A self-learning-based mill steel biting impact speed drop compensation method

Abstract

The present application belongs to the field of automation control and rolling technology, and particularly relates to a rolling mill steel biting impact speed drop pre-compensation method based on self-learning. According to the impact speed drop historical data corresponding to the current preparation rolling steel grade, the impact speed drop compensation value is added to the rolling mill speed given value before the steel biting. The total time needing compensation is set, and the total time is segmented. In each time segment, the difference between the rolling mill speed given value and the actual value is calculated. The self-learning system is used, and the time delay of the actuator is considered, so that the difference value is compensated in the rolling mill speed given value corresponding to the next strip steel time. After the compensation is completed, the impact speed drop compensation curve of the current steel grade is recorded, and is called when the same specification steel grade is rolled later. The present application can effectively improve the rolling mill steel biting impact speed drop problem, and improve the production stability and head size precision.

Description

Technical Field

[0001] This invention belongs to the field of automation control and rolling technology, specifically relating to a self-learning-based pre-compensation method for the impact rate drop of steel biting in rolling mills. Background Technology

[0002] In the production process of hot-rolled strip steel, the rolling mill may experience impact speed drop, where the mill speed suddenly decreases due to the impact of the strip at the moment of bite. This can cause speed fluctuations in the rolling mill and excessive strip head thickness. In severe cases, it can disrupt the normal speed relationship between stands, causing strip accumulation and affecting the stable operation of the production line. Adopting a speed pre-compensation method is an effective measure to solve this problem. This involves adding the impact speed drop compensation value to the preset speed setpoint of the rolling mill before the bite moment. Then, using a self-learning system, the difference between the setpoint and the actual speed is calculated segment by segment to correct the impact speed drop compensation.

[0003] In existing compensation methods, the accuracy of the mill impact descent compensation value is not high. Most methods do not correct for the actual mill speed, and the timing of reversing the given descent compensation value is difficult to control. Too much or too little descent compensation value, or too fast or slow timing, will affect the quality of the finished rolled product. Summary of the Invention

[0004] To address the aforementioned problems, this invention provides a self-learning-based mill bite impact rate drop compensation method, which improves upon the low accuracy and large timing deviation of existing impact rate drop pre-compensation methods. By combining a self-learning system, the difference between the given and actual mill speed is calculated in segments to correct the impact rate drop compensation, thereby minimizing mill speed fluctuations.

[0005] The technical solution adopted in this invention is as follows:

[0006] A self-learning-based pre-compensation method for impact rate drop in rolling mill bite, the method comprising:

[0007] (1) Based on the historical data of the impact speed drop compensation value corresponding to the steel grade to be rolled, pre-compensate the speed setting of the rolling mill before the steel bites;

[0008] (2) Set the total time that needs to be compensated, and divide the total time into segments;

[0009] (3) Calculate the difference between the given value and the actual value of the mill speed in each time period, and calculate the average value of the speed difference. Use the average value of the speed difference in each time period as the correction value for the corresponding time period.

[0010] The correction value is compensated in the impact velocity drop compensation value of the next strip steel corresponding to the time period according to a certain proportion to obtain the corrected impact velocity drop compensation value; the corrected impact velocity drop compensation value is compensated in the mill speed setting of the next strip steel at the corresponding time.

[0011] (4) After the compensation is completed, record the impact velocity drop compensation curve of the current steel grade.

[0012] Further, step (1) specifically involves: before the strip enters the rolling mill to be rolled, identifying the corresponding steel grade, calling up the historical data of the impact speed drop compensation value corresponding to the currently rolled steel grade, and adding the impact speed drop compensation value to the rolling mill speed setting before the steel bites.

[0013] Further, in step (1), the historical data of the impact rate drop compensation value is the impact rate drop data of the same specification steel grade that was previously recorded; if the current steel grade is rolled for the first time and there is no corresponding data, then a fixed value is set as the impact rate drop compensation value based on the rolling experience, and it is called as historical data.

[0014] Furthermore, in step (2), the total time required for impact velocity reduction compensation is set and divided into segments of 5ms to 15ms; the starting point of the compensation time is the moment of steel biting, and the ending point is the total time of compensation.

[0015] Furthermore, in step (3), the formula for calculating the average value of the speed difference between the given and actual mill speed values ​​within each time interval is as follows:

[0016]

[0017] in, v is the average difference between the setpoint and actual mill speed during the current time period. acti v is the actual rotational speed of the rolling mill at the corresponding moment. refi is the given rotational speed of the rolling mill at the corresponding time; n is the number of data sets collected by the data acquisition system within the current time period.

