Intelligent roller type plate straightening machine

Abstract

An intelligent roller type plate straightening machine is characterized in that a screw-down mechanism of a feeding roller or a bidirectional feeding roller or a beam is provided with a load pressure sensor, a displacement pressure curve or broken line, actually measured plate thickness and rolling reduction of each roller are recorded when a workpiece enters the first three rollers and is automatically fed and pressed down to a preset position, and the qualified degree of workpiece discharging unevenness is measured; inputting the data into a data storage unit according to part numbers; the highest workpiece unevenness qualification degree is used as a target value, related programs of machine learning are used for learning and training the original related data, and when a new same workpiece is corrected, according to workpiece parameters such as actually measured plate thickness, plate width and displacement pressure curves or broken lines when a feeding roller is automatically fed and pressed to a preset position when the workpiece enters the first three rollers, the workpiece unevenness qualification degree is corrected according to the workpiece parameters such as the actually measured plate thickness, the actually measured plate width and the actually measured displacement pressure curves or broken lines. The rolling reduction of each roller is automatically selected, and the rolling reduction of individual or a small number of feeding rollers at the discharging position is finely adjusted according to the actually measured workpiece discharging unevenness qualification degree, so that the optimal unevenness qualification degree of the workpiece is guaranteed.

Description

Technical Field

[0001] The present invention belongs to forging equipment for processing metal materials with working rollers, and relates to the field of roller plate straightening machines, and specifically to an intelligent method and device for roller plate straightening machines, and an intelligent roller plate straightening machine. Background Art

[0002] Most of the existing roller plate straightening machines are manually controlled. CNC straightening machines can only control the synchronization and position accuracy of the working rollers, perform corrections according to calculated or preset process parameters, and provide feedback and corrections afterwards. The accuracy and efficiency are low, and the actual measurement of the workpiece unevenness can only be carried out after the workpiece has been discharged for a certain length. Summary of the invention

[0003] The problem to be solved by the present invention is to provide an intelligent roller plate straightening machine which can automatically control the straightening in real time and can provide feedback correction when the workpiece starts to be discharged, thereby improving the straightening accuracy and efficiency.

[0004] In order to achieve the above purpose, an intelligent roller plate straightening machine is provided, which is characterized in that: a load pressure sensor is provided on the pressing mechanism of the feed roller or the bidirectional feed roller or the beam, and a displacement pressure curve or a broken line is recorded when the feed roller automatically feeds and presses down to a preset position when the workpiece enters the first three rollers, and the plate thickness, the pressing amount of each roller, and the qualified degree of the workpiece discharge unevenness are measured, and the above data are input into a data storage unit according to the part number; the highest qualified degree of the workpiece unevenness is taken as the target value, and the above original related data is learned and trained with the relevant procedures of machine learning, and when correcting a new workpiece of the same kind, according to Measure the plate thickness, plate width, displacement pressure curve or broken line of the workpiece when the feed roller automatically feeds and presses down to the preset position when the workpiece enters the first 3 rollers, and automatically select the pressing amount of each roller, and according to the measured workpiece discharge unevenness qualification, fine-tune the pressing amount of individual or a small number of feed rollers at the discharge to ensure the best unevenness qualification of the workpiece; the measurement of workpiece unevenness can be online or after the fact, can be automatically measured by sensor or manually measured, the degree of workpiece qualification can be scored, and the score can be positive or negative, convex is positive, concave is negative, and the absolute value is used to judge the qualification.

[0005] Usually, the mechanical properties of materials are measured by using a tensile test bar method to measure a stress-strain curve. The present invention uses a bending method to measure a displacement-pressure curve or a broken line. As the feed amount of the working roll increases, the workpiece also undergoes an elastic deformation stage, a yield stage, and a strengthening stage. Like the tensile test bar, as the feed amount of the working roll increases until a fixed preset feed amount is reached, the stress and strain also gradually increase and sequentially undergo the above three stages. Using the bending method to measure the displacement-pressure curve or a broken line is also to measure the mechanical properties of the material, but the material has not been processed. Different plate thicknesses have different effects, and mastering the parameters of the plate thickness can reduce the effects.

