing temperature T and process speed s) of each model group are taken as the average values under the same running state condition, and the failure time t 故障 is calculated according to the daily average failure time of the previous year * the prediction period (days).

[0111]

[0112]

[0113] On this basis, the ton steel consumption (note: SUM(w) in the formula does not include the transition material) can be calculated as follows:

[0114] Ton steel medium consumption

[0115] Ton steel comprehensive standard energy consumption SQ_std' = SQ' / SUM(w)

[0116] Among them, f m，j (·) is the unit medium consumption prediction model: the energy consumption prediction model of medium j under the mth model group. Taking weight w, process attribute (representing temperature T and process speed s), and time t as variables, the consumption level q of medium j is predicted. m，j F m(·) is the standard energy consumption prediction model of the unit: the standard energy consumption prediction model of the unit in the mth model group. The weight w, process attribute (representing temperature T, process speed s), time t, etc. are used as variables to predict the standard energy consumption level Sq m .

[0117] The energy consumption prediction process in a certain period (usually more than a week) is shown in Figure 4 The variable configuration table includes the model group product structure configuration table and the maintenance plan configuration table 5:

[0118] Table 5 Energy consumption prediction product structure and maintenance plan configuration table (example)

[0119]

[0120] Note: Fault time estimation t_ fault = average daily fault time of last year * prediction duration

[0121] Then the production optimization adjustment strategy is recommended.

[0122] The platform layer is based on the energy medium consumption prediction model to build the production strategy optimization adjustment function, which supports the optimal strategy recommendation of the operation plan (including product structure strategy and maintenance plan strategy) under the quantitative condition of energy medium consumption (total energy consumption or single medium consumption). The specific process is shown in Figure 5 .

[0123] When the energy consumption prediction result does not reach the control target, the maintenance plan is adjusted first, and then the prediction result is evaluated; if the control target cannot be reached again, the product structure is adjusted. If the control target requirement cannot be met based on the maintenance plan and product structure adjustment, select N (usually more than 3, according to the control needs) specified prediction project consumption minimum week from Table 4 according to the prediction period, and use the corresponding product structure and fixed maintenance performance in the N weeks for product structure and fixed maintenance plan adjustment reference. According to the ratio of actual consumption level to control target level, the target proximity is calculated and sorted from large to small.

[0124] Target proximity a = actual consumption level / target control level * 100%.

[0125] The product structure and maintenance plan adjustment scheme recommendation output is shown in Table 6.

[0126]

[0127] The preferred embodiments of the present application have been described above in detail. It should be understood that modifications and variations to the present application can be affected by those skilled in the art without departing from the scope of the application. Accordingly, it is intended that all possible modifications and alterations be included within the scope of the present application as defined by the following claims.

Claims

1. A method for adjusting the operation plan of a continuous heat treatment production line for steel, characterized in that the method... The method comprises the following steps: S1, collecting energy medium flow data of a continuous heat treatment production line; S2, grouping and aligning the energy medium flow data according to the basic properties of the production materials and the production state of the unit, to obtain energy medium flow data of different groups; the unit state and the material group of the same group of data are the same, and the same material group indicates that the basic properties and process properties of the materials are the same; S3, constructing a unit energy medium consumption theme table based on the grouping of S2 and the heat value level of each energy medium; S4, constructing energy consumption prediction models corresponding to different groups based on the unit energy medium consumption theme table, and updating the energy consumption prediction models at regular intervals; S5, obtaining an optimal adjustment strategy for the production line operation plan based on the prediction results of the energy consumption prediction models obtained in S4.

2. The method of adjusting the operation plan of a steel continuous heat treatment production line according to claim 1, characterized in that, The specific steps of S3 are as follows: S31, calculate the medium consumption q of the medium j of the i-th steel coil j,i ; S32, calculating the total consumption of medium j in a single period, the comprehensive standard energy consumption in a single period, the medium consumption per ton of steel in a single period, and the comprehensive standard energy consumption per ton of steel in a single period corresponding to each group of data; S33, setting a single period as a month, a quarter, half a year, and a year respectively, executing S32 for each period, and rolling calculating the total consumption of medium j, the comprehensive standard energy consumption, the medium consumption per ton of steel, and the comprehensive standard energy consumption per ton of steel to form a unit energy medium consumption theme table with a month, a quarter, half a year, and a year as a period respectively.

3. The method of claim 2, wherein the method is characterized by: Medium consumption q of medium j of i-th steel coil j,i is: q j,i = L j,i - L j,i-1 wherein q j,i represents the medium j consumption amount of the i-th coil, L j,i and L j,i-1 respectively represent the medium j measurement data at the time of the i-th coil and the i-1-th coil going off line.