[0018] Furthermore, in step (3), the calculation method for the corrected impact velocity drop compensation value is as follows:

[0019] Based on the average value of the speed difference during the current time period. A self-learning system is introduced to correct the original impact velocity drop compensation value for the current time period. The correction formula is as follows:

[0020]

[0021] Where, N dj N is the corrected impact drop compensation value used in the current time period. d1jThis is the impact velocity drop compensation value for the current time period corresponding to the previous rolling. denoted as the average value of the speed difference, and 'a' as the proportional coefficient.

[0022] Furthermore, the value of 'a' ranges from 10% to 20%.

[0023] Furthermore, in step (3), the time delay of the actuator needs to be considered, that is, the time required for the actual motor speed to meet the requirements from the given speed in the program needs to be considered, and then the corrected impact speed drop compensation value is compensated into the mill speed given when rolling the next piece of steel of the same specification.

[0024] Furthermore, the requirement to consider the time delay of the actuator is based on the historical mill speed curve, where the total delay time t is set. d During the compensation process, the corrected impact velocity drop compensation value is advanced by t. d The time compensation is incorporated into the mill speed setting.

[0025] Furthermore, in step (4), the recording of the impact velocity drop compensation curve of the current steel grade is to record the impact velocity drop compensation curve in the data of the corresponding steel grade after the impact velocity drop compensation is completed, and to call it as historical data when rolling steel grades of the same specification in the future.

[0026] The beneficial effects of the technical solutions provided by the embodiments of the present invention include at least the following:

[0027] The method provided by this invention compensates the impact velocity drop value into the mill speed setting before the steel bites, based on historical data of the impact velocity drop value corresponding to the steel grade to be rolled. The compensation method is to gradually add the compensation value to the speed setting before the steel bites, so as to prevent the mill speed from surging in a short period of time, and record the speed setting value.

[0028] The method provided by this invention pre-compensates for the impact drop of steel biting in the rolling mill by combining self-learning, which can more accurately compensate for the rolling mill speed and make the process of canceling the compensation value smoother, reducing the fluctuation of the rolling mill speed.

[0029] The method provided by this invention records the impact velocity drop compensation curve of the current steel grade, and calls it when rolling steel grades of the same specification in the future. It uses a large amount of data to optimize the impact velocity drop compensation curve, so that the compensation value is more accurate.

[0030] The method provided by this invention can effectively improve the problem of impact speed drop when steel is bitten in rolling mills, and improve production stability and head size accuracy. Attached Figure Description

[0031] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0032] Figure 1 This is a flowchart illustrating a self-learning-based rolling mill impact speed reduction compensation method in an embodiment of the present invention.

[0033] Figure 2 This is a schematic curve of the impact speed drop of the rolling mill in an embodiment of the present invention. In the figure, the vertical axis represents the rotational speed of the rolling mill, and the horizontal axis represents time.

[0034] Figure 3 This refers to the time required for the actual motor speed to meet the requirements from the given speed of the rolling mill in this embodiment of the invention. In the figure, the vertical axis represents the rolling mill speed, the horizontal axis represents time, the dashed line is a schematic curve of the given speed of the rolling mill, and the solid line is a schematic curve of the actual speed of the rolling mill. d This represents the total delay time.

[0035] Figure 4 This is a schematic curve of the mill speed after compensation in an embodiment of the present invention. In the figure, the vertical axis is the mill speed, the horizontal axis is time, the dashed line is the schematic curve of the mill speed before compensation, and the solid line is the schematic curve of the mill speed after compensation. Detailed Implementation

[0036] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.

[0037] This invention provides a self-learning-based pre-compensation method for the impact rate drop of steel biting in rolling mills, such as... Figure 1 The diagram shown is a flowchart of the method in an embodiment of the present invention. The method pre-compensates for the impact rate drop of steel biting in the rolling mill by incorporating a self-learning approach.

[0038] The method includes:

[0039] (1) Based on the historical data of the impact speed drop compensation value corresponding to the steel grade to be rolled, pre-compensate the speed setting of the rolling mill before the steel bites;

[0040] (2) Set the total time that needs to be compensated, and divide the total time into segments;

[0041] (3) Calculate the difference between the given value and the actual value of the mill speed in each time period, and calculate the average value of the speed difference. Use the average value of the speed difference in each time period as the correction value for the corresponding time period.

[0042] The correction value is compensated in the impact velocity drop compensation value of the next strip steel corresponding to the time period according to a certain proportion to obtain the corrected impact velocity drop compensation value; the corrected impact velocity drop compensation value is compensated in the mill speed setting of the next strip steel at the corresponding time.

