A method and device for detecting faults in the collection of a bar in a downstream collection

Abstract

The application discloses a kind of bar post collection fault detection method and device, the method includes: the image on conveying roller way is collected, the bar image of each bar in bar group is extracted from current frame image;From the bar image, the bar end surface image of all bars is extracted, and the bar end surface image of all bars is converted to roller way surface coordinate system;When conveying roller way is in running state, the bar group is detected based on the bar end surface movement of all bars, including the detection of skidding failure and tailing failure;Based on visual detection, the accurate positioning of single bar is realized, and the skidding failure and tailing failure of bar group are realized, the fault type of this link is maximized to cover.

Description

Technical Field

[0001] This invention belongs to the field of hot rolling technology, and more specifically, this invention relates to a fault detection method and device for downstream collection of bar stock. Background Technology

[0002] In the hot rolling production process of bar stock, after the bars exit the cooling bed, they need to be transported to the cold shearing process via conveyor rollers for length cutting. This step is the core process for subsequent bar stock collection. During roller conveyor transport, two types of core faults are prone to occur: First, slippage, where the roller conveyor motor drives the rollers to run idle, but the bar stock group does not move synchronously with the rollers, causing the bars to fail to reach the cold shearing station on time, resulting in production rhythm disruption; Second, tailing, where the bars at the edges of the bar stock group become stuck due to loose contact with foreign objects such as roller conveyor baffles, resulting in the middle bars arriving while the edge bars remain stuck. This can easily lead to bar bending, roller conveyor blockage, and even serious production accidents such as damage to the cold shearing equipment and production line shutdown.

[0003] Currently, fault detection in this stage generally relies on manual monitoring on-site, requiring on-site workers to conduct inspections throughout the entire process. This places extremely high demands on the workers' concentration, resulting in high labor intensity, slow response time, and a high rate of missed or false detections, making it unsuitable for the high-speed, continuous production pace of bar stock. Summary of the Invention

[0004] This invention provides a fault detection method for downstream collection of bar stock, aiming to solve at least one of the above-mentioned problems.

[0005] This invention is implemented as follows: a fault detection method for downstream collection of bar stock, the method comprising:

[0006] (1) Acquire images on the conveyor rollers and extract the bar images of each bar in the bar group from the current frame image;

[0007] (2) Extract the end face images of all bars from the bar images and convert the end face images of all bars to the roller surface coordinate system;

[0008] (3) When the conveyor roller is in operation, the fault detection of the bar group is performed based on the end face movement of all bars, including the detection of slippage fault and tailing fault.

[0009] In this coordinate system, the y-axis is defined by the running direction of the conveyor rollers.

[0010] Furthermore, the method for detecting slippage faults in bar stock is as follows:

[0011] The displacement of the bar assembly is detected within the first T seconds. If the displacement is less than the set displacement value C, it is determined that the bar assembly is currently experiencing a slippage fault. The set displacement value is determined based on the set operating speed V of the conveyor roller and the time T.

[0012] Furthermore, the process for determining the displacement of the bar assembly within the first T seconds is as follows:

[0013] Read the y-values ​​of all bar end faces in the roller table coordinate system and calculate the average y-values. Use the average y-values ​​as the y-values ​​of the bar group in the current frame image.

[0014] Calculate the difference between the y-value of the bar group in the current image frame and the y-value in the image frame T seconds ago, and use this difference as the displacement of the bar group in the previous T seconds.

[0015] Furthermore, the detection method for tailing failure of bar stock is as follows:

[0016] Read the y-values ​​of all bar end faces in the coordinate system of the roller conveyor surface, determine the maximum and minimum y-values, calculate the absolute value of the difference between the maximum and minimum values, and if the absolute value of the difference is greater than the set difference threshold Thr, then the bar group is identified as having a tailing fault.

[0017] Furthermore, the set displacement value C = V × T × K, where K represents the proportional coefficient, with a value range of 0.7 to 0.9, V represents the set operating speed of the conveyor rollers, and T represents the operating time of the conveyor rollers.

[0018] Furthermore, the current acquisition time and the y-value of the bar group in the current frame image are stored in a FIFO queue. The FIFO queue only saves the acquisition time of the bar group in the first T seconds and its corresponding y-value.