[0006] The intelligent roller plate straightening machine is characterized in that: a rotary encoder or other sensor for measuring the rotation angle and rotation speed of the working roller is provided on the main transmission device of the relevant roller; when the relevant roller is a passive roller, a measuring transmission device is provided to drive the rotary encoder or other sensor for measuring the rotation angle and rotation speed of the working roller; the pulse data Mn collected by the sensor on the main transmission device of the last working roller at the discharge position is related to the length of the surface passed after the workpiece rebounds and the unevenness of the workpiece; the pulse data Mi collected by the sensor on the main transmission device of the other working rollers except the feed side roller is related to the length and curvature of the surface passed before the workpiece bends and rebounds; the unevenness qualification of the workpiece after actual correction is measured, and the correlation between the unevenness qualification of the workpiece and the pulse data collected by the sensor on the main transmission device of the working roller is established; at the same time, the plate thickness, the displacement pressure curve or broken line when the feed roller automatically feeds and presses down to the preset position when the workpiece enters the first 3 rollers, the pressing amount of each working roller and other parameters are recorded, and the above data are entered into the data according to the part number Storage unit; taking the highest qualified degree of workpiece unevenness as the target value, using the relevant procedures of machine learning to learn and train the above original related data, when correcting a new workpiece of the same kind, according to the measured plate thickness, plate width, displacement pressure curve or broken line when the feed roller is pressed down to the preset position when the workpiece enters the first 3 rollers and other workpiece parameters, automatically select the correlation between the amount of each roller's depression and the workpiece unevenness qualified degree and the pulse data collected by the sensor on the working roller main transmission device; before the workpiece can measure the qualified degree of unevenness, monitor and control the workpiece unevenness, and according to the measured qualified degree of workpiece discharge unevenness, fine-tune the depression amount of individual or a small number of feed rollers at the discharge to ensure the best qualified degree of unevenness of the workpiece; or when the correction machine does not or cannot measure the qualified degree of unevenness, use the correlation between the qualified degree of workpiece unevenness and the pulse data collected by the sensor on the working roller main transmission device to fine-tune the depression amount of individual or a small number of feed rollers at the discharge to ensure the best qualified degree of unevenness of the workpiece.

[0007] Since the working rollers are elastically bent and deformed when loaded, the axial curvature of the workpiece is inconsistent and the thickness and material of the workpiece are uneven. The pulse data collected by the sensors on each roller are not completely equal to the actual length of the workpiece surface, but are related. For workpieces with relatively uniform thickness and material and a fixed plate rolling machine, this correlation is stable. This method is used before the unevenness of the workpiece can be measured online, or used in conjunction with claim 1.

[0008] The intelligent roller plate straightening machine is characterized in that: the workpiece unevenness measurement is set on one side or both sides of the discharge roller, and the longitudinal unevenness qualification and convexoconcave of the workpiece can be quickly measured as soon as the workpiece is discharged, and immediate feedback is given. The adjustment of the feed roller at the discharge mainly improves the longitudinal accuracy of the workpiece, and the workpiece unevenness measurement is set on one side or both sides of the discharge roller. The workpiece longitudinal unevenness qualification and convexoconcave can be quickly measured as soon as the workpiece is discharged, realizing rapid feedback and improving the correction accuracy.

[0009] A workpiece centering device is installed on the feed roller table. The workpiece surface should be clean and free of water, snow, oil and burrs. The working roller should have sufficient rigidity, processing accuracy and cleanliness to ensure the stability and repeatability of Mi.

[0010] The beneficial effects of the present invention are: intelligent control of workpiece correction in real time, improving the degree of correction automation, improving the correction accuracy of the workpiece, and improving the correction efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 is a schematic diagram of embodiment 1 of the present invention;

[0012] Figure 2 is a schematic diagram of Embodiment 2 of the present invention;

[0013] Figure 3 is a schematic diagram of embodiment 3 of the present invention;

[0014] Figure 4 is a schematic diagram of embodiment 4 of the present invention;

[0015] In the figure, Mn-i is the number of rotary encoder pulses collected by the ni-th roller in a certain period of time, which is related to the length and curvature of the inner surface of the workpiece before rebounding; Mn is the number of rotary encoder pulses collected by the working roller at the discharge side in a certain period of time, which is related to the length and unevenness of the surface of the workpiece after rebounding; DETAILED DESCRIPTION

[0016] Embodiment 1:

