Method and system for automatically detecting opening degree of sliding nozzle of continuous casting ladle

Abstract

The invention discloses a method and system for automatically detecting the opening degree of a continuous casting ladle sliding nozzle, and belongs to the field of automatic control. Comprising the following steps: acquiring a follow-up target image connected with a ladle lower sliding plate water gap in real time, and forming digital signal output; transmitting the acquired follow-up target digital image signal to an image processing unit; the obtained follow-up target image connected with the ladle lower sliding plate nozzle is analyzed and processed, and the detection quantity reflecting the opening degree of the ladle lower sliding plate nozzle in real time is obtained; and the steel ladle sliding nozzle opening detection quantity is output to a user system through the signal output unit and is used for displaying, alarming and controlling. Through an image recognition mode, a follow-up target image connected with a steel ladle lower sliding plate nozzle is directly collected and analyzed, and the detection quantity reflecting the opening degree of the steel ladle sliding nozzle is obtained in real time; the automatic optimization control of the opening degree of the sliding nozzle before final casting of the continuous casting ladle is realized, the method can also be used for automatic alarm and interlocking control of the opening degree abnormity of the sliding nozzle of the continuous casting ladle, and the monitoring precision and the automation degree of the production process are improved.

Description

Technical Field

[0001] This invention belongs to the field of automatic control, and in particular relates to an automatic detection method and system for the opening of the sliding gate nozzle of a continuous casting ladle. Background Art

[0002] In the continuous casting production process, molten steel flows from the sliding gate at the bottom of the ladle through the long gate into the tundish for continuous casting.

[0003] The ladle sliding gate consists of an upper sliding gate and a lower sliding gate. The upper sliding gate is a static gate that connects to the gate at the bottom of the ladle. The lower sliding gate is a dynamic gate that is driven to open and close by a hydraulic cylinder. The long gate is connected to the lower sliding gate through a clamping device to establish a closed flow channel for molten steel.

[0004] The opening of the ladle sliding gate is determined by the size of the through hole between the upper and lower sliding gates. The larger the gate opening, the greater the amount of molten steel flowing out of the ladle. The smaller the gate opening, the smaller the amount of molten steel flowing out of the ladle. When the gate opening becomes too small, the molten steel will no longer flow out of the ladle, thus forming the final pour.

[0005] The opening of the ladle sliding gate is the most important means of controlling the amount of molten steel flowing out of the ladle. It directly affects the height of the molten steel level in the tundish, as well as the stability of continuous casting production and quality.

[0006] Traditional methods for detecting the opening of the sliding gate mainly rely on displacement sensors inside the hydraulic cylinder connected to the lower sliding gate. However, due to the harsh environment around the continuous casting ladle pouring location, including high-temperature molten metal and flames, the displacement sensors have a very short lifespan and are easily damaged. Therefore, in actual production, the size of the ladle sliding gate opening is still largely determined by manual experience. This not only results in large errors and difficulty in achieving precise control, affecting the control of excess steel in the ladle and the molten steel level in the tundish, but also, with slight negligence, the gate may close completely, leading to "final pouring" (short for "termination of casting"), which has a very adverse impact on the stability of continuous casting production and quality.

[0007] The invention patent CN104999043B, authorized on April 26, 2017, discloses "An online measuring device and method for the opening of the sliding gate nozzle in continuous casting ladles." The device includes: a pressure sensor installed on the ladle turret to measure the current signal of the current total weight of the ladle; a signal processing unit connected to the pressure sensor to process the current signal into the total weight value of the ladle; a process signal interface unit that converts the currently cast steel grade signal into a steel grade code and outputs it along with the net weight of the currently cast ladle; and a calculation controller connected to both the signal processing unit and the process signal interface unit to read the current total weight value of the ladle, the steel grade code, and the net weight signal of the ladle, and calculates the opening of the sliding gate nozzle. This technical solution measures the gradient and rate of change in the weight of molten steel in the ladle, combined with parameters such as the static pressure and viscosity of the molten steel at the outlet, to calculate the opening value of the sliding gate nozzle. The device is not easily damaged, ensuring the smooth progress of the continuous casting process.

[0008] However, this technical solution is based on the measurement of the gradient and rate of the change trend of molten steel weight in the ladle. For the opening value of the sliding gate, it is an indirect calculation and measurement, which is greatly affected by parameters such as weight, steel grade and molten steel viscosity, resulting in low accuracy and low real-time performance.