[0019] Furthermore, the pixel coordinates of the bar end face in the image coordinate system are read, and the pixel coordinates of the bar end face in the image coordinate system are transformed to the camera coordinate system based on the intrinsic parameter matrix of the industrial camera. The coordinates of the feature points in the camera coordinate system are transformed to the roller surface coordinate system based on the extrinsic parameter matrix of the industrial camera.

[0020] This invention is implemented as follows: a fault detection device for downstream collection of bar stock, the device comprising:

[0021] The system includes an image acquisition unit, a processing unit connected to the image acquisition unit, and a prompting unit connected to the processing unit. The image acquisition unit is used to acquire images on the conveyor roller in real time and send them to the processing unit. The processing unit detects whether there is a slippage or tailing fault in the bar group based on the fault detection method collected in the downstream of the bar. If so, it issues a reminder through the prompting unit.

[0022] Furthermore, the image acquisition unit includes at least one industrial camera, and the field of view of the image acquisition unit covers the entire conveyor roller conveyor.

[0023] Based on visual inspection, precise positioning of individual bars is achieved, while also detecting slippage and tailing faults in bar groups, covering the maximum range of fault types in this process. In addition, a non-contact visual inspection solution is adopted, and components that come into direct contact with the roller conveyor and bars can adapt to the harsh environment of high temperature, high dust, and high impact in the steel rolling mill, eliminating the need for manual inspection and significantly reducing the labor intensity and labor costs of on-site workers. At the same time, the system parameters can be flexibly adjusted according to the bar specifications and roller conveyor speed, adapting to bar production lines of different specifications. Attached Figure Description

[0024] Figure 1 A flowchart of a fault detection method for downstream bar material collection provided in an embodiment of the present invention;

[0025] Figure 2 This is a schematic diagram of the structure of the fault detection device for downstream collection of bar stock provided in an embodiment of the present invention. Detailed Implementation

[0026] The specific embodiments of the present invention will be further described in detail below with reference to the accompanying drawings, so as to help those skilled in the art to have a more complete, accurate and in-depth understanding of the inventive concept and technical solution of the present invention.

[0027] Figure 1 The flowchart of the fault detection method for downstream bar stock collection provided in this embodiment of the invention is as follows:

[0028] (1) Acquire images on the conveyor rollers and extract the bar images of each bar in the bar group from the current frame image;

[0029] In this embodiment of the invention, the conveyor roller is located between the cooling bed and the cold shear. The bar material coming out of the cooling bed is transported to the cold shear via the conveyor roller. At least one industrial camera is arranged above the conveyor roller. The field of view of the industrial camera covers the entire output roller, and the spatial resolution of the industrial camera is not less than 5mm / pixel. The industrial camera acquires images of the conveyor roller in real time, including the bar material. The bar material image is extracted from the current frame image.

[0030] The current frame image is input into the trained Mask R-CNN network, which extracts the image of each rod in the current frame image.

[0031] (2) Extract the end face images of all bars from the bar images and convert the end face images of all bars to the roller surface coordinate system;

[0032] After extracting the bar image from the image, the pixel coordinates of the bar end face in the image coordinate system are read. Based on the intrinsic parameter matrix of the industrial camera, the pixel coordinates of the bar end face in the image coordinate system are transformed to the camera coordinate system. Based on the extrinsic parameter matrix of the industrial camera, the coordinates of the feature points in the camera coordinate system are transformed to the roller surface coordinate system. The roller surface coordinate system, also known as the world coordinate system, is established with the running direction (length) of the conveyor roller as the y-axis and the lateral direction as the x-axis.

[0033] This invention uses a chessboard calibration algorithm to calculate the intrinsic parameter matrix A of each industrial camera. Then, a calibration board is placed on the roller conveyor surface. Based on the intrinsic parameter matrix A and the world coordinates of the calibration board, the extrinsic parameter matrix B of the camera with the roller conveyor surface as a reference is calculated, thus completing the conversion from pixel coordinates to roller conveyor surface coordinates.

[0034] (3) When the conveyor roller is in operation, the fault detection of the bar group is performed based on the end face movement of all bars, including the detection of slippage fault and tailing fault.

