Intelligent Monitoring Method and System for a Forging Production Line

Abstract

This application relates to the technical field of forging production, and particularly to an intelligent monitoring method and system for a forging production line. The method includes: monitoring the stamping accuracy of multiple stamping strokes at any position point during the forging process of the forging, and taking the average value of the stamping accuracy as the forging quality of the position point; constructing a stamping area for each position point; weighted-summing the forging quality of the target stamping area of any position point to obtain the weighted forging quality of each position point, where the target stamping area is the stamping area containing the position point; and issuing a warning in response to the weighted forging quality of any position point being less than a preset quality value. Through the technical solution of this application, defects of forgings can be accurately identified, and the accuracy of the production quality monitoring results of the forging production line can be improved.

Description

Technical Field

[0001] This application relates to the technical field of forging production, and particularly to an intelligent monitoring method and system for a forging production line. Background Art

[0002] Forging is a processing technology that plastically deforms metal materials by applying pressure to obtain the required shape and performance; the forging process usually includes multiple processes such as blanking, heating, forging, cooling, and post-treatment, and the produced forgings are applied in fields such as aviation, automobiles, machinery manufacturing, petrochemicals, etc.

[0003] Currently, the patent application document with the publication number CN116977780A discloses a forging defect recognition method based on machine vision. The method includes: collecting and preparing normal and abnormal forging video samples; preprocessing and annotating the video samples; using deep learning technology to construct a forging defect abnormal recognition model; inputting the annotated samples into the forging defect abnormal recognition model for training. After the training is completed, the model can accurately identify the defects; preprocessing the real-time video of the forging process, and inputting the preprocessed real-time pictures into the model for recognition; identifying the defects in the forging process and automatically sending an alarm signal to the alarm device.

[0004] The above method directly inputs the preprocessed real-time pictures in the forging process into the forging defect abnormal recognition model to obtain the defects in the forging process. However, the real-time pictures contain a large number of features irrelevant to the forging defect abnormal recognition. Directly using the real-time pictures as the input of the forging defect abnormal recognition model will introduce a large amount of noise, resulting in the inability to accurately identify the defects in the forging process and inaccurate quality monitoring results of the forging production line. Summary of the Invention

[0005] In order to solve the technical problem of inaccurate quality monitoring results of the forging production line, this application provides an intelligent monitoring method and system for a forging production line, which can accurately identify the defects of forgings and improve the accuracy of the quality monitoring results of the forging production line.

[0006] In the first aspect of the present application, an intelligent monitoring method for a forging production line is provided. The monitoring method includes: monitoring the stamping accuracy of multiple stamping strokes at any position point during the forging process of the forging, and taking the average value of the stamping accuracy as the forging quality of the position point, including: obtaining the slider area and the workbench area of each forging image during a single stamping stroke, and taking the slope consistency between the lower edge of the slider area and the upper edge of the workbench area as the stamping parallelism, where the slider is connected to a pressure sensor; collecting the stamping parallelism sequence, the pressure value sequence, and the movement trajectory of the center point of the slider area from the start of the stamping stroke to the time when the pressure sensor reaches the maximum value; calculating the stamping accuracy of the stamping stroke, where the stamping accuracy is positively correlated with the average value of the stamping parallelism sequence, negatively correlated with the variance of the non-zero pressure values in the pressure value sequence, and negatively correlated with the offset of the movement trajectory of the center point of the slider area from the vertical direction; constructing the stamping area of each position point, where the size of the stamping area is related to the size of the slider and the average offset of multiple stamping strokes at the position point; weighted summing the forging quality of the target stamping area of any position point to obtain the weighted forging quality of each position point, where the target stamping area is the stamping area containing the position point; and issuing a warning in response to the weighted forging quality of any position point being less than a preset quality value.

