Steel descaling method and system

By generating candidates for steel surface detection and selecting the ratio of mechanical and chemical descaling processes based on material and surface grade scores, the problems of low scale descaling efficiency and environmental pollution in the existing technology are solved, and efficient and environmentally friendly steel descaling treatment is achieved .

CN118735916BActive Publication Date: 2024-11-26SHENGSHI ORIENTAL (LIAONING) TECH DEV CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411216335.7
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-09-02
Publication Date
2024-11-26
Estimated Expiration
2044-09-02

AI Technical Summary

Technical Problem

Existing steel descaling methods mostly use a single method, which leads to low efficiency or incompleteness. Incorrect treatment of waste acid during chemical descaling will pollute the environment.

Method used

By detecting the annealed steel surface, multiple candidate solutions are generated, combining material categories, surface grades and area to be treated, the optimal ratio of mechanical descaling and chemical descaling processes is selected to improve the descaling efficiency and effect, and Reduce acid liquid usage and contamination through electrocatalytic cleaning.

Benefits of technology

It has achieved the selection of the best descaling solution based on the steel material and surface grade, improving the descaling efficiency and effect, reducing the amount of acid liquid, and reducing the risk of environmental pollution.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118735916B_ABST
    Figure CN118735916B_ABST
Patent Text Reader

Abstract

This application relates to the field of steel production technology, and provides a method and system for descaling steel. The steel descaling method includes feeding annealed steel into a descaling machine to descale and obtain pre-treated steel. It involves acquiring surface image data of the annealed steel and determining target detection results. Multiple candidate schemes are determined based on the target detection results, and a first score is calculated based on the material category, steel surface grade, and area to be treated, along with their corresponding weighting coefficients. The estimated energy consumption of the candidate schemes is calculated based on the steel running speed and the amount of acid used. A second score is obtained by scoring the candidate schemes based on the estimated energy consumption. An optimal scheme is determined based on the first and second scores of the candidate schemes, and the descaling process of the optimal scheme is performed on the pre-treated steel. This application can select the optimal descaling scheme for different steel materials and surface grades to improve the descaling efficiency or effect.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the technical field of steel production, and in particular to a steel descaling method and system. Background Art

[0002] Steel is the most basic material for manufacturing various mechanical equipment, and is the material basis for developing the national economy and national defense construction. It has penetrated into all walks of life. With the continuous improvement of global industrial production levels, major companies have higher and higher requirements for the product quality of plates. During the hot rolling and cold rolling processes, steel will form a layer of oxide skin, i.e. "scale", on the surface due to contact with oxygen, water vapor and other media in the air. This layer of scale is mainly composed of iron oxide, and its formation not only affects the surface quality of the steel, but may also have an adverse effect on subsequent processing.

[0003] In order to reduce the formation of scale on the surface of steel, the industry often takes a variety of preventive measures, such as strengthening the rust removal of raw materials, controlling the rolling temperature and time, and using a protective atmosphere. However, although these measures can reduce the formation of scale to a certain extent, they cannot be completely avoided. Technicians mostly use mechanical descaling or chemical descaling methods to remove scales on steel. Mechanical descaling is to remove the scale layer on the surface of steel by physical means. Commonly used methods include shot peening descaling and high-pressure water descaling. Mechanical descaling has the advantages of good descaling effect and fast processing speed, but it may also cause certain damage to the surface of steel. Chemical descaling uses chemical reactions to remove the scale layer on the surface of steel, such as pickling descaling. Pickling can effectively remove oxides on the surface of steel and improve the surface quality of steel. However, improper treatment of waste acid generated during pickling will pollute the environment.

[0004] Due to the different materials and surface grades (surface oxide scale grades) of steel, different descaling methods are required. However, currently, the industry mostly uses a single method to remove scales produced on steel. Failure to select the best descaling method results in low descaling efficiency or incomplete descaling. Summary of the invention

[0005] The present application provides a steel descaling method and system, which selects the best descaling solution according to different steel materials and surface grades (surface oxide scale grades) to improve the descaling efficiency and / or descaling effect of the steel.

[0006] In a first aspect, the present application provides a method for descaling a steel material, comprising:

[0007] The annealed steel is sent to a descaling machine to descale and obtain pretreated steel;

[0008] Acquire the surface image data of the steel after annealing, and determine the target detection result based on the surface image data of the steel; the target detection result includes the material category result, the steel surface grade result and the area to be processed result;

[0009] Determine a plurality of candidate solutions based on the material category result, the steel surface grade result, and the area to be processed result respectively;

[0010] Calculate a first score according to the material category result in the candidate solution, the steel surface grade result, and the area to be processed result and their corresponding weight coefficients;

[0011] Estimate the steel running speed and acid usage in the candidate solution;

[0012] Calculating the estimated energy consumption value of the candidate solution based on the steel running speed and the acid consumption;

[0013] Scoring the candidate solution based on the estimated energy consumption value of the candidate solution to obtain a second score;

[0014] An optimal solution is determined based on the first score of the candidate solutions and the second score, and a descaling process in the optimal solution is performed on the pretreated steel material.

[0015] This application detects the surface of the steel after annealing and generates multiple candidate solutions based on the detection results. The candidate solutions are scored according to their steel material, surface grade and estimated energy consumption value, and the candidate solution with the highest score is selected. This can reasonably and efficiently perform descaling treatment based on the steel's own scale layer to improve energy efficiency.

[0016] In some possible implementations, determining multiple candidate solutions based on the material category result, the steel surface grade result, and the to-be-processed area result includes:

[0017] When the score of the target detection result is higher than a first threshold, the candidate solution includes a first process; wherein the first process includes a descaling operation of a shot blasting machine;

[0018] When the score of the target detection result is lower than the first threshold and higher than the second threshold, the candidate solution includes a first process and a second process; wherein the second process includes electrolytic tank pickling and / or sulfuric acid pickling operations;

[0019] When the score of the target detection result is lower than a second threshold, the candidate solutions include a first process, a second process and a third process; wherein the third process includes a mixed acid cleaning operation.

[0020] The present application generates multiple candidate solutions by combining the material category, steel surface grade and area to be treated in the target detection results. The candidate solutions are achieved by compound allocation of mechanical descaling and chemical descaling processes, and setting process conditions for different descaling processes according to the material category, steel surface grade and area to be treated. By selecting the best solution among the candidate solutions, the proportion of scale layers removed by mechanical descaling and chemical descaling can be efficiently allocated, energy efficiency can be improved, and acid emissions can be reduced.

[0021] In some possible implementations, determining the optimal solution based on the first score of the candidate solution and the second score includes:

[0022] Determining the process level of the descaling process in the candidate solution;

[0023] When the process levels in the candidate solutions are inconsistent, the first score is multiplied by the first weight coefficient to obtain a first product, and the second score is multiplied by the second weight coefficient to obtain a second product; the first product and the second product are summed to obtain a first sum value, and the candidate solution with the maximum value in the first sum value is determined as the optimal solution.

[0024] The present application determines the process level of the descaling process in the candidate schemes. When the candidate schemes contain multiple processes of different levels, the candidate scheme with the highest score is taken as the optimal scheme by performing weighted summation through weight allocation of different processes. The descaling process that is most suitable for the steel is selected, and the proportion of scale layers removed by mechanical descaling and chemical descaling is efficiently allocated according to the actual situation of the steel (material type, steel surface grade and area to be treated), thereby improving the descaling efficiency of the steel.