[0043] (4) After the compensation is completed, record the impact velocity drop compensation curve of the current steel grade, and call it when rolling the same steel grade later.

[0044] In this embodiment, step (1) specifically involves: before the strip enters the rolling mill to be rolled, identifying the corresponding steel grade, calling up the historical data of the impact velocity drop compensation value corresponding to the currently rolled steel grade, and adding the impact velocity drop compensation value to the rolling mill speed setting before the strip bites. In this step, the historical data of the impact velocity drop compensation value is the impact velocity drop data of the same specification steel grade that was previously recorded; if the current steel grade is being rolled for the first time and there is no corresponding data, then a fixed value is set as the impact velocity drop compensation value based on rolling experience, and this is called up as historical data.

[0045] In this embodiment, in step (2), the total time required for impact velocity reduction compensation is set, and divided into segments of 10ms each; the starting point of the compensation time is the moment of steel biting, and the ending point is the set total compensation time. That is, as follows... Figure 2 The range from t1 to t2 shown represents the total compensation time;

[0046] Specifically, in this embodiment, in step (2), the total time for impact velocity reduction compensation is set to 300ms based on historical impact velocity reduction curves, and divided into 30 segments of 10ms each. Each segment is calculated independently and then compensated separately. The set total compensation time can be modified according to specific circumstances during implementation.

[0047] Step (3) specifically includes the following steps:

[0048] Step S3.1: According to step (2), the total compensation time is divided into 30 segments, each segment lasting 10ms; if the data acquisition system collects data once every 1ms, it can collect 10 sets of data in each segment.

[0049] Step S3.2: Calculate the speed difference between the setpoint and the actual speed of the rolling mill collected in each time period, and take the average of the speed difference of 10 sets of data as the correction value for this time period;

[0050] Considering the time delay of the actuator, the correction value is compensated in the impact velocity drop compensation value of the next strip steel corresponding to the time period according to a certain proportion, so as to obtain the corrected impact velocity drop compensation value.

[0051] Step S3.3: The corrected impact speed drop compensation value is incorporated into the mill speed setting at the corresponding moment of the next piece of steel to prevent the impact speed drop compensation value from being withdrawn too early or too late, so that the mill speed can be kept as stable as possible.

[0052] Specifically, step S3.2 is implemented as follows:

[0053] Step S3.2.1: The formula for calculating the average difference between the setpoint and actual mill speed for each of the 10 data sets within each time period is as follows:

[0054]

[0055] Among them, the calculation results v is the average difference between the setpoint and actual mill speed during the current time period. acti v is the actual rotational speed of the rolling mill at the corresponding moment. refi This represents the given rotational speed of the rolling mill at the corresponding moment.

[0056] Step S3.2.2: Based on the calculated average value of the speed difference over the current time period. A self-learning system is introduced to correct the original impact velocity drop compensation value for the current time period. The correction formula is as follows:

[0057]

[0058] Where, N dj N is the corrected impact drop compensation value used in the current time period. d1j This is the impact velocity drop compensation value for the current time period corresponding to the previous rolling. This is the average value of the speed difference. The 20% in the formula is to prevent excessive fluctuations in the mill speed caused by compensating for the entire difference at once. This percentage can be adjusted according to actual production.

[0059] Step S3.2.3: The corrected impact speed drop compensation value calculated in step S3.2.2 is used to compensate the mill speed setting when rolling the next piece of steel of the same specification, according to the time corresponding to the current impact speed drop value and taking into account the time required for the actual motor speed to meet the requirements from the given speed in the program.

[0060] In step (3) of this embodiment, the time delay of the actuator needs to be considered, that is, the time required for the actual motor speed to meet the requirements from the given speed of the program needs to be considered, and the corrected impact speed drop compensation value is compensated into the mill speed given when rolling the next piece of steel of the same specification.

[0061] In this embodiment, the need to consider the time delay of the actuator, such as Figure 3 As shown, specifically, the total delay time t is set based on historical rolling mill speed curves. d During the compensation process, the advance time is t. d The time will be used to compensate the corrected impact speed drop value into the mill speed setpoint.

[0062] In step (4) of this embodiment, the recording of the impact velocity drop compensation curve of the current steel grade is to record the impact velocity drop compensation curve in the data of the corresponding steel grade after the impact velocity drop compensation is completed, and then call it as historical data when rolling steel grades of the same specification in the future.

[0063] In a preferred embodiment of the present invention, the acquisition of the setpoint and actual values ​​of the rolling mill speed, as well as the identification of the moment of steel bite, are achieved using software such as ibaAnalyzer and ODG that monitor rolling production data. At each moment, the difference between the setpoint and actual rolling mill speed, the average value of the data for each time period, and the correction of the impact speed drop are all implemented using PLC programming.