[0007] During the forging process of the forging, it is necessary to perform multiple stamping strokes on each position point on the surface of the forging. The stamping accuracy of the stamping stroke directly affects the forging quality of each position point. Monitor the stamping accuracy of multiple stamping strokes at a position point during the forging process of the forging, and take the average value of the stamping accuracy as the forging quality of the position point; during the stamping of the position point the impact force generated by the slider not only acts on the position point but also acts on the surrounding area of the position point causing deformation of other position points in the surrounding area. Therefore, construct the stamping area of each position point according to the size of the slider and the average offset of multiple stamping strokes at the position point. The stamping area covers other position points that can be affected by the stamping stroke at the corresponding position point; a position point will appear in multiple stamping areas at the same time. Denote the multiple stamping areas where the position point is located as the target stamping area of the position point, and perform weighted summation on the forging quality of the target stamping area to obtain the weighted forging quality of each position point; the weighted forging quality can truly and accurately reflect the real forging quality of each position point. When the weighted forging quality of any position point is less than the preset quality value, it indicates that the position point is a defective point, and a warning is issued in time to improve the accuracy of the production quality monitoring result of the forging production line.

[0008] Preferably, the calculation method of the stamping parallelism includes: dividing the slider area into multiple slider sub-areas along the vertical direction, performing linear fitting on the lowest points of each slider sub-area to obtain the slope of the lower edge of the slider area; dividing the workbench area into multiple workbench sub-areas along the vertical direction, performing linear fitting on the highest points of each workbench sub-area to obtain the slope of the upper edge of the workbench area; the stamping parallelism is negatively correlated with the absolute value of the difference between the slope of the lower edge and the slope of the upper edge.

[0009] By comparing the slope of the lower edge of the slider area and the slope of the upper edge of the workbench area, the stamping parallelism is accurately quantified, and the stamping parallelism directly affects the stamping accuracy of the stamping stroke.

[0010] Preferably, the forging image of the stamping parallelism is: , and are respectively the slope of the lower edge and the slope of the upper edge in the forging image .

[0011] Preferably, the method for obtaining the offset of the movement trajectory of the center point of the slider area from the vertical direction includes: calculating the shortest distance from each center point of the slider area on the movement trajectory to a preset straight line, where the preset straight line is any vertical straight line; calculating the absolute value of the difference between each shortest distance and the average shortest distance to obtain the offset distance of each center point of the slider area, and taking the average value of the offset distances as the offset of the movement trajectory in the vertical direction.

[0012] The offset of the movement trajectory of the center point of the slider area from the vertical direction can reflect the lateral offset degree of the slider in the horizontal direction during the movement of the slider; the larger the offset, the greater the shaking degree of the slider during the stamping stroke, realizing the accurate quantification of the shaking degree of the slider during the stamping stroke.

[0013] Preferably, the stamping accuracy satisfies the relational expression:

[0014] , is the mean value of the stamping parallelism sequence, is the variance of the non-zero pressure values in the pressure value sequence, is the offset of the movement trajectory of the center point of the slider area from the vertical direction.

[0015] Preferably, the stamping area of the position point is a rectangular area centered on the position point , the width of the rectangular area, the length of the rectangular area, and are the width and length of the slider respectively, is the position point The average offset of each stamping stroke.

[0016] The average offset of the stamping stroke at any position point can reflect the swaying degree of the slider at that position point. Furthermore, the stamping areas of each position point can be accurately located according to the swaying degree, ensuring the accuracy of the weighted forging quality of each subsequent position point and avoiding the omission of defective position points.

[0017] Preferably, the position point The weighted forging quality is:

[0018] , is the number of target stamping areas, is the target stamping area The forging quality, is the position point And the target stamping area The Euclidean distance from the center point.

[0019] The stamping stroke at a position point on the surface of the forging will also affect other position points within the stamping area of that position point. Obtain all the stamping areas containing the position point , and the forging quality of the corresponding position points in these stamping areas will all affect the position point . Therefore, by comprehensively considering the forging quality of all the stamping areas containing the position point , the accurate monitoring of the true forging quality of the position point in the forging is realized, and the defects of the position point in the forging are accurately identified.