[0025] In some possible implementations, estimating the steel running speed in the candidate solution includes:

[0026] preparing a plurality of samples based on the pretreated steel material, and feeding the samples into a shot blasting machine;

[0027] By adjusting the running speed of the steel material, the removal rate of the sample is determined;

[0028] Output the steel running speed corresponding to the sample with the highest removal rate.

[0029] In some possible implementations, determining the amount of acid used in the candidate solution includes:

[0030] preparing a plurality of samples based on the pretreated steel material, and feeding the samples into a pickling tank;

[0031] By adjusting the amount of acid used in each pickling tank, the removal rate of the sample is determined;

[0032] When the removal rate is 100%, the amount of acid used is calculated based on the mass of the acid solution in the pickling tank.

[0033] This application sets the process of shot blasting descaling to obtain the running speed of the steel when the surface removal rate of the steel is a fixed threshold; in the process of chemical descaling, the acid consumption value of the residual scale layer per unit area is a fixed threshold to obtain the amount of acid used; in the actual process, the main parameters of pickling can also be adjusted according to the surface quality inspection of the steel at the pickling outlet to form a control closed loop. This application determines the estimated energy consumption values ​​corresponding to different candidate solutions by calculating the running speed of the steel and the amount of acid used, and scores each candidate solution according to the estimated energy consumption value, thereby controlling costs, reducing energy consumption, and reducing acid consumption.

[0034] In some possible implementations, the image data of the steel surface after annealing is collected by an image acquisition device; after acquiring the image data of the steel surface after annealing, the method includes:

[0035] Importing the steel surface image data into a target detection model; the target detection model includes one of mobilenet_v3_small, mobilenet_v3_large, squeezenet1_0, squeezenet1_1, resnet18, resnet50, and efficientnet_v2_s;

[0036] The target detection model outputs a target detection result based on the area to be processed and the material category of the annealed steel.

[0037] This application uses deep learning target detection technology and a trained target detection model to split the collected steel surface images into detection tasks and classification tasks according to the priority of the surface detection task objectives. The detection task is to confirm the surface grade of the steel and the area to be processed as the primary goal, and the classification task is to confirm the material category as the primary goal. Through task combination, the descaling position (defect position) and area are first located in the detection task, and then auxiliary classification is performed in the classification task, so as to effectively determine the specific situation of the steel to be dephosphorized.

[0038] In some possible implementations, the second process further includes:

[0039] The steel after dephosphorization by the shot blasting machine is sent to an electrocatalytic cleaning tank; the electrocatalytic cleaning tank includes active components and water, wherein the mass concentration of the active components is 15-20%, and the solution temperature in the cleaning tank is 70-80°C.

[0040] In some possible implementations, the active components include active component a, active component b, active component c and active component d; the active component a is one of potassium sulfate and lithium sulfate; the active component b is one of sodium tartrate and tetrasodium ethylenediaminetetraacetate; the active component c includes one of potassium diisopropyl dithiophosphate and potassium dibutyl dithiophosphate; the active component d includes one of diethanolamine and triethanolamine; wherein the mass ratio of active component a, active component b, active component c and active component d is 13~15:1~5:0.2~2:0.5~1.

[0041] In this application, a green and environmentally friendly electrocatalytic cleaning tank is selected, which can reduce the amount of acid used in sulfuric acid pickling and mixed acid pickling in chemical descaling, and the electrocatalytic liquid in the electrocatalytic cleaning tank can be recycled without generating pollution. The selection of non-hazardous and non-polluting active components can effectively remove iron oxide scale on the surface of steel and improve the descaling effect.

[0042] In some possible implementations, the sulfuric acid concentration in the electrolytic cell for pickling in the second process is 180-200 g / L, the metal ion concentration is 40-50 g / L, the acid solution temperature is 65-75°C, the current is 950-1050A, and the strip speed is 40-45 m / min.

[0043] In some possible implementations, the sulfuric acid concentration in the pickling tank for sulfuric acid pickling in the second process is 280-300 g / L, the metal ion concentration is 40-50 g / L, the acid solution temperature is 75-85°C, and the strip speed is 39-45 m / min.

[0044] In some possible implementations, the nitric acid concentration in the pickling tank of the mixed acid pickling in the third process is 95-105 g / L, the hydrofluoric acid concentration is 25-30 g / L, the metal ion concentration is 15-20 g / L, the acid solution temperature is 55-60°C, and the strip speed is 40-45 m / min.

[0045] In a second aspect, the present application also provides a steel descaling system, the system comprising:

[0046] An image acquisition device, used to obtain image data of the steel surface after annealing;

[0047] A mechanical descaling device is used to perform mechanical dephosphorization on the annealed steel; the mechanical descaling device comprises a descaling machine, a drying device and a shot blasting machine which are sequentially connected;

[0048] A chemical descaling device is used to chemically dephosphorize the steel after being treated by the mechanical dephosphorization device; the inlet of the chemical descaling device is connected to the shot blasting machine; the chemical descaling device includes a pre-cleaning device, an electrolytic cell, a sulfuric acid pickling device and a mixed acid cleaning device which are sequentially connected;

[0049] The controller is configured as:

[0050] The annealed steel is sent to a descaling machine to descale and obtain pretreated steel;

[0051] Acquire the surface image data of the steel after annealing, and determine the target detection result based on the surface image data of the steel; the target detection result includes the material category result, the steel surface grade result and the area to be processed result;

[0052] Determine a plurality of candidate solutions based on the material category result, the steel surface grade result, and the area to be processed result respectively;

[0053] Calculate a first score according to the material category result in the candidate solution, the steel surface grade result, and the area to be processed result and their corresponding weight coefficients;

[0054] Estimate the steel running speed and acid usage in the candidate solution;

[0055] Calculating the estimated energy consumption value of the candidate solution based on the steel running speed and the acid consumption;

[0056] Scoring the candidate solution based on the estimated energy consumption value of the candidate solution to obtain a second score;

[0057] An optimal solution is determined based on the first score of the candidate solutions and the second score, and a descaling process in the optimal solution is performed on the pretreated steel material.

[0058] The steel descaling system provided in the present application scores candidate solutions from two perspectives: the adaptability of the target detection results and the estimated energy consumption value of the solution. It takes into account the steel material, surface grade and area to be processed, and the energy consumption of the candidate solutions, and combines these two parts to select the candidate solution that is suitable for the steel material and has low energy consumption as the optimal solution, thereby improving the descaling effect and saving costs.