[0035] In this embodiment of the invention, the current operating status signal of the conveyor roller is read. When the conveyor roller is currently in a stopped state, fault detection is not started. After receiving the roller running signal, fault detection is started, and the slippage and tailing faults of the bar group are detected based on the movement of the bar end face.

[0036] In this embodiment of the invention, the method for detecting slippage faults in bar stock is as follows:

[0037] The displacement of the bar stock assembly within the first T seconds is detected. If the displacement is less than the set displacement value C, it is determined that the bar stock assembly currently has a slippage fault. The set displacement value is determined based on the set operating speed V of the conveyor roller and the time T. The set displacement value C = V × T × K, where K represents the proportional coefficient, and the value ranges from 0.7 to 0.9. When the bar stock assembly has a slippage fault, the displacement of the bar stock assembly within the first T seconds will be less than the set displacement value.

[0038] In this embodiment of the invention, the process for determining the displacement of the bar assembly within the first T seconds is as follows:

[0039] Read the y-values ​​of all bar end faces in the roller conveyor coordinate system and calculate the average y-value. Use the average y-value as the y-value of the bar group in the current frame image. Store the current acquisition time and the y-value of the bar group in the current frame image. Calculate the difference between the y-value of the bar group in the current image frame and the y-value in the image frame T seconds ago. Use this difference as the displacement of the bar group in the previous T seconds. Store the current acquisition time and the y-value of the bar group in the current frame image into a FIFO queue. The FIFO queue follows the first-in, first-out principle, and only stores the acquisition time and corresponding y-value of the bar group in the previous T seconds.

[0040] In this embodiment of the invention, the method for detecting tailing faults in bar stock is as follows:

[0041] Read the y-values ​​of all bar end faces in the coordinate system of the roller conveyor surface, determine the maximum and minimum y-values, calculate the absolute value of the difference between the maximum and minimum values, and if the absolute value of the difference is greater than the set difference threshold Thr, then the bar group is identified as having a tailing fault. The difference threshold Thr ranges from 100mm to 300mm and can be adjusted according to the bar specifications and production line process parameters.

[0042] (4) When a tailing fault or slippage fault is detected in the bar stock assembly, fault handling shall be performed, including alarm mode and interlock mode.

[0043] In alarm mode, while issuing a fault alert, the system saves video data of the time of the fault and N seconds before the fault occurred, and writes the fault information into the log to complete full traceability of fault data.

[0044] In the interlocking mode, a fault alert is issued, video data of the fault occurrence and N seconds prior to the fault is saved, and the fault information is written to the log. At the same time, an interlocking control signal is sent to the PLC system to trigger the PLC to execute the preset fault handling logic, including stopping the roller rotation, suspending the bundling plan of the bar group, and pausing the cold shearing process.

[0045] Figure 2 This is a schematic diagram of a fault detection device for downstream bar stock collection provided in an embodiment of the present invention. For ease of explanation, only the parts relevant to the embodiment of the present invention are shown. The device includes:

[0046] The system includes an image acquisition unit, a processing unit connected to the image acquisition unit, and a prompting unit connected to the processing unit. The image acquisition unit is used to acquire images on the conveyor roller in real time and send them to the processing unit. The processing unit detects whether there is a slippage or tailing fault in the bar group based on the fault detection method collected in the downstream of the bar. If so, it issues a reminder through the prompting unit.

[0047] In this embodiment of the invention, the image acquisition unit includes at least one industrial camera, a matching bracket, and a terminal box. The industrial camera is positioned above the conveyor rollers, providing a complete field of view covering the entire range of the rollers. The spatial resolution of a single camera is no less than 5mm / pixel, and it is used to acquire video images of the bar stock conveyor in real time.

[0048] Based on the received current frame image, the processing unit sequentially completes the extraction of the bar end face, the y-value of the bar group, the confirmation of the displacement of the bar group in the first T seconds, and the determination of slippage and tailing faults.