[0020] Preferably, the position points with weighted forging quality less than the preset quality value are marked as defective points to obtain the forging defects of the forging.

[0021] Preferably, a semantic segmentation network is used to obtain the slider area and the workbench area of each forging image during one stamping stroke.

[0022] In the second aspect of the present application, an intelligent monitoring system for a forging production line is further provided, including a processor and a memory. The memory stores computer program instructions, and when the computer program instructions are executed by the processor, an intelligent monitoring method for a forging production line according to the first aspect of the present application is implemented.

[0023] The technical solution of the present application has the following beneficial technical effects:

[0024] During the forging process of forgings, multiple stamping strokes need to be performed on each position point on the surface of the forging. The stamping accuracy of the stamping stroke directly affects the forging quality of each position point. Therefore, monitor the stamping accuracy of multiple stamping strokes at a position point during the forging process of the forging, and use the average value of the stamping accuracy as the forging quality of this position point; during the stamping of the position point In the process, the impact force generated by the slider will not only act on the position point but also act on the surrounding area of the position point causing deformation of other position points in the surrounding area. Therefore, construct the stamping area of each position point according to the slider size and the average offset of multiple stamping strokes at the position point. This stamping area covers other position points that can be affected by the stamping stroke at the corresponding position point; a position point will appear in multiple stamping areas at the same time. Denote the multiple stamping areas where the position point is located as the target stamping area of this position point, and perform a weighted sum of the forging quality of the target stamping area to obtain the weighted forging quality of each position point; the weighted forging quality can truly and accurately reflect the true forging quality of each position point. When the weighted forging quality of any position point is less than the preset quality value, it indicates that this position point is a defective point, and an alarm is issued in time to improve the accuracy of the production quality monitoring result of the forging production line. Brief Description of the Drawings

[0025] Figure 1 is a schematic diagram of the stamping stroke of the forging machine according to an embodiment of the present application;

[0026] Figure 2 is a flowchart of an intelligent monitoring method for a forging production line according to an embodiment of the present application;

[0027] Figure 3 is a structural block diagram of an intelligent monitoring system for a forging production line according to an embodiment of the present application. Detailed Embodiment

[0028] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are some, but not all, of the embodiments of the present application.

[0029] According to the first aspect of the present application, the present application provides an intelligent monitoring method for a forging production line, which is used to monitor the production quality of forgings on the forging production line; the forging machine is a key device on the forging production line, which is used to perform multiple stamping strokes on each position point of the metal material, so as to forge the metal material into a forging. Therefore, the stamping accuracy of the stamping stroke directly affects the forging quality of the forging. Please refer to Figure 1, which is a schematic diagram of the stamping stroke of a forging machine according to an embodiment of the present application. The forging machine includes a slider and a workbench. A metal material is placed on the workbench, and the position of the metal material on the workbench is controlled by a manipulator. The slider is disposed above the workbench and can reciprocate in the vertical direction. In one stamping stroke, the slider starts from the highest point and moves in the vertical direction, and finally falls on the metal material at a relatively high speed. The impact force generated by the slider will change the shape of the corresponding position point of the metal material, and then the slider returns to the highest point to complete one stamping stroke. A position point will go through multiple stamping strokes. After completing the stamping stroke of a position point, the manipulator will move the metal material, so as to complete the forging of any position point in the metal material and obtain a forging.

[0030] Figure 2 is a flowchart of an intelligent monitoring method for a forging production line according to an embodiment of the present application. As Figure 2 shown, the intelligent monitoring method for the forging production line includes steps S101 to S104, which are described in detail below.

[0031] S101, monitor the stamping accuracy of multiple stamping strokes at any position point during the forging process of the forging, and use the average value of the stamping accuracy as the forging quality of the position point.