[0059] It can be seen from the above technical scheme that the present application provides a steel descaling method and system, which includes sending the annealed steel into a scale breaker for descaling to obtain pretreated steel; obtaining surface image data of the annealed steel, and determining the target detection result based on the steel surface image data; determining multiple candidate schemes based on the material category result, the steel surface grade result and the area to be processed result respectively; calculating a first score according to the material category result, the steel surface grade result and the area to be processed result in the candidate scheme and their corresponding weight coefficients; estimating the steel running speed and acid usage in the candidate scheme; calculating the estimated energy consumption value of the candidate scheme based on the steel running speed and acid usage; scoring the candidate scheme based on the estimated energy consumption value of the candidate scheme to obtain a second score; determining the optimal scheme based on the first score and the second score of the candidate scheme, and performing the descaling process in the optimal scheme on the pretreated steel. By testing the steel surface and generating multiple candidate solutions based on the test results, the candidate solutions are scored according to their steel material, surface grade and estimated energy consumption value, and the candidate solution with the highest score is selected. This can reasonably and efficiently perform descaling according to the steel scale layer itself to improve energy efficiency. The best descaling solution is selected according to different steel materials and surface grades to improve the descaling efficiency or descaling effect of steel. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] In order to more clearly illustrate the technical solution of the present application, the drawings required for use in the embodiments are briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0061] Figure 1 A flow chart of a steel descaling method provided in an embodiment of the present application;

[0062] Figure 2 This is a target detection reasoning result diagram of the original image of an embodiment of the present application. DETAILED DESCRIPTION

[0063] The following embodiments are described in detail, and examples thereof are shown in the accompanying drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following embodiments do not represent all embodiments consistent with the present application.

[0064] Steel is the most basic material for manufacturing various mechanical equipment, and is the material basis for developing the national economy and national defense construction. It has penetrated into all walks of life. With the continuous improvement of global industrial production levels, major companies have higher and higher requirements for the product quality of plates. During the hot rolling and cold rolling processes, steel will form a layer of oxide skin, i.e. "scale", on the surface due to contact with oxygen, water vapor and other media in the air. This layer of scale is mainly composed of iron oxide, and its formation not only affects the surface quality of the steel, but may also have an adverse effect on subsequent processing.

[0065] In order to reduce the formation of scale on the surface of steel, the industry often takes a variety of preventive measures, such as strengthening the rust removal of raw materials, controlling the rolling temperature and time, and using a protective atmosphere. However, although these measures can reduce the formation of scale to a certain extent, they cannot be completely avoided. Technicians mostly use mechanical descaling or chemical descaling methods to remove scale on steel. Mechanical descaling is to remove the scale layer on the surface of steel by physical means. Commonly used methods include shot peening descaling and high-pressure water descaling. Mechanical descaling has the advantages of good descaling effect and fast processing speed, but it may also cause certain damage to the surface of steel. Chemical descaling uses chemical reactions to remove the scale layer on the surface of steel, such as pickling descaling. Pickling can effectively remove oxides on the surface of steel and improve the surface quality of steel. However, improper treatment of waste acid generated during pickling will pollute the environment. During the mechanical descaling process, due to the high intensity of treatment, scratches, pits and other damages may be caused to the surface of the steel, affecting the subsequent processing and performance of the steel. In addition, mechanical descaling equipment is complex and has high maintenance costs. In the process of chemical descaling, the treatment of waste acid is a difficult problem. If not handled properly, the harmful substances in the waste acid may pollute the environment. At the same time, the pickling process may also cause excessive corrosion to the steel surface, affecting the strength and corrosion resistance of the steel.

[0066] Due to the different materials and surface grades (surface oxide scale grades) of steel, different descaling methods are required. However, currently, the industry mostly uses a single method to remove scales produced on steel. Failure to select the best descaling method results in low descaling efficiency or incomplete descaling.

[0067] Based on this, some embodiments of the present application provide a steel descaling method and system, which detects the steel surface and generates multiple descaling candidate solutions based on the detection results, scores the candidate solutions based on the material category results, steel surface grade results and the area to be processed results in the candidate solutions, and selects the candidate solution with the highest score, which can reasonably and efficiently allocate the scale ratio of mechanical descaling and chemical descaling according to the scale layer of the steel itself, thereby improving energy efficiency. The best descaling method is selected according to the different steel materials and surface grades to improve the descaling efficiency or descaling effect of the steel.

[0068] like Figure 1 As shown, some embodiments of the present application provide a steel descaling method comprising the following steps:

[0069] S1: The annealed steel is sent to a scale breaker to be scaled to obtain pre-treated steel.

[0070] After annealing, the steel is simply cleaned to remove dirt and loose oxide scale on the steel plate surface, and then enters the descaling machine. Under water spraying, part of the oxide layer on the steel is broken and dropped under bending stress and tensile stress. After the iron powder is taken away by the spraying water, solid-liquid separation is performed and the spraying water is recycled. In some embodiments, the elongation rate of the descaling machine is controlled at 1.8% to 2.0%.

[0071] S2: Acquire surface image data of the steel material after annealing, and determine the target detection result based on the surface image data of the steel material.

[0072] The steel surface image data in the present application is acquired by an image acquisition device in a steel descaling system, and the image acquisition device may be a visual sensor.

[0073] This application uses deep learning target detection technology and a trained target detection model to split the collected steel surface images into detection tasks and classification tasks according to the priority of the surface detection task objectives. Among them, the detection task has the primary goal of confirming the surface grade of the steel and the area to be processed, and the classification task has the primary goal of confirming the material category. Through task combination, the descaling position (defect position) and area are first located in the detection task, and then auxiliary classification is performed in the classification task, so as to effectively determine the specific situation of the steel to be dephosphorized.

[0074] In some embodiments of the present application, the target detection model needs to undergo relevant training and learning before use.

[0075] The training and learning process of the target detection model is as follows:

[0076] A steel surface data set is established according to the original steel image data; the data set includes a training set and a test set; wherein the original steel image data may be historical steel surface image data and may be obtained by downloading from the network.

[0077] The more defects the steel has, the lower the surface grade of the steel. Therefore, detection can be carried out based on the surface defects of the steel to determine the surface grade of the steel. The surface defects of steel are divided into four major categories: cracks, rolling, scratches, and pits. Each major defect category is divided into several minor defects. The minor defect categories are used as defect labels, including scars, inclusions, foreign body intrusion, heavy skin, transverse cracks, longitudinal cracks, scratches, pits, etc. The collected original image data are annotated according to the defect labels to establish a steel surface data set.

[0078] The steel surface data set is sent to the constructed target detection model for training, and a trained target detection model, namely the training model, is obtained.

[0079] The constructed target detection models include the YOLO (You Only Look Once, target detection algorithm) series represented by CNN (Convolutional Neural Network) and the DETR (DEtection Transformer, end-to-end target detection) series represented by Transformer.

[0080] This embodiment uses the yolov8 model as the target detection model, trains the steel surface data set, and saves the model weight file with the best test effect.

[0081] The test set is used as the data to be detected and sent to the trained training model for reasoning to obtain the target detection reasoning result.

[0082] like Figure 2 As shown, the original image data is sent to the trained yolov8 model for inference to locate the location and category of the defect. Figure 2 In (a), two scar defects were detected. Figure 2 In (b), one inclusion defect was detected.

[0083] In some embodiments of the present application, the target detection model selects at least one of mobilenet_v3_small, mobilenet_v3_large, squeezenet1_0, squeezenet1_1, resnet18, resnet50, and efficientnet_v2_s.

[0084] In some embodiments, the steel surface grade results are divided into level one, level two and level three, and the corresponding scores are 0.9, 0.8 and 0.7 points respectively. In some embodiments, the area to be treated results are divided into small, medium and large, and the corresponding scores are 0.9, 0.8 and 0.7 points respectively. In some embodiments, the score corresponding to alloy steel in the material classification results is 0.9, and the score corresponding to carbon steel is 0.8. The alloy steel can be: 20CrMnTi or 42CrMo. The carbon steel can be Q235B or Q345B.