4. The steel continuous heat treatment production line operation plan adjustment method according to claim 3, characterized in that, The total consumption of medium j in a single period of a group of data is: wherein k represents the number of steel coils.

5. The method of adjusting the operation plan of a steel continuous heat treatment production line according to claim 4, characterized in that, The comprehensive standard energy consumption in a single period of a group of data is: where c j is the standard energy consumption conversion factor for medium j.

6. The method of claim 1, wherein the method further comprises: determining a target temperature of the steel product; and adjusting the target temperature of the steel product based on the target temperature and the temperature of the steel product. The specific steps of constructing energy consumption prediction models corresponding to different groups based on the unit energy medium consumption theme table are as follows: The variables of the energy consumption prediction model are weight w, temperature T, process speed s and time t, and the output of the mth prediction model is the consumption Q of medium j j,m ′ and the comprehensive standard energy consumption level SQ m ′; the energy consumption prediction model is divided into a unit medium consumption prediction model and a unit standard energy consumption prediction model, the output of the mth unit medium consumption prediction model is the consumption Q of medium j j,m ′, and the output of the mth unit standard energy consumption prediction model is the comprehensive standard energy consumption level SQ m ′; Based on the mth group of data in the unit energy medium consumption theme table, a plurality of algorithms are used to generate an initial prediction model of the mth unit medium consumption and an initial prediction model of the mth unit standard energy consumption, and the optimal initial prediction model of medium consumption and the optimal initial prediction model of standard energy consumption are selected from the initial prediction models, the optimal initial prediction model of medium consumption is taken as the mth medium consumption prediction model corresponding to the mth group of data, and the optimal initial prediction model of standard energy consumption is taken as the mth standard energy consumption prediction model corresponding to the mth group of data.

7. The method of adjusting the operation plan of a steel continuous heat treatment production line according to claim 6, characterized in that, The specific steps of updating the energy consumption prediction models at regular intervals are as follows: detecting whether all groups of data in the unit energy medium consumption theme table have corresponding medium consumption prediction models and standard energy consumption prediction models, if all groups of data have corresponding medium consumption prediction models and standard energy consumption prediction models, judging whether an evaluation period is reached, if the evaluation period is reached, constructing energy consumption prediction models corresponding to different groups again, if the evaluation period is not reached, continuing to judge whether the evaluation period is reached; if the data of a certain group do not have corresponding medium consumption prediction models or standard energy consumption prediction models, constructing energy consumption prediction models corresponding to the data of the group.

8. The method of claim 7, wherein the method further comprises: determining a target temperature of the steel product; and adjusting the target temperature of the steel product based on the target temperature and the temperature of the steel product. The specific steps of obtaining the prediction results of the energy consumption prediction models based on S4 are as follows: The energy consumption prediction model is updated periodically. Then, based on the energy consumption prediction model obtained in S4, the energy consumption Q of medium j predicted by each model is obtained. j,m ′ and comprehensive standard energy consumption level SQ m ′, the consumption Q of medium j predicted by each model j,m Adding them together, we get the total consumption Q of medium j. j The combined standard energy consumption level SQ predicted by each model is ''. m Adding them together, we get the overall standard energy consumption level SQ′, and the total consumption Q based on medium j. j The total energy consumption level SQ′ and the comprehensive standard energy consumption level are used to calculate the medium consumption per ton of steel, Q. j _std′ and the comprehensive standard energy consumption per ton of steel SQ_std′, the total consumption of medium j Q j ′、Comprehensive standard energy consumption level SQ′、Medium consumption per ton of steel Q j _std′ and the comprehensive standard energy consumption per ton of steel, SQ_std′, are the predicted results.

9. The method of claim 8, wherein the method further comprises: determining a target temperature of the steel product; and adjusting the target temperature of the steel product based on the target temperature and the temperature of the steel product. 9 The specific steps of obtaining the optimal adjustment strategy for the production line operation plan based on the prediction results of the energy consumption prediction models obtained in S4 are as follows: judging whether any value in the prediction results is out of standard, if yes, adjusting the maintenance plan or adjusting the variety structure, otherwise, returning to S1.

10. The method of claim 9, wherein the method further comprises: determining a target temperature of the steel product; and adjusting the target temperature of the steel product based on the target temperature and the temperature of the steel product. 10 The specific steps for adjusting the maintenance plan are adjusting the maintenance time t, and the specific steps for adjusting the variety structure are adjusting the weight w of the single group of materials. The specific steps for adjusting the maintenance plan are adjusting the maintenance time t, and the specific steps for adjusting the variety structure are adjusting the weight w of the single group of materials.

Images