[0032] In one embodiment, during the processing of the forging, multiple stamping strokes will be sequentially performed on each position point of the forging. An image acquisition device is used to acquire the forging images during each stamping stroke, and then the stamping accuracy of each stamping stroke is monitored based on the forging images. The greater the stamping accuracy, the more accurately the stamping stroke can change the shape of the corresponding position point in the forging. Therefore, for any position point in the forging, the average value of the stamping accuracy of multiple stamping strokes at this position point is used as the forging quality of this position point. The greater the forging quality, the closer the size of this position point is to the standard size.

[0033] Among them, the position points in the forging correspond one-to-one with the positions of the manipulator. When the position of the manipulator remains unchanged, the position points of the stamping stroke in the forging remain unchanged.

[0034] In one embodiment, the method for obtaining the stamping accuracy of the stamping stroke includes: obtaining the slider area and the workbench area of each forging image in one stamping stroke, and using the slope consistency between the lower edge of the slider area and the upper edge of the workbench area as the stamping parallelism. The slider is connected to a pressure sensor; from the start of the stamping stroke to the end when the pressure sensor reaches the maximum value, collect the stamping parallelism sequence, the pressure value sequence, and the movement trajectory of the center point of the slider area; calculate the stamping accuracy of the stamping stroke. The stamping accuracy is positively correlated with the average value of the stamping parallelism sequence, negatively correlated with the variance of the non-zero pressure values in the pressure value sequence, and negatively correlated with the offset of the movement trajectory of the center point of the slider area from the vertical direction.

[0035] Among them, the slider area and the workbench area of each forging image in one stamping stroke are obtained by using a semantic segmentation network. The semantic segmentation network can classify the pixel points in the forging image into three categories: the slider area, the workbench area, and the background area. Therefore, the output result of the semantic segmentation network includes 3 binary images, corresponding to the three categories respectively; in the binary image corresponding to the slider area, the pixel value of the pixel points located in the slider area is 1, and the pixel value of the pixel points not located in the slider area is 0. The semantic segmentation network can adopt a fully convolutional neural network such as UNet or FCN.

[0036] Among them, the calculation method of the stamping parallelism includes: dividing the slider area into multiple slider sub-areas along the vertical direction, performing a linear fitting on the lowest points of each slider sub-area to obtain the lower edge slope of the slider area; dividing the workbench area into multiple workbench sub-areas along the vertical direction, performing a linear fitting on the highest points of each workbench sub-area to obtain the upper edge slope of the workbench area; the stamping parallelism is negatively correlated with the absolute value of the difference between the lower edge slope and the upper edge slope.

[0037] Specifically, forging image 's stamping parallelism is: , and are respectively the lower edge slope and the upper edge slope in the forging image .

[0038] It can be understood that when the stamping parallelism is equal to 1, it means that after the slider falls on the workbench, the lower edge of the slider and the upper edge of the workbench can be perfectly matched, and the impact force of the slider can act well on the forging on the workbench; when the stamping parallelism is not equal to 1, there is a gap between the lower edge of the slider and the upper edge of the workbench. When the impact force of the slider acts on the workbench, the forging will have dimensional errors due to the gap between the lower edge of the slider and the upper edge of the workbench, resulting in a decrease in the stamping accuracy of the stamping stroke; therefore, the stamping accuracy is positively correlated with the mean value of the stamping parallelism sequence. The stamping parallelism sequence includes the stamping parallelism of each forging image during the period from the start of the stamping stroke to the time when the pressure sensor reaches the maximum value. When the mean value is larger, it means that the matching degree between the lower edge of the slider and the upper edge of the workbench is better, and the stamping accuracy is higher.