[0085] S3: Determine multiple candidate solutions based on the material category results, steel surface grade results, and area to be processed results.

[0086] Based on the material category results, steel surface grade results and the area to be processed results, multiple candidate solutions are determined, including:

[0087] When the score of the target detection result is higher than the first threshold, the candidate solution includes a first process; wherein the first process includes a descaling operation of a shot blasting machine;

[0088] When the score of the target detection result is lower than the first threshold and higher than the second threshold, the candidate solution includes the first process and the second process; wherein the second process includes electrolytic tank pickling and / or sulfuric acid pickling operation;

[0089] When the score of the target detection result is lower than the second threshold, the candidate solutions include the first process, the second process and the third process; wherein the third process includes a mixed acid cleaning operation.

[0090] The present application generates multiple candidate solutions by combining the material category, steel surface grade and area to be treated in the target detection results. The candidate solutions are achieved by compound allocation of mechanical descaling and chemical descaling processes, and setting process conditions for different descaling processes according to the material category, steel surface grade and area to be treated. By selecting the best solution among the candidate solutions, the proportion of scale layers removed by mechanical descaling and chemical descaling can be efficiently allocated, energy efficiency can be improved, and acid emissions can be reduced.

[0091] S4: Calculate the first score according to the material category results, steel surface grade results, and area results to be processed in the candidate solutions and their corresponding weight coefficients. The specific calculation process of the first score is: multiply the material category results, steel surface grade results, and area results to be processed in the candidate solutions with their corresponding weight coefficients to obtain multiple product values, sum the multiple product values ​​and take the average value, which is the first score.

[0092] S5: Estimate the steel running speed and acid usage in the candidate solutions.

[0093] S6: Calculate the estimated energy consumption of the candidate solution based on the steel running speed and the amount of acid used.

[0094] S7: Score the candidate solutions based on the estimated energy consumption values ​​of the candidate solutions to obtain a second score.

[0095] S8: Determine the optimal solution based on the first score and the second score of the candidate solutions.

[0096] In some embodiments, determining the optimal solution based on the first score and the second score of the candidate solutions includes:

[0097] Determine the process level of the descaling process in the candidate solution; the first process has the highest level, followed by the second process, and finally the third process;

[0098] When the process levels in the candidate solutions are consistent, that is, the candidate solutions contain processes of the same level, since the second process and the third process cannot exist alone in the candidate solution, that is, the candidate solution only contains one descaling process (the first process), then the candidate solution is the optimal solution;

[0099] When the process levels in the candidate solutions are inconsistent, that is, the candidate solutions contain multiple processes of different levels, the first score is multiplied by the first weight coefficient to obtain a first product, and the second score is multiplied by the second weight coefficient to obtain a second product; the first product and the second product are summed to obtain a first sum value, and the candidate solution with the maximum value in the first sum value is determined as the optimal solution.

[0100] This application determines the process level of the descaling process in the candidate scheme. When the candidate scheme contains multiple processes of different levels, the weight allocation of different processes is used to perform weighted summation, and the candidate scheme with the highest score is taken as the optimal scheme. The descaling process that is most suitable for the steel is selected, and the proportion of scale layers removed by mechanical descaling and chemical descaling is efficiently allocated according to the actual situation of the steel (material category, steel surface grade and area to be treated), so as to improve the descaling efficiency of the steel. For example, when the finished product is NO.1 / 2E, generally only sulfuric acid pickling is used without sodium sulfate electrolysis. When the raw material oxide scale is thicker, sodium sulfate electrolysis can be added. However, when the finished product is 2B / 2D, in order to ensure the surface grade, only sodium sulfate electrolysis can be used, and sulfuric acid pickling cannot be used.

[0101] In some embodiments, estimating the steel material running speed in the candidate solution includes:

[0102] preparing a plurality of test specimens based on the pre-treated steel material, and feeding the test specimens into a shot blasting machine;

[0103] By adjusting the running speed of the steel, the removal rate of the sample is determined;

[0104] Output the steel running speed corresponding to the sample with the highest removal rate.

[0105] When the removal rate is less than the removal rate threshold, the current steel running speed is reduced; the shot blasting and descaling and sample preparation are repeated until the removal rate of the sample is greater than or equal to the removal rate threshold.

[0106] In some embodiments, determining the amount of acid used in the candidate solution comprises:

[0107] preparing a plurality of test specimens based on pre-treated steel and feeding the test specimens into a pickling tank;

[0108] By adjusting the amount of acid used in each pickling tank, the removal rate of the sample is determined;

[0109] When the removal rate is 100%, the amount of acid used is calculated based on the mass of the acid solution in the pickling tank.

[0110] In some embodiments, by adjusting the amount of acid used in each pickling tank, determining the removal rate of the sample includes:

[0111] The sample is immersed in a pickling tank containing acid solutions of different mass concentrations for pickling treatment; wherein the mass of the acid solution in each pickling tank is the same;

[0112] When the removal rate is 100%, determine the mass percentage of acid in each pickling tank, and calculate the acid consumption of steel according to the formula , , is the mass percentage of the original acid solution; is the mass percentage of the acid solution in the i-th pickling tank after the pickling treatment is completed; is the mass of acid in the pickling tank.

[0113] In some embodiments, the estimated energy consumption value is calculated based on the following function:

[0114] ;

[0115] Among them, A is the energy consumption parameter of shot blasting and descaling, B is the base value of pickling, is the pickling energy consumption parameter; is the strip running speed corresponding to the candidate solution; In some embodiments, the energy consumption parameter of shot blasting is the product of the energy consumption coefficient of shot blasting and the length of the steel, the energy consumption parameter of pickling is the product of the energy consumption coefficient of acid treatment and the area of ​​the steel, and the pickling base is the product of the length and width of the pickling sample.

[0116] This application sets the process of shot blasting descaling to obtain the steel running speed when the steel surface removal rate is a fixed threshold; in the process of chemical descaling, the acid consumption value of the residual scale layer per unit area is a fixed threshold to obtain the acid consumption; in the actual process, the main parameters of pickling can also be adjusted according to the surface quality inspection of the steel at the pickling outlet to form a control closed loop. This application determines the estimated energy consumption values ​​corresponding to different candidate solutions by calculating the steel running speed and the acid consumption, and scores each candidate solution according to the estimated energy consumption value, thereby controlling costs, improving energy consumption, and reducing acid consumption.

[0117] By scoring the candidate solutions from two perspectives, namely the adaptability of the target detection results and the estimated energy consumption of the solutions, the steel material, surface grade and area to be treated are taken into consideration, and the energy consumption of the candidate solutions is combined to select the candidate solution that is suitable for the steel material and has low energy consumption as the optimal solution, thereby improving the descaling effect and saving costs.

[0118] In some embodiments, the shot blasting machine has a projection rate of 1000-1050 kg / min and a rotation speed of 2050-2100 rpm.

[0119] In some embodiments, the sulfuric acid concentration in the electrolytic cell is 180-200 g / L, the metal ion concentration is 40-50 g / L, the acid solution temperature is 65-75°C, the current is 950-1050 A, and the strip speed is 40-45 m / min. A catalyst is added to the solution at a concentration of 20-25 g / L.