[0017] Figure 1It is a nine-roller plate straightening machine, with 4 on top and 5 on the bottom, bidirectional feeding, and hydraulically pressed side rollers. It has displacement sensors and pressure sensors. When the workpiece enters the first 3 rollers, the feed rollers automatically feed and press down to the preset position, and then record the displacement pressure curve or broken line of the working rollers, which is equivalent to measuring the mechanical properties of the workpiece, recording the pressure of the feed rollers at the discharge end, measuring the plate thickness, recording the plate width and the pressing amount of each roller, and measuring the unevenness qualification of the workpiece after correction. The above data are input into the data storage unit according to the part number, and a correction parameter data set for the workpiece with the part number is established; the highest unevenness qualification of the workpiece is taken as the target value, and the above original related data are learned and trained with the relevant programs of machine learning, and the correction is carried out. When checking a new workpiece of the same kind, the pressure reduction of each roller is automatically selected based on the measured plate thickness, plate width, displacement pressure curve or broken line when the feed roller automatically feeds and presses down to the preset position when the workpiece enters the first 3 rollers and other workpiece parameters. After the workpiece is discharged for a certain length, the pressure reduction of individual or a small number of feed rollers at the discharge end is automatically fine-tuned according to the measured unevenness qualification of the workpiece discharge and the pressure of the feed roller at the discharge end to ensure the best unevenness qualification of the workpiece. The measurement of the workpiece unevenness can be measured online or afterwards, with automatic measurement sensors or manual measurement. The degree of qualification of the workpiece can be scored, and the score can be positive or negative, with convexity being positive and concave being negative, and the absolute value is used to judge the qualification.

[0018] Embodiment 2:

[0019] Figure 2 It is the same nine-roller leveling machine as in Example 1. A rotary encoder is provided on the transmission device of the n-2, n-1, and n-th rollers. The number of pulses emitted is Mn-2, Mn-1, and Mn. Mn is related to the length of the surface after the workpiece rebounds and the unevenness of the workpiece; Mn-2 and Mn-1 are related to the length and curvature of the surface passed by the workpiece before bending and rebounding at the n-2 and n-1 working rollers; the unevenness qualification of the workpiece after actual correction is measured, and the correlation between the unevenness qualification of the workpiece and the pulse data collected by the sensor on the main transmission device of the working roller is established. At the same time, the plate thickness, the displacement pressure curve or broken line when the feed roller automatically feeds and presses down to the preset position when the workpiece enters the first three rollers, the pressure of the feed roller at the discharge end, the pressing amount of each working roller and other parameters are recorded. The above parameters are combined according to the part number. Data is input into a data storage unit; with the highest qualified degree of workpiece unevenness as the target value, the above original related data are learned and trained with relevant programs of machine learning. When correcting a new workpiece of the same kind, the correlation between the amount of each roller pressed down and the qualified degree of workpiece unevenness and the pulse data collected by the sensor on the main transmission device of the working roller is automatically selected according to the measured plate thickness, plate width, displacement pressure curve or broken line when the feed roller is pressed down to the preset position when the workpiece enters the first 3 rollers and other workpiece parameters. Before the qualified degree of unevenness of the workpiece can be measured, the unevenness of the workpiece is monitored and controlled. After the workpiece is corrected to a certain length, according to the measured qualified degree of unevenness of the workpiece discharge and the pressure of the feed roller at the discharge end, the amount of pressure of individual or a small number of feed rollers at the discharge is fine-tuned to ensure the best qualified degree of unevenness of the workpiece.

[0020] In the same time period, Mn>Mn-1>Mn-2. When the workpiece is concave, Mn increases, and when the workpiece is convex, Mn decreases. When the pressure of other rollers is constant, fine-tune the pressure of the feed roller at the discharge point. Increasing the pressure will reduce Mn-2 and Mn-1, and reducing the pressure will increase Mn-2 and Mn-1. However, the change of Mn is uncertain. If the workpiece is concave, Mn needs to be reduced. If the workpiece is convex, Mn needs to be increased. Only the machine can learn by itself. When it is measured that Mn-2, Mn-1, and Mn meet a certain proportion under the condition of known pressure reduction, the workpiece unevenness is qualified. Mn represents the length through which the neutral layer passes.