[0039] During a stamping stroke, when the pressure sensor reaches its maximum value, it indicates that the stamping stroke has reached the lowest point. At this time, the maximum value of the pressure sensor is the impact force generated by the slider on the forging. Ideally, the slider only generates an impact force on the forging at the lowest point. That is to say, ideally, there is only a non-zero pressure value at the lowest point in the pressure value sequence, and the pressure values at other times are all 0. When there is relative movement between the slider and the workbench, the pressure sensor connected to the slider will collect non-zero pressure values other than the maximum pressure value. At this time, the variance of the non-zero pressure values in the pressure value sequence will increase, and the relative movement between the slider and the workbench will cause dimensional errors in the forging. Therefore, the larger the variance of the non-zero pressure values in the pressure value sequence, the greater the relative movement between the slider and the workbench, and the lower the stamping accuracy of this stamping stroke.

[0040] The offset of the center point of the slider area from the vertical direction in the movement trajectory can reflect the degree of lateral offset of the slider in the horizontal direction during the vertical movement of the slider. The larger the offset, the greater the degree of sway of the slider during the stamping stroke, and the lower the stamping accuracy of the stamping stroke.

[0041] Specifically, the method for obtaining the offset of the center point of the slider area from the vertical direction in the movement trajectory includes: calculating the shortest distance from each center point of the slider area on the movement trajectory to a preset straight line, where the preset straight line is any vertical straight line; calculating the absolute value of the difference between each shortest distance and the average shortest distance to obtain the offset distance of each center point of the slider area, and taking the average value of the offset distances as the offset in the vertical direction of the movement trajectory.

[0042] In one embodiment, during a stamping stroke, after obtaining the mean value of the stamping parallelism sequence, the variance of the non-zero pressure values in the pressure value sequence, and the offset of the center point of the slider area from the vertical direction in the movement trajectory, the stamping accuracy of this stamping stroke can be quantified. The stamping accuracy satisfies the relationship: , is the mean value of the stamping parallelism sequence, is the variance of the non-zero pressure values in the pressure value sequence, is the offset of the center point of the slider area from the vertical direction in the movement trajectory.

[0043] Among them, when the mean value of the stamping parallelism sequence is 1, the variance of the non-zero pressure values in the pressure value sequence is 0, and the offset of the center point of the slider area from the vertical direction in the movement trajectory is 0, the stamping accuracy reaches the maximum value, and the maximum value of the stamping accuracy is 1.

[0044] In this way, according to the stamping parallelism sequence, pressure value sequence, and the movement trajectory of the center point of the slider area during the stamping stroke, the stamping accuracy during the stamping stroke is accurately quantified, and the stamping accuracy of each stamping stroke is obtained. For any position point in the forging, the average value of the stamping accuracies of all stamping strokes at this position point is used as the forging quality of the position point.

[0045] S102. Construct the stamping area for each position point. The size of the stamping area is related to the size of the slider and the average offset of multiple stamping strokes at the position point.

[0046] In one embodiment, during the process of forging a forging, after multiple stamping strokes at an arbitrary position point are collected, the stamping area of this position point can be constructed. The stamping area of a position point is the superposition result of multiple stamping strokes at this position point and is the influence area of multiple stamping strokes at this position point.

[0047] For the position point in terms of, the larger the size of the slider, the larger the contact area between the slider and the forging, and the stamping area of the position point is larger; the offset of the stamping stroke can reflect the shaking degree of the slider during the stamping stroke. When the average offset of each stamping stroke at the position point is equal to 0, the slider can accurately land on the position point . At this time, the stamping area of the position point is equal to the size of the slider; when the average offset of each stamping stroke at the position point is relatively large, it indicates that the slider has a large shaking, and the position where the slider lands on the surface of the forging will deviate from the position point . At this time, the stamping area of the position point will be larger than the size of the slider.

[0048] Specifically, the stamping area of the position point is a rectangular area centered on the position point . The width of the rectangular area, the length of the rectangular area, and are the width and length of the slider respectively, is the average offset of each stamping stroke at the position point .