[0120] In some embodiments, the catalyst includes hydrogen peroxide or manganese dioxide.

[0121] In some embodiments, the sulfuric acid concentration in the sulfuric acid pickling tank is 280-300 g / L, the metal ion concentration is 40-50 g / L, the acid solution temperature is 75-85° C., and the strip speed is 39-45 m / min.

[0122] In some embodiments, the nitric acid concentration in the mixed acid pickling tank is 95-105 g / L, the hydrofluoric acid concentration is 25-30 g / L, the metal ion concentration is 15-20 g / L, the acid solution temperature is 55-60° C., and the strip speed is 40-45 m / min.

[0123] In some embodiments, after the shot blasting machine descaling, grinding cleaning is also included, and the residual shot blasting material on the surface of the steel is cleaned by a pre-cleaning device; wherein, during the grinding cleaning process, the motor speed is 800~1000rpm, and the motor current is 320~330A.

[0124] In some embodiments, the second process also includes electrocatalytic cleaning, and the device used for electrocatalytic cleaning is composed of a multi-layer vertical electrochemical cleaning tank, an electrocatalytic liquid, and an electrode sheet. The cleaning tank for electrocatalytic cleaning includes at least one active component and water, the total mass concentration of the active components is 15-20%, and the solution temperature in the cleaning tank is 70-80°C. The electrocatalytic cleaning operation is performed after shot blasting and before electrolytic tank pickling and / or sulfuric acid pickling.

[0125] In some embodiments, the electrocatalyst liquid is prepared from active component a, active component b, active component c, active component d and water; wherein the mass ratio of active component a, active component b, active component c and active component d is 13~15:1~5:0.2~2:0.5~1.

[0126] The active component a includes one of potassium sulfate and lithium sulfate; the active component b includes one of sodium tartrate and tetrasodium ethylenediaminetetraacetate; the active component c includes one of potassium diisopropyl dithiophosphate and potassium dibutyl dithiophosphate; and the active component d includes one of diethanolamine and triethanolamine.

[0127] In this application, a green and environmentally friendly electrocatalytic cleaning tank is selected, which can reduce the amount of acid used in sulfuric acid pickling and mixed acid pickling in chemical descaling, and the electrocatalytic liquid in the electrocatalytic cleaning tank can be recycled without generating pollution. The selection of non-hazardous and non-polluting active components can remove iron oxide on the surface of steel and improve the descaling effect.

[0128] By testing the steel surface and generating multiple descaling candidate solutions based on the test results, scoring the candidate solutions and selecting the candidate solution with the highest score, it is possible to reasonably and efficiently allocate the scale ratios of mechanical descaling and chemical descaling based on the scale layer of the steel itself, thereby improving energy efficiency.

[0129] In a second aspect, the present application also provides a steel descaling system, the system comprising:

[0130] An image acquisition device, used to obtain image data of the steel surface after annealing;

[0131] The mechanical descaling device is used to perform mechanical dephosphorization on the annealed steel; the mechanical descaling device includes a descaling machine, a drying device and a shot blasting machine which are connected in sequence;

[0132] A chemical descaling device is used to chemically dephosphorize the steel after being treated by the mechanical dephosphorization device; the inlet of the chemical descaling device is connected to the shot blasting machine; the chemical descaling device includes a pre-cleaning device, an electrolytic cell, a sulfuric acid pickling device and a mixed acid cleaning device which are sequentially connected;

[0133] The controller is configured as:

[0134] The annealed steel is sent to a descaling machine to descale and obtain pretreated steel;

[0135] Acquire the surface image data of the steel after annealing, and determine the target detection result based on the surface image data of the steel; the target detection result includes the material category result, the steel surface grade result and the area to be processed result;

[0136] Determine multiple candidate solutions based on the material category results, steel surface grade results, and area to be processed results;

[0137] Calculate the first score according to the material category results, steel surface grade results and area to be processed results in the candidate solutions and their corresponding weight coefficients;

[0138] Estimate the steel running speed and acid usage in the candidate solutions;

[0139] Calculate the estimated energy consumption of the candidate solutions based on the steel running speed and the amount of acid used;

[0140] Scoring the candidate solutions based on their estimated energy consumption values ​​to obtain a second score;

[0141] An optimal solution is determined based on the first score and the second score of the candidate solutions, and a descaling process in the optimal solution is performed on the pretreated steel material.

[0142] Each device in the chemical descaling device in the present application is equipped with a corresponding circulation tank, and the solution between each device and the circulation tank circulates to chemically descaling the steel.

[0143] Specifically, a circulation tank connected to a sulfuric acid pickling device is provided with a filtering device. The circulating liquid in the circulation tank is sent to a solid-liquid separation device, and the separated liquid is returned to the circulation tank. The filtered solid is sent to a sintering process for recycling. The circulating liquid between the sulfuric acid circulation tank and the sulfuric acid pickling tank is sulfuric acid, which is convenient for chemical descaling of steel.

[0144] The circulating tank connected to the mixed acid pickling device is equipped with a filtering device. The circulating liquid in the circulating tank is sent to the solid-liquid separation device, and the separated liquid is returned to the circulating tank. The filtered solid is sent to the sintering process for recycling. The circulating liquid between the mixed acid circulating tank and the mixed acid pickling tank is circulating acid, which is convenient for chemical descaling of steel.

[0145] In some embodiments, the steel descaling system further includes a waste acid regeneration device, which regenerates nitric acid and hydrofluoric acid in the waste acid based on the waste acid spray roasting regeneration technology, recovers metal ions in the waste mixed acid in the form of oxides, and reduces acid consumption. Specifically, the SCR (Selective Catalytic Reduction) denitrification technology is used to treat the mixed acid waste, and a rare earth low-temperature catalyst is used, which has a stable catalytic effect and is not a hazardous waste.

[0146] In some embodiments, the steel descaling system further comprises an electrocatalytic cleaning device, and the electrocatalytic cleaning device comprises a multi-layer vertical electrochemical cleaning tank, an electrocatalytic liquid and an electrode sheet. The multi-layer vertical electrocatalytic tank has a tank shape structure of a rectangular tank body, the bottom of which is inclined from front to back to the middle and is provided with one to five layers of a ship-shaped tank body space structure, a tank cover is provided on the tank body, and an air intake pipe and an exhaust atomizer are provided, and the electrode sheet is divided into a cathode and an anode, the cathode material is stainless steel, and the anode material is titanium plate.

[0147] The controller in the steel descaling system is also used to execute the steel descaling method in any of the above embodiments, which will not be described in detail here. The steel descaling system provided in the present application detects the collected steel surface image data through a built-in target detection model, and determines multiple candidate solutions based on the target detection results obtained by the detection; and scores the candidate solutions based on the steel material, surface grade and estimated energy consumption value of the candidate solutions, thereby determining the optimal solution, and descaling the steel according to the optimal solution to improve the descaling efficiency or descaling effect of the steel.

[0148] Example 1

[0149] The annealed steel is sent to the descaling machine in the steel descaling system. The elongation rate of the descaling machine is 1.9%. Under the water spray, part of the oxide layer on the steel is broken and falls off.

[0150] The image acquisition device in the steel descaling system obtains the surface image data of the steel after annealing, and the controller imports the surface image data of the steel into the target detection model, and the target detection model is mobilenet_v3_large combined with efficientnet_v2_s.