[0009] How to solve the problems of easy damage and short lifespan of current displacement sensors, and large errors and poor reliability of other detection methods, is a technical problem that needs to be solved in actual field work. Summary of the Invention

[0010] The technical problem to be solved by this invention is to provide an automatic detection method and system for the opening degree of the sliding gate nozzle in continuous casting ladles. It adopts a non-contact detection mode using image recognition, directly acquiring and analyzing images of a moving target connected to the sliding gate nozzle at the bottom of the ladle, and obtaining the detected quantity reflecting the opening degree of the sliding gate nozzle in real time. This achieves real-time and automatic detection of the opening degree of the sliding gate nozzle in continuous casting ladles, effectively solving the problems of easily damaged and short-life sensors, large errors, and poor reliability in current measurement methods. It can also be used for automatic alarm and interlock control of abnormal sliding gate nozzle opening in continuous casting ladles, improving the monitoring accuracy and automation level in the production process.

[0011] The technical solution of the present invention is: to provide an automatic detection method for the opening degree of the sliding gate nozzle in a continuous casting ladle, characterized by including the following steps:

[0012] 1) The image acquisition unit acquires real-time images of the servo target connected to the slide gate nozzle of the ladle and generates digital signal output;

[0013] 2) The acquired digital image signal of the servo target is transmitted to the image processing unit;

[0014] 3) The image processing unit automatically analyzes and processes the obtained image of the moving target connected to the ladle slide gate nozzle to obtain the detection quantity that reflects the opening degree of the ladle slide gate nozzle in real time;

[0015] 4) The detected opening value of the ladle sliding gate is output to the user system through the signal output unit for display, alarm and control.

[0016] The automatic detection method for the opening of the sliding gate of the continuous casting ladle described in this invention directly acquires and analyzes the image of the moving target connected to the sliding gate of the ladle through a non-contact detection mode of image recognition, and obtains the detection quantity reflecting the opening of the sliding gate of the ladle in real time, thereby realizing the automatic detection of the opening of the sliding gate of the continuous casting ladle.

[0017] Specifically, the follow-up target connected to the ladle's lower slide gate nozzle includes a long nozzle that is tightly connected to the ladle's lower slide gate nozzle.

[0018] Furthermore, in step 3), the image processing unit automatically analyzes and processes the obtained image of the moving target connected to the ladle's sliding gate nozzle to obtain a detection quantity that reflects the real-time opening degree of the ladle's sliding gate nozzle, specifically including:

[0019] 3.1) Identify the detection target from the acquired image of the moving target of the sliding gate nozzle under the connected steel ladle;

[0020] 3.2) Process the acquired image containing the target to obtain the pixel coordinates reflecting the location of the target;

[0021] 3.3) Convert the pixel coordinates of the target into a detection quantity that reflects the opening of the ladle sliding gate.

[0022] Furthermore, step 3.1) of identifying the detection target from the acquired image of the moving target of the connecting ladle slide gate nozzle specifically includes:

[0023] 3.1.1) Select a portion of the image samples from the collected images of the moving target at the bottom slide gate of the connected steel ladle as template images for target detection and recognition;

[0024] 3.1.2) Traverse and search similar sub-images from the real-time acquired images according to the size of the template image to identify the target to be detected;

[0025] 3.1.3) If the target exists, distinguish between a cold water inlet and a hot water inlet based on the brightness of the target image; if the target does not exist, the detection ends.

[0026] Furthermore, in step 3.2), the acquired image containing the detected target is processed to obtain the pixel coordinates reflecting the location of the detected target, specifically including:

[0027] 3.2.1) Extract the effective detection selection area from the acquired image where the target is to be detected;

[0028] 3.2.2) Perform preliminary denoising on the image of the effective detection area to initially filter out the influence of Martian smoke on the image;

[0029] 3.2.3) Binarize the image of the effective detection selection area to obtain the target contour;

[0030] 3.2.4) Perform secondary denoising on the target contour image to further filter out the influence of external burrs, small sparks, and other bright light on the contour;

[0031] 3.2.5) Based on the position of the target contour after the secondary denoising process, generate the corresponding target contour bounding rectangle and determine its center coordinate point; use the horizontal pixel coordinates of the center coordinate point to represent the pixel coordinates P of the detected target.