[0049] In this way, the average offset of the stamping stroke at any position point can reflect the shaking degree of the slider at this position point. Furthermore, the stamping area of each position point is accurately located according to the shaking degree, ensuring the accuracy of the weighted forging quality of each subsequent position point and avoiding the missed inspection of defective position points.

[0050] S103. Weighted sum the forging quality of the target stamping area at any position point to obtain the weighted forging quality of each position point. The target stamping area is the stamping area containing the position point.

[0051] In one embodiment, after one forging is completed, the stamping area and forging quality of each position point on the forging can be collected. A position point will appear in multiple stamping areas. All stamping areas containing this position point are regarded as the target stamping area of this position point. Further, weighted sum the forging quality of the target stamping area to obtain the weighted forging quality of this position point.

[0052] Specifically, the weighted forging quality of the position point

[0053] is: is the number of target stamping areas, is the forging quality of the target stamping area , is the Euclidean distance between the position point and the center point of the target stamping area .

[0054] In this way, the stamping stroke at a position point on the surface of the forging will also affect other position points within the stamping area of this position point. Obtain all stamping areas containing the position point . The forging quality of the corresponding position points in these stamping areas will all affect the position point . Therefore, when monitoring the quality of the position point in the forging, comprehensively consider the forging quality of all stamping areas containing the position point to achieve accurate monitoring of the true forging quality of the position point in the forging and accurately identify the defects of the position point in the forging.

[0055] S104. In response to the weighted forging quality of any position point being less than the preset quality value, issue a warning.

[0056] In one embodiment, when the weighted forging quality of any position point is less than the preset quality value, mark this position point as a defective point, indicating that there is a defect in the forging. At this time, issue a warning to remind the maintenance personnel to repair the forging machine in time to avoid subsequent unqualified forging quality. Among them, the preset quality value is 0.6.

[0057] It should be noted that after obtaining the weighted forging quality of each position point, mark the position points with weighted forging quality less than the preset quality value as defective points, and the forging defects of the forging can be obtained.

[0058] The technical principle and implementation details of an intelligent monitoring method for a forging production line of the present application are introduced through specific embodiments above. During the forging process of forgings, multiple stamping strokes need to be performed on each position point on the surface of the forging. The stamping accuracy of the stamping stroke directly affects the forging quality of each position point. Therefore, monitor the stamping accuracy of multiple stamping strokes at a position point during the forging process of the forging, and use the average value of the stamping accuracy as the forging quality of this position point; when stamping at the position point during the process, the impact force generated by the slider will not only act on the position point but also act on the surrounding area of the position point causing deformation of other position points in the surrounding area. Therefore, construct the stamping area of each position point according to the slider size and the average offset of multiple stamping strokes at the position point. This stamping area covers other position points that can be affected by the stamping stroke at the corresponding position point; a position point will appear in multiple stamping areas at the same time. Denote the multiple stamping areas where the position point is located as the target stamping area of this position point, and perform a weighted sum of the forging quality of the target stamping area to obtain the weighted forging quality of each position point; the weighted forging quality can truly and accurately reflect the true forging quality of each position point. When the weighted forging quality of any position point is less than the preset quality value, it indicates that this position point is a defective point, and an alarm is issued in a timely manner to improve the accuracy of the production quality monitoring results of the forging production line.

[0059] According to the second aspect of the present application, the present application also provides an intelligent monitoring system for a forging production line. Figure 3 is a structural block diagram of an intelligent monitoring system for a forging production line according to an embodiment of the present application. As shown in Figure 3 the figure, the system 50 includes a processor and a memory. The memory stores computer program instructions, and when the computer program instructions are executed by the processor, they implement an intelligent monitoring method for a forging production line according to the first aspect of the present application. The system also includes other components well-known to those skilled in the art such as a communication bus and a communication interface. Their settings and functions are known in the art, so they will not be described in detail here.

[0060] It should be noted that for those of ordinary skill in the art, without departing from the concept of the present application, several deformations and improvements can still be made, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application should be subject to the appended claims.