[0151] The target detection model outputs target detection results, which include material classification results (20CrMnTi), steel surface grade results (secondary) and area to be processed results (large). The scores of the three results are 0.9, 0.8 and 0.7 respectively. The total score of the target detection result is (0.9+0.8+0.7) / 3=0.8. 0.8 is less than the first threshold value of 0.85 and higher than the second threshold value of 0.75. Therefore, the candidate solution is the first process combined with the second process.

[0152] Based on the material classification results, steel surface grade results and the area to be processed results, multiple candidate solutions are determined:

[0153] Solution 1: Shot blasting descaling + electrolysis in electrolytic tank.

[0154] Option 2: Shot blasting descaling + electrocatalytic cleaning.

[0155] Option 3: Shot blasting and descaling + sulfuric acid pickling.

[0156] Solution 4: Shot blasting descaling + electrolysis in electrolytic tank + sulfuric acid pickling.

[0157] The weight coefficient of shot blasting and descaling for material classification results is set to 0.3, the weight coefficient for steel surface grade results is set to 0.2, and the weight coefficient for the area to be treated results is set to 0.5.

[0158] The weight coefficient of electrolysis in the electrolytic cell for the material classification result is set to 0.2, the weight coefficient for the steel surface grade result is set to 0.6, and the weight coefficient for the area to be processed result is set to 0.2.

[0159] The weight coefficient of electrocatalytic cleaning for material classification results is set to 0.4, the weight coefficient for steel surface grade results is set to 0.4, and the weight coefficient for the area to be treated results is set to 0.2.

[0160] The weight coefficient of sulfuric acid pickling for material classification results is set to 0.3, the weight coefficient for steel surface grade results is set to 0.3, and the weight coefficient for the area to be treated results is set to 0.4.

[0161] The first score of scheme 1 = [(0.9×0.3+0.8×0.2+0.7×0.5)+(0.9×0.2+0.8×0.6+0.7×0.2)] / 2=0.79;

[0162] The first score of scheme 2 = [(0.9×0.3+0.8×0.2+0.7×0.5)+(0.9×0.4+0.8×0.4+0.7×0.2)] / 2=0.8;

[0163] The first score of scheme 3 = [(0.9×0.3+0.8×0.2+0.7×0.5)+(0.9×0.3+0.8×0.3+0.7×0.4)] / 2=0.785;

[0164] The first score of Plan 4 = [(0.9×0.3+0.8×0.2+0.7×0.5)+(0.9×0.2+0.8×0.6+0.7×0.2)+(0.9×0.3+0.8×0.3+0.7×0.4)] / 3=0.79.

[0165] The estimated steel running speed is 70m / min.

[0166] The estimated energy consumption is calculated based on the following function:

[0167] ;

[0168] Among them, A is the energy consumption parameter of shot blasting and descaling, B is the base value of pickling, is the pickling energy consumption parameter; is the strip running speed corresponding to the candidate solution; In some embodiments, the energy consumption parameter of shot blasting is the product of the energy consumption coefficient of shot blasting and the length of the steel, the energy consumption parameter of pickling is the product of the energy consumption coefficient of acid treatment and the area of ​​the steel, and the pickling base is the product of the length and width of the pickling sample.

[0169] Since no acid was used in Scheme 2, the amount of acid used was 0. The amounts of acid used in Schemes 1, 3 and 4 were calculated based on the actual acid used.

[0170] The estimated energy consumption of Scheme 1 is 20.5kw, the estimated energy consumption of Scheme 2 is 14.25kw, the estimated energy consumption of Scheme 3 is 26.85kw, and the estimated energy consumption of Scheme 4 is 31.35kw. The second score of Scheme 1 is 0.81, the second score of Scheme 2 is 0.87, the second score of Scheme 3 is 0.74, and the second score of Scheme 4 is 0.65.

[0171] The first score of scheme 1 is 0.79 and the second score is 0.81;

[0172] The first score of scheme 2 is 0.8 and the second score is 0.87;

[0173] The first score of scheme 3 is 0.785 and the second score is 0.74;

[0174] The first score of scheme 4 is 0.79 and the second score is 0.65.

[0175] Since Schemes 1 to 4 are all combinations of the first process and the second process, the process levels in the candidate schemes are inconsistent, so the optimal scheme is determined based on the weighted sum of the first scores and the second scores of Schemes 1 to 4.

[0176] The first sum of scheme 1 = 0.79 × 0.6 + 0.81 × 0.4 = 0.798;

[0177] The first sum of scheme 2 = 0.8 × 0.6 + 0.87 × 0.4 = 0.828;

[0178] The first sum of scheme 3 = 0.785 × 0.6 + 0.74 × 0.4 = 0.767;

[0179] The first sum value of Scheme 4 = 0.79 × 0.6 + 0.65 × 0.4 = 0.734.

[0180] Solution 2 has the largest first sum, so it is the best solution.

[0181] Therefore, the steel descaling method in this embodiment is:

[0182] The steel is sent to a shot blasting machine for dephosphorization; the shot blasting machine has a projection rate of 1000 kg / min and a rotation speed of 2050 rpm;

[0183] The steel after dephosphorization by the shot blasting machine is sent to an electrocatalytic cleaning tank; the electrocatalytic cleaning tank includes active components and water, wherein the mass concentration of the active components is 15%, and the solution temperature in the cleaning tank is 70°C. The active components include 10 kg potassium sulfate, 1 kg sodium tartrate, 1.5 kg potassium diisopropyl dithiophosphate and 0.5 kg triethanolamine.

[0184] Example 2

[0185] The annealed steel is sent to the scale breaker in the steel descaling system. The elongation rate of the scale breaker is 2.0%. Under the water spray, part of the oxide layer just applied is broken and falls off.

[0186] The image acquisition device in the steel descaling system obtains the surface image data of the steel after annealing, and imports the surface image data of the steel into the target detection model, and the target detection model is mobilenet_v3_large combined with squeezenet1_1.

[0187] The target detection model outputs target detection results, which include material classification results (Q345B), steel surface grade results (level three) and area to be processed results (large). The scores of the three results are 0.8, 0.7 and 0.7 respectively. The total score of the target detection result is (0.8+0.7+0.7) / 3=0.73. 0.73 is less than the second threshold value 0.75, so the candidate solutions are the first process + the second process + the third process.

[0188] Based on the material classification results, steel surface grade results and the area to be processed results, multiple candidate solutions are determined:

[0189] Solution 1: Shot blasting descaling + electrolysis in electrolytic tank + mixed acid cleaning.

[0190] Option 2: shot blasting descaling + electrocatalytic cleaning + mixed acid pickling.

[0191] Option 3: shot blasting descaling + sulfuric acid pickling + mixed acid pickling.

[0192] Solution 4: Shot blasting descaling + electrolysis in electrolytic tank + sulfuric acid pickling + mixed acid pickling.

[0193] Solution 5: Shot blasting descaling + electrolysis in electrolytic tank + electrocatalytic cleaning + mixed acid pickling.

[0194] Solution 6: shot blasting descaling + electrocatalytic cleaning + sulfuric acid pickling + mixed acid pickling.