[0021] When the leveling machine does not have or cannot measure the unevenness qualification, the correlation between the unevenness qualification of the workpiece and the pulse data collected by the sensor on the main transmission device of the working roller is used to fine-tune the depression amount of individual or a small number of feed rollers at the discharge point to ensure the best unevenness qualification of the workpiece; or as an auxiliary tool of claim 1, monitor and control the unevenness of the workpiece.

[0022] According to the continuity principle and the properties of the neutral layer, the length of the neutral layer passed by each roller is the same in the same time period, so Mn is related to the length of the surface passed by the workpiece after rebound and the unevenness of the workpiece; Mn-2 and Mn-1 are related to the surface length and curvature of the workpiece before bending and rebounding at the n-2 and n-1 working rollers. The n-2, n-1, and n rollers are single roller main drives.

[0023] Embodiment 3:

[0024] Figure 3 and Figure 1For the same nine-roller plate straightening machine, a rotary encoder is set on the transmission device of the n-2 and n rollers, and the number of pulses sent is Mn-2 and Mn, where Mn is related to the length of the surface passed by the workpiece after rebound and the unevenness of the workpiece; Mn-2 is related to the surface length and curvature passed by the workpiece before bending and rebounding at the n-2 working roller; the unevenness qualification of the workpiece after actual correction is measured, and the correlation between the unevenness qualification of the workpiece and the pulse data collected by the sensor on the main transmission device of the working roller is established. At the same time, the plate thickness, the displacement pressure curve or broken line when the feed roller automatically feeds and presses down to the preset position when the workpiece enters the first 3 rollers, the pressure of the feed roller at the discharge end, the pressing amount of each working roller and other parameters are recorded, and the above data are input into the data storage unit according to the part number. ; Taking the highest qualified degree of workpiece unevenness as the target value, the above original related data are learned and trained with the relevant programs of machine learning. When correcting a new workpiece of the same kind, according to the measured plate thickness, plate width, the displacement pressure curve or broken line when the feed roller is pressed down to the preset position when the workpiece enters the first 3 rollers and other workpiece parameters, the correlation between the pressure reduction amount of each roller and the qualified degree of workpiece unevenness and the pulse data collected by the sensor on the main transmission device of the working roller is automatically selected. Before the qualified degree of unevenness of the workpiece can be measured, the unevenness of the workpiece is monitored and controlled. After the workpiece is corrected to a certain length, according to the measured qualified degree of unevenness of the workpiece discharge and the pressure of the feed roller at the discharge end, the pressure reduction amount of individual or a small number of feed rollers at the discharge is fine-tuned to ensure the best qualified degree of unevenness of the workpiece.

[0025] In the same time period, Mn>Mn-2, except for the n-1 roller, the pressure of each roller remains unchanged, and Mn-2 is fixed. For example, when the machine obtains Mn-1 of 1000 pulses, Mn1 of 150 pulses, the roughness qualification of the workpiece is the best, and the roughness of the workpiece with positive and negative 2 pulses of Mn is within the qualified range. When the roughness qualification of the workpiece cannot be measured, keep Mn-1 at 1150 pulses of Mn for every 1000 pulses, and keep the positive and negative 2 pulses unchanged, and adjust the pressure of the feed roller at the discharge point.

[0026] When the leveling machine does not have or cannot measure the unevenness qualification, the correlation between the unevenness qualification of the workpiece and the pulse data collected by the sensor on the main transmission device of the working roller is used to fine-tune the depression amount of individual or a small number of feed rollers at the discharge point to ensure the best unevenness qualification of the workpiece; or as an auxiliary tool of claim 1, monitor and control the unevenness of the workpiece.

[0027] Embodiment 4:

[0028] Figure 4In the nine-roller plate leveling machine shown, a workpiece unevenness measuring device is provided on one or both sides of the nth roller of the discharging roller. As soon as the workpiece is discharged, the longitudinal unevenness qualification and convexoconcavity of the workpiece can be quickly measured, and feedback can be given immediately. The adjustment of the feed roller at the discharging place mainly improves the longitudinal accuracy of the workpiece. The workpiece unevenness measurement is provided on one or both sides of the discharging roller. As soon as the workpiece is discharged, the longitudinal unevenness qualification and convexoconcavity of the workpiece can be quickly measured, and rapid feedback can be achieved to improve the correction accuracy.