Claims

1. An intelligent monitoring method for a forging production line, characterized in that, The monitoring method includes: Monitoring the stamping accuracy of multiple stamping strokes at any position point during the forging process of the forging, and taking the average value of the stamping accuracy as the forging quality of the position point, including: obtaining the slider area and the workbench area of each forging image during a single stamping stroke, and taking the slope consistency between the lower edge of the slider area and the upper edge of the workbench area as the stamping parallelism, where the slider is connected to a pressure sensor; collecting the stamping parallelism sequence, the pressure value sequence, and the movement trajectory of the center point of the slider area from the start of the stamping stroke until the pressure sensor reaches the maximum value; calculating the stamping accuracy of the stamping stroke, where the stamping accuracy is positively correlated with the average value of the stamping parallelism sequence, negatively correlated with the variance of the non-zero pressure values in the pressure value sequence, and negatively correlated with the offset of the movement trajectory of the center point of the slider area from the vertical direction; Constructing the stamping area of each position point, where the size of the stamping area is related to the size of the slider and the average offset of multiple stamping strokes at the position point; Weighted summing the forging quality of the target stamping area at any position point to obtain the weighted forging quality of each position point, where the target stamping area is the stamping area containing the position point; Issuing an alarm in response to the weighted forging quality of any position point being less than a preset quality value; The calculation method of the stamping parallelism includes: Dividing the slider area into multiple slider sub-areas along the vertical direction, and performing linear fitting on the lowest points of each slider sub-area to obtain the slope of the lower edge of the slider area; Dividing the workbench area into multiple workbench sub-areas along the vertical direction, and performing linear fitting on the highest points of each workbench sub-area to obtain the slope of the upper edge of the workbench area; The stamping parallelism is negatively correlated with the absolute value of the difference between the lower edge slope and the upper edge slope; Forging image The stamping parallelism is as follows: , and are respectively the lower edge slope and the upper edge slope in the forging image .

2. The intelligent monitoring method of a forging production line according to claim 1, characterized in that, The method for obtaining the offset of the movement trajectory of the center point of the slider area from the vertical direction includes: Calculating the shortest distance from each center point of the slider area on the movement trajectory to a preset straight line, where the preset straight line is any vertical straight line; Calculating the absolute value of the difference between each shortest distance and the average shortest distance to obtain the offset distance of each center point of the slider area, and taking the average value of the offset distances as the offset in the vertical direction of the movement trajectory.

3. The intelligent monitoring method of a forging production line according to claim 1, characterized in that, The stamping accuracy Satisfies the relation: , is the mean of the stamping parallelism sequence, is the variance of the non-zero pressure values in the pressure value sequence, is the offset of the movement trajectory of the center point of the slider area from the vertical direction.

4. The intelligent monitoring method for a forging production line according to claim 1, characterized in that Position point The stamping area of is a rectangular area centered on the position point The width of the rectangular area is , and are the width and length of the slider respectively, is the average offset of each stamping stroke of the position point .

5. The intelligent monitoring method for a forging production line according to claim 1, wherein Position point weighted forging quality is as follows: , is the number of target stamping areas, is the forging quality of the target stamping area , is the position point and the Euclidean distance from the center point of the target stamping area .

6. The intelligent monitoring method of a forging production line according to claim 1, characterized in that, Marking the position points with weighted forging quality less than the preset quality value as defect points to obtain the forging defects of the forging.

7. The intelligent monitoring method for a forging production line according to claim 1, wherein, Using a semantic segmentation network to obtain the slider area and the workbench area of each forging image during a single stamping stroke.

8. An intelligent monitoring system for a forging production line, characterized in that, Including a processor and a memory, where the memory stores computer program instructions, and when the computer program instructions are executed by the processor, an intelligent monitoring method for a forging production line according to any one of claims 1 to 7 is implemented.

Images