[0195] Option 7, shot blasting descaling + electrocatalytic cleaning + sulfuric acid pickling + electrolytic cell electrolysis + mixed acid pickling (the process is complex and consumes a lot of energy, so it is not considered).

[0196] The weight coefficient of shot blasting and descaling for material classification results is set to 0.3, the weight coefficient for steel surface grade results is set to 0.2, and the weight coefficient for the area to be treated results is set to 0.5.

[0197] The weight coefficient of electrolysis in the electrolytic cell for the material classification result is set to 0.2, the weight coefficient for the steel surface grade result is set to 0.6, and the weight coefficient for the area to be processed result is set to 0.2.

[0198] The weight coefficient of electrocatalytic cleaning for material classification results is set to 0.4, the weight coefficient for steel surface grade results is set to 0.4, and the weight coefficient for the area to be treated results is set to 0.2.

[0199] The weight coefficient of sulfuric acid pickling for material classification results is set to 0.3, the weight coefficient for steel surface grade results is set to 0.3, and the weight coefficient for the area to be treated results is set to 0.4.

[0200] The weight coefficient of mixed acid pickling for material classification results is set to 0.35, the weight coefficient for steel surface grade results is set to 0.35, and the weight coefficient for the area to be treated results is set to 0.3.

[0201] The first score of scheme 1 = [(0.8×0.3+0.7×0.2+0.7×0.5)+(0.8×0.2+0.7×0.6+0.7×0.2)+(0.8×0.35+0.7×0.35+0.7×3)] / 3=0.728;

[0202] The first score of scheme 2 = [(0.8×0.3+0.7×0.2+0.7×0.5)+(0.8×0.4+0.7×0.4+0.7×0.2)+(0.8×0.35+0.7×0.35+0.7×3)] / 3=0.735;

[0203] The first score of scheme 3 = [(0.8×0.3+0.7×0.2+0.7×0.5)+(0.8×0.3+0.7×0.3+0.7×0.4)+(0.8×0.35+0.7×0.35+0.7×3)] / 3=0.731;

[0204] The first score of scheme 4 = [(0.8×0.3+0.7×0.2+0.7×0.5)+(0.8×0.2+0.7×0.6+0.7×0.2)+(0.8×0.3+0.7×0.3+0.7×0.4)+(0.8×0.35+0.7×0.35+0.7×3)] / 4=0.728;

[0205] The first score of scheme 5 = [(0.8×0.3+0.7×0.2+0.7×0.5)+(0.8×0.2+0.7×0.6+0.7×0.2)+(0.8×0.4+0.7×0.4+0.7×0.2)+(0.8×0.35+0.7×0.35+0.7×3)] / 4=0.731;

[0206] The first score of Plan 6 = [(0.8×0.3+0.7×0.2+0.7×0.5)+(0.8×0.4+0.7×0.4+0.7×0.2)+(0.8×0.3+0.7×0.3+0.7×0.4)+(0.8×0.35+0.7×0.35+0.7×3)] / 4=0.733.

[0207] The estimated steel running speed is 70m / min.

[0208] The estimated energy consumption is calculated based on the following function:

[0209] ;

[0210] Among them, A is the energy consumption parameter of shot blasting and descaling, B is the base value of pickling, is the pickling energy consumption parameter; is the strip running speed corresponding to the candidate solution; In some embodiments, the energy consumption parameter of shot blasting is the product of the energy consumption coefficient of shot blasting and the length of the steel, the energy consumption parameter of pickling is the product of the energy consumption coefficient of acid treatment and the area of ​​the steel, and the pickling base is the product of the length and width of the pickling sample.

[0211] The estimated energy consumption of Scheme 1 is 32.13kw, the estimated energy consumption of Scheme 2 is 28.40kw, the estimated energy consumption of Scheme 3 is 35.62kw, the estimated energy consumption of Scheme 4 is 45.23kw, the estimated energy consumption of Scheme 5 is 39.18kw, and the estimated energy consumption of Scheme 6 is 42.57kw. The second score of Scheme 1 is 0.61, the second score of Scheme 2 is 0.67, the second score of Scheme 3 is 0.57, the second score of Scheme 4 is 0.47, the second score of Scheme 5 is 0.53, and the second score of Scheme 6 is 0.50.

[0212] Since Schemes 1 to 6 are all combinations of the first process, the second process and the third process, and the process levels in the candidate schemes are inconsistent, the optimal scheme is determined based on the weighted sum of the first scores and the second scores of Schemes 1 to 4.

[0213] The first sum of scheme 1 = 0.728 × 0.6 + 0.61 × 0.4 = 0.444;

[0214] The first sum of scheme 2 = 0.735 × 0.6 + 0.69 × 0.4 = 0.507;

[0215] The first sum of scheme 3 = 0.731 × 0.6 + 0.57 × 0.4 = 0.417;

[0216] The first sum of scheme 4 = 0.728 × 0.6 + 0.47 × 0.4 = 0.342;

[0217] The first sum of scheme 5 = 0.728 × 0.6 + 0.53 × 0.4 = 0.387;

[0218] The first sum value of Scheme 6 = 0.728 × 0.6 + 0.50 × 0.4 = 0.366.

[0219] Solution 2 has the largest first sum, so it is the best solution.

[0220] Therefore, the steel descaling method in this embodiment is:

[0221] The steel is sent to a shot blasting machine for dephosphorization; the shot blasting machine has a projection rate of 1050kg / min and a rotation speed of 2100rpm;

[0222] The steel after dephosphorization by the shot blasting machine is sent to an electrocatalytic cleaning tank; the electrocatalytic cleaning tank includes active components and water, wherein the mass concentration of the active components is 20%, and the solution temperature in the cleaning tank is 80° C. The active components include 20 kg of lithium sulfate, 1 kg of tetrasodium ethylenediaminetetraacetate, 1 kg of potassium diisopropyl dithiophosphate, and 0.6 kg of diethanolamine.

[0223] The pretreated steel after electrocatalytic cleaning is sent to a mixed acid pickling tank for descaling. The mixed acid pickling tank has a nitric acid concentration of 105 g / L, a hydrofluoric acid concentration of 30 g / L, a metal ion concentration of 20 g / L, an acid solution temperature of 60°C, and a strip speed of 40 m / min.

[0224] It can be seen from the above embodiments that the present application provides a steel descaling method and system, which includes sending the annealed steel into a scale breaker for descaling to obtain pretreated steel; obtaining surface image data of the annealed steel, and determining the target detection result based on the steel surface image data; determining multiple candidate schemes based on the material category result, the steel surface grade result and the area to be processed result respectively; calculating a first score according to the material category result, the steel surface grade result and the area to be processed result in the candidate scheme and their corresponding weight coefficients; estimating the steel running speed and acid usage in the candidate scheme; calculating the estimated energy consumption value of the candidate scheme based on the steel running speed and acid usage; scoring the candidate scheme based on the estimated energy consumption value of the candidate scheme to obtain a second score; determining the optimal scheme based on the first score and the second score of the candidate scheme, and performing the descaling process in the optimal scheme on the pretreated steel. By testing the steel surface and generating multiple candidate solutions based on the test results, the candidate solutions are scored according to their steel material, surface grade and estimated energy consumption value, and the candidate solution with the highest score is selected. This can reasonably and efficiently perform descaling according to the steel scale layer itself to improve energy efficiency. The best descaling method is selected according to different steel materials and surface grades to improve the descaling efficiency or descaling effect of steel.