[0064] This invention improves upon existing impact velocity drop pre-compensation methods, which suffer from low accuracy and inability to compensate for impact velocity drop at corresponding moments. It pre-compensates for the impact velocity drop caused by steel biting in rolling mills by incorporating a self-learning approach. The compensated rolling mill speed is as follows: Figure 4 As shown, this allows for more precise compensation of the mill speed and a smoother process for canceling the compensation value, reducing mill speed fluctuations.

[0065] The method provided by this invention employs a self-learning approach in the mill impact speed drop compensation value, and considers the time delay of the actuator. It adds correction amounts at corresponding moments, enabling more accurate compensation of the rotational speed at the moment of mill speed drop, resulting in a more stable mill rotational speed.

[0066] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A self-learning-based pre-compensation method for impact rate drop in rolling mill bite, characterized in that, The method includes: (1) Based on the historical data of the impact speed drop compensation value corresponding to the steel grade to be rolled, pre-compensate the speed setting of the rolling mill before the steel bites; (2) Set the total time that needs to be compensated, and divide the total time into segments; (3) Calculate the difference between the given value and the actual value of the mill speed in each time period, and calculate the average value of the speed difference. Use the average value of the speed difference in each time period as the correction value for the corresponding time period. The correction value is compensated in the impact velocity drop compensation value of the next strip steel corresponding to the time period according to a certain proportion to obtain the corrected impact velocity drop compensation value; the corrected impact velocity drop compensation value is compensated in the mill speed setting of the next strip steel at the corresponding time. (4) After the compensation is completed, record the impact velocity drop compensation curve of the current steel grade; In step (2), the total time required for impact drop compensation is set and divided into segments of 5ms to 15ms. The starting point of the compensation time is the moment of steel biting, and the ending point is the total time of compensation. In step (3), the formula for calculating the average value of the speed difference between the given and actual mill speed values ​​within each time period is as follows: in, v is the average difference between the setpoint and actual mill speed during the current time period. acti v is the actual rotational speed of the rolling mill at the corresponding moment. refi is the given rotational speed of the rolling mill at the corresponding time; n is the number of data sets collected by the data acquisition system within the current time period; In step (3), the calculation method for the corrected impact velocity drop compensation value is as follows: Based on the average value of the speed difference during the current time period. A self-learning system is introduced to correct the original impact velocity drop compensation value for the current time period. The correction formula is as follows: Where, N dj N is the corrected impact drop compensation value used in the current time period. d1j This is the impact velocity drop compensation value for the current time period corresponding to the previous rolling. is the average value of the speed difference, and 'a' is the proportionality coefficient; In step (4), the recording of the impact velocity drop compensation curve of the current steel grade is to record the impact velocity drop compensation curve in the data of the corresponding steel grade after the impact velocity drop compensation is completed, and then call it as historical data when rolling steel grades of the same specification in the future.

2. The pre-compensation method for impact rate drop in rolling mill bite based on self-learning as described in claim 1, characterized in that, Step (1) is as follows: before the strip enters the rolling mill to be rolled, the corresponding steel grade is identified, the historical data of the impact speed drop compensation value corresponding to the currently rolled steel grade is called, and the impact speed drop compensation value is added to the speed setting of the rolling mill before the steel bites.

3. The pre-compensation method for impact rate drop in rolling mill bite based on self-learning as described in claim 2, characterized in that, In step (1), the historical data of the impact velocity drop compensation value is the impact velocity drop data of the same specification steel grade that was previously recorded; if the current steel grade is rolled for the first time and there is no corresponding data, then a fixed value is set as the impact velocity drop compensation value based on the rolling experience, and it is called as historical data.

4. The pre-compensation method for impact rate drop in rolling mill bite based on self-learning as described in claim 1, characterized in that, The value of 'a' ranges from 10% to 20%.

5. The pre-compensation method for impact rate drop in rolling mill bite based on self-learning as described in claim 1, characterized in that, In step (3), the time delay of the actuator needs to be considered, that is, the time required for the actual motor speed to meet the requirements from the given speed in the program needs to be considered, and then the corrected impact speed drop compensation value is compensated into the mill speed given when rolling the next piece of steel of the same specification.

6. The pre-compensation method for impact rate drop in rolling mill bite based on self-learning as described in claim 5, characterized in that, The need to consider the time delay of the actuator is based on the historical mill speed curve, where the total delay time t is set. d During the compensation process, the corrected impact velocity drop compensation value is advanced by t. d The time compensation is incorporated into the mill speed setting.

Images