[0029] According to the displacement pressure curve or broken line of the feed roller when the workpiece enters the first 3 rollers and is automatically fed and pressed to the preset position, the pressure of the feed roller at the discharge end, the measured plate thickness, plate width and the pressing amount of each roller and the measured longitudinal unevenness qualification and convexity and concavity of the workpiece, the pressing amount of the feed roller at the discharge end is automatically fine-tuned to achieve the best longitudinal unevenness qualification of the workpiece.

[0030] If a device for measuring the degree of lateral unevenness is provided, the pressing amount of the feed roller and the supporting roller at the feeding point can be automatically adjusted.

Claims

1. An intelligent roller plate straightening machine, characterized by: The pressing mechanism of the feed roller or bidirectional feed roller or beam is equipped with a load pressure sensor to record the displacement pressure curve or broken line when the feed roller automatically feeds and presses down to the preset position when the workpiece enters the first three rollers, measure the plate thickness, the pressing amount of each roller, and the qualified degree of the workpiece discharge unevenness, and input the above data into the data storage unit according to the part number; Taking the highest qualified degree of workpiece unevenness as the target value, the above original related data are learned and trained with the relevant programs of machine learning. When correcting a new workpiece of the same kind, the pressure reduction amount of each roller is automatically selected according to the measured plate thickness, plate width, displacement pressure curve or broken line when the feed roller automatically feeds and presses to the preset position when the workpiece enters the first 3 rollers and other workpiece parameters, and according to the measured qualified degree of workpiece discharge unevenness, the pressure reduction amount of individual or a small number of feed rollers at the discharge is fine-tuned to ensure the best qualified degree of workpiece unevenness; the measurement of workpiece unevenness can be online measurement or post-measurement, automatic measurement sensor or manual measurement, the qualified degree of the workpiece can be scored, the score is positive and negative, convex is positive, concave is negative, and the absolute value is taken to judge the qualified degree.

2. The intelligent roller plate straightening machine according to claim 1 is characterized in that: A rotary encoder or other sensor for measuring the rotation angle and rotation speed of the working roll is provided on the main transmission device of the relevant roll. When the relevant roll is a passive roll, a measuring transmission device is provided to drive the rotary encoder or other sensor for measuring the rotation angle and rotation speed of the working roll; the pulse data MO collected by the sensor on the main transmission device of the last working roll at the discharge position is related to the length of the surface passed by the workpiece after rebound and the unevenness of the workpiece; the pulse data Mi collected by the sensor on the main transmission device of the other working rolls except the feed side roll is related to the length and curvature of the surface passed by the workpiece before bending and rebounding; Measure the roughness qualification of the workpiece after actual correction, establish the correlation between the roughness qualification of the workpiece and the pulse data collected by the sensor on the main transmission device of the working roll, and record the plate thickness, the displacement pressure curve or broken line when the feed roll automatically feeds and presses down to the preset position when the workpiece enters the first three rolls, the pressing amount of each working roll and other parameters, and input the above data into the data storage unit according to the part number; Taking the highest qualified degree of workpiece unevenness as the target value, the above original related data are learned and trained by the relevant procedures of machine learning. When correcting a new workpiece of the same kind, the pressure reduction amount of each roller and the correlation between the qualified degree of workpiece unevenness and the pulse data collected by the sensor on the main transmission device of the working roller are automatically selected according to the measured plate thickness, plate width, displacement pressure curve or broken line when the feed roller is pressed down to the preset position when the workpiece enters the first 3 rollers and other workpiece parameters. Before the qualified degree of unevenness of the workpiece can be measured, the unevenness of the workpiece is monitored and controlled, and according to the measured qualified degree of unevenness of the workpiece discharge, the pressure reduction amount of individual or a small number of feed rollers at the discharge is fine-tuned to ensure the best qualified degree of unevenness of the workpiece; or when the correcting machine does not have or cannot measure the qualified degree of unevenness, the correlation between the qualified degree of workpiece unevenness and the pulse data collected by the sensor on the main transmission device of the working roller is used to fine-tune the pressure reduction amount of individual or a small number of feed rollers at the discharge to ensure the best qualified degree of unevenness of the workpiece.

3. The intelligent roller plate straightening machine according to claim 1 or 2, characterized in that: the workpiece The unevenness measurement is set on one side or both sides of the discharge roller. As soon as the workpiece is discharged, the longitudinal unevenness qualification and convexity and concavity of the workpiece can be quickly measured and feedback can be given immediately.

Images