[0225] Similar parts between the embodiments provided in this application can be referenced to each other. The specific implementation methods provided above are only a few examples under the overall concept of this application and do not constitute a limitation on the scope of protection of this application.

Claims

1. A method for descaling a steel material, characterized in that: include: The annealed steel is sent to a descaling machine to descale and obtain pretreated steel; Acquiring surface image data of the steel material after annealing, and determining a target detection result based on the surface image data of the steel material; The target detection results include material category results, steel surface grade results and to-be-processed area results; wherein, the steel surface grade is determined based on the steel surface defects to obtain the steel surface grade results. The more defects the steel has, the lower the steel surface grade is; the steel surface grade results are divided into level one, level two and level three; Determine a plurality of candidate solutions based on the material category result, the steel surface grade result, and the area to be processed result respectively; Calculate a first score according to the material category result, the steel surface grade result and the area to be processed result in the candidate solution and their corresponding weight coefficients; Estimate the steel running speed and acid usage in the candidate solution; Calculating the estimated energy consumption value of the candidate solution based on the steel running speed and the acid consumption; Scoring the candidate solution based on the estimated energy consumption value of the candidate solution to obtain a second score; An optimal solution is determined based on the first score of the candidate solutions and the second score, and a descaling process in the optimal solution is performed on the pretreated steel material.

2. The steel descaling method according to claim 1, characterized in that: Determining multiple candidate solutions based on the material category result, the steel surface grade result and the area to be processed result respectively includes: When the score of the target detection result is higher than a first threshold, the candidate solution includes a first process; wherein the first process includes a descaling operation of a shot blasting machine; When the score of the target detection result is lower than the first threshold and higher than the second threshold, the candidate solution includes a first process and a second process; wherein the second process includes electrolytic tank pickling and / or sulfuric acid pickling operations; When the score of the target detection result is lower than a second threshold, the candidate solutions include a first process, a second process and a third process; wherein the third process includes a mixed acid cleaning operation.

3. The steel descaling method according to claim 2, characterized in that: Determining the optimal solution based on the first score of the candidate solution and the second score includes: Determining the process level of the descaling process in the candidate solution; When the process levels in the candidate solutions are inconsistent, the first score is multiplied by the first weight coefficient to obtain a first product, and the second score is multiplied by the second weight coefficient to obtain a second product; the first product and the second product are summed to obtain a first sum value, and the candidate solution with the maximum value in the first sum value is determined as the optimal solution.

4. The steel descaling method according to claim 1, characterized in that: Estimation of the steel running speed and acid usage for the candidate solutions include: preparing a plurality of samples based on the pretreated steel material, and feeding the samples into a shot blasting machine; By adjusting the running speed of the steel material, the removal rate of the sample is determined; Output the steel running speed corresponding to the sample with the highest removal rate.

5. The steel descaling method according to claim 1, characterized in that: Estimation of the steel running speed and acid usage for the candidate solutions include: preparing a plurality of samples based on the pretreated steel material, and feeding the samples into a pickling tank; By adjusting the amount of acid used in each pickling tank, the removal rate of the sample is determined; When the removal rate is 100%, the amount of acid used is calculated based on the mass of the acid solution in the pickling tank.

6. The steel descaling method according to claim 1, characterized in that: The surface image data of the steel material after annealing is collected by an image collection device; after obtaining the surface image data of the steel material after annealing, the method comprises: Importing the steel surface image data into a target detection model; the target detection model includes one of mobilenet_v3_small, mobilenet_v3_large, squeezenet1_0, squeezenet1_1, resnet18, resnet50, and efficientnet_v2_s; The target detection model outputs a target detection result based on the area to be processed and the material category of the annealed steel.

7. The steel descaling method according to claim 2, characterized in that: The second process further comprises: The steel after dephosphorization by the shot blasting machine is sent to an electrocatalytic cleaning tank; the electrocatalytic cleaning tank includes active components and water, wherein the mass concentration of the active components is 15-20%, and the solution temperature in the cleaning tank is 70-80°C.

8. The steel descaling method according to claim 7, characterized in that: The active components include active component a, active component b, active component c and active component d; the active component a is one of potassium sulfate and lithium sulfate; the active component b is one of sodium tartrate and tetrasodium ethylenediaminetetraacetate; the active component c includes one of potassium diisopropyl dithiophosphate and potassium dibutyl dithiophosphate; the active component d includes one of diethanolamine and triethanolamine; wherein the mass ratio of active component a, active component b, active component c and active component d is 13~15:1~5:0.2~2:0.5~1.

9. The steel descaling method according to claim 2, characterized in that: In the second process, the concentration of sulfuric acid in the electrolytic tank for pickling is 180-200 g / L, the concentration of metal ions is 40-50 g / L, the temperature of the acid solution is 65-75°C, the current is 950-1050A, and the strip speed is 40-45 m / min.

10. The steel descaling method according to claim 2, characterized in that: In the second process, the sulfuric acid concentration in the pickling tank of sulfuric acid pickling is 280-300 g / L, the metal ion concentration is 40-50 g / L, the acid solution temperature is 75-85°C, and the strip speed is 39-45 m / min.

11. The steel descaling method according to claim 2, characterized in that: In the third process, the nitric acid concentration in the pickling tank of the mixed acid pickling is 95-105 g / L, the hydrofluoric acid concentration is 25-30 g / L, the metal ion concentration is 15-20 g / L, the acid solution temperature is 55-60°C, and the strip speed is 40-45 m / min.

12. A steel descaling system, characterized in that: The system comprises: An image acquisition device, used to obtain image data of the steel surface after annealing; A mechanical descaling device is used to perform mechanical dephosphorization on the annealed steel; the mechanical descaling device comprises a descaling machine, a drying device and a shot blasting machine which are sequentially connected; A chemical descaling device is used to chemically dephosphorize the steel after being treated by the mechanical dephosphorization device; the inlet of the chemical descaling device is connected to the shot blasting machine; the chemical descaling device includes a pre-cleaning device, an electrolytic cell, a sulfuric acid pickling device and a mixed acid cleaning device which are sequentially connected; The controller is configured as: The annealed steel is sent to a descaling machine to descale and obtain pretreated steel; Acquire the surface image data of the steel after annealing, and determine the target detection result based on the steel surface image data; the target detection result includes the material category result, the steel surface grade result and the to-be-processed area result; wherein, based on the surface defects of the steel, the surface grade of the steel is determined to obtain the surface grade result of the steel, the more defects the steel has, the lower the surface grade of the steel; the surface grade result of the steel is divided into level one, level two and level three; Determine a plurality of candidate solutions based on the material category result, the steel surface grade result, and the area to be processed result respectively; Calculate a first score according to the material category result, the steel surface grade result and the area to be processed result in the candidate solution and their corresponding weight coefficients; Estimate the steel running speed and acid usage in the candidate solution; Calculating the estimated energy consumption value of the candidate solution based on the steel running speed and the acid consumption; Scoring the candidate solution based on the estimated energy consumption value of the candidate solution to obtain a second score; An optimal solution is determined based on the first score of the candidate solutions and the second score, and a descaling process in the optimal solution is performed on the pretreated steel material.