[0049] The fault detection method for downstream collection of bar stock provided by this invention has the following beneficial technical effects:

[0050] Based on visual inspection, precise positioning of individual bars is achieved, while also detecting slippage and tailing faults in bar groups, covering the maximum range of fault types in this process. In addition, a non-contact visual inspection solution is adopted, and components that come into direct contact with the roller conveyor and bars can adapt to the harsh environment of high temperature, high dust, and high impact in the steel rolling mill, eliminating the need for manual inspection and significantly reducing the labor intensity and labor costs of on-site workers. At the same time, the system parameters can be flexibly adjusted according to the bar specifications and roller conveyor speed, adapting to bar production lines of different specifications.

[0051] The present invention has been described by way of example. Obviously, the specific implementation of the present invention is not limited to the above-described manner. Any non-substantial improvements made using the inventive concept and technical solution of the present invention, or the direct application of the inventive concept and technical solution of the present invention to other occasions without modification, are all within the protection scope of the present invention.

Claims

1. A fault detection method for downstream bar stock collection, characterized in that, The method includes: (1) Acquire images on the conveyor rollers and extract the bar images of each bar in the bar group from the current frame image; (2) Extract the end face images of all bars from the bar images and convert the end face images of all bars to the roller surface coordinate system; (3) When the conveyor roller is in operation, the fault detection of the bar group is performed based on the end face movement of all bars, including the detection of slippage fault and tailing fault. In this coordinate system, the y-axis is defined by the running direction of the conveyor rollers.

2. The fault detection method for downstream bar stock collection as described in claim 1, characterized in that, The method for detecting slippage faults in bar stock is as follows: The displacement of the bar assembly is detected within the first T seconds. If the displacement is less than the set displacement value C, it is determined that the bar assembly is currently experiencing a slippage fault. The set displacement value is determined based on the set operating speed V of the conveyor roller and the time T.

3. The fault detection method for downstream collection of bar stock as described in claim 2, characterized in that, The process for determining the displacement of the bar assembly within the first T seconds is as follows: Read the y-values ​​of all bar end faces in the roller table coordinate system and calculate the average y-values. Use the average y-values ​​as the y-values ​​of the bar group in the current frame image. Calculate the difference between the y-value of the bar group in the current image frame and the y-value in the image frame T seconds ago, and use this difference as the displacement of the bar group in the previous T seconds.

4. The fault detection method for downstream collection of bar stock as described in claim 1, characterized in that, The detection method for trailing faults in bar stock is as follows: Read the y-values ​​of all bar end faces in the coordinate system of the roller conveyor surface, determine the maximum and minimum y-values, calculate the absolute value of the difference between the maximum and minimum values, and if the absolute value of the difference is greater than the set difference threshold Thr, then the bar group is identified as having a tailing fault.

5. The fault detection method for downstream collection of bar stock as described in claim 1, characterized in that, The set displacement value C = V × T × K, where K represents the proportional coefficient, with a value range of 0.7 to 0.9, V represents the set operating speed of the conveyor rollers, and T represents the operating time of the conveyor rollers.

6. The fault detection method for downstream collection of bar stock as described in claim 1, characterized in that, Store the current acquisition time and the y-value of the bar group in the current frame image into a FIFO queue. The FIFO queue only saves the acquisition time of the bar group in the first T seconds and its corresponding y-value.

7. The fault detection method for downstream bar stock collection as described in claim 1, characterized in that, Read the pixel coordinates of the bar end face in the image coordinate system, transform the pixel coordinates of the bar end face in the image coordinate system to the camera coordinate system based on the intrinsic parameter matrix of the industrial camera, and transform the coordinates of the feature points in the camera coordinate system to the roller surface coordinate system based on the extrinsic parameter matrix of the industrial camera.

8. A fault detection device for downstream collection of bar stock, characterized in that, The device includes: The image acquisition unit, the processing unit connected to the image acquisition unit, and the prompting unit connected to the processing unit are provided. The image acquisition unit is used to acquire images on the conveyor roller in real time and send them to the processing unit. The processing unit detects whether there is a slippage fault or a tailing fault in the bar stock group based on the fault detection method for bar stock collection in any one of claims 1 to 7. If there is a fault, the prompting unit issues a reminder.

9. The fault detection device for downstream collection of bar stock as described in claim 8, characterized in that, The image acquisition unit includes at least one industrial camera, and its field of view covers the entire conveyor roller conveyor.

Images