[0032] Specifically, the effective detection area is pre-selected according to the displacement change range of the detection target;

[0033] The preliminary denoising process for the effective detection selection area image is performed using a Gaussian filtering algorithm.

[0034] The binarization processing of the effective detection area image includes selecting different binarization thresholds based on whether the water inlet is cold or hot.

[0035] The target contour image undergoes secondary denoising processing using erosion and dilation algorithms.

[0036] Furthermore, in step 3.3), the pixel coordinates of the detected target are converted into a detection quantity that reflects the opening degree of the ladle sliding gate, specifically including:

[0037] 3.3.1) After the ladle reaches the casting position and the long nozzle is installed, zero-position calibration is performed with the sliding nozzle fully closed. The target pixel coordinates P at this time are extracted as the zero-position pixel coordinates P0 of the current ladle.

[0038] 3.3.2) After the ladle is opened for casting, the fully open position calibration is performed with the sliding gate fully open, and the target pixel coordinate P at this time is extracted as the fully open position pixel coordinate P1 of the current ladle.

[0039] 3.3.3) Based on the real-time change of the target pixel coordinate P of the ladle as the sliding gate opening changes during casting, the ladle sliding gate opening detection quantity SG is determined:

[0040] The measured value of the sliding gate opening of the ladle is SG = [(P-P0) / (P1-P0)] × 100%;

[0041] Where P is the target pixel coordinate, P1 is the fully open pixel coordinate of the ladle, and P0 is the zero pixel coordinate of the ladle.

[0042] The technical solution of the present invention also provides an automatic detection system for the opening degree of the sliding gate nozzle in continuous casting ladle, characterized in that:

[0043] It consists of an image acquisition unit, a signal transmission unit, an image processing unit, and a signal output unit;

[0044] The image acquisition unit is used to acquire images of the moving target connected to the sprue nozzle of the ladle in real time and generate digital signal output.

[0045] The signal transmission unit is connected to the image acquisition unit and is used to transmit the digital image signal emitted by the image acquisition unit to the image processing unit.

[0046] The image processing unit is connected to the signal transmission unit and is used to receive image information from the image acquisition unit in real time, automatically analyze and process the target image that moves with the ladle slide gate nozzle, and form a detection quantity reflecting the opening degree of the ladle slide gate nozzle.

[0047] The signal output unit is used to output the ladle sliding gate opening detection value generated by the image processing unit to the user system through the signal output interface for display, alarm and control;

[0048] The aforementioned automatic detection system for the opening of the sliding gate of the continuous casting ladle directly acquires and analyzes images of the moving target connected to the sliding gate of the ladle through a non-contact detection mode of image recognition, thereby obtaining the detection quantity reflecting the opening of the sliding gate of the ladle in real time and realizing the automatic detection of the opening of the sliding gate of the continuous casting ladle.

[0049] Specifically, the image acquisition unit includes an industrial camera, a protective cover, and a pan-tilt bracket;

[0050] The industrial camera is used to acquire real-time images of the moving target connected to the sprue nozzle of the ladle and generate a digital signal output suitable for network transmission.

[0051] The protective cover is used to protect industrial cameras from dust, splashes, and high-temperature environments.

[0052] The pan-tilt bracket is used to easily adjust the angle and position of the industrial camera capturing images.

[0053] Specifically, the signal transmission unit includes an optoelectronic switch, a network cable, and an optical fiber cable.

[0054] Furthermore, the image processing unit is composed of an industrial control computer, which has a central processing unit, a memory, and a network interface.

[0055] Specifically, the image processing unit includes a target recognition module, a target detection module, and an aperture conversion module;

[0056] The target recognition module is used to identify the target to be detected from the acquired image of the moving target of the connecting ladle slide gate nozzle;

[0057] The target detection module is used to process the image of the target to obtain pixel coordinates reflecting the location of the target;

[0058] The opening conversion module is used to convert the pixel coordinates of the detected target into a detection quantity that reflects the opening of the ladle sliding gate.

[0059] Furthermore, the signal output interface adopts an Ethernet communication interface and outputs the ladle sliding nozzle opening detection data to the remote monitoring screen of the user system via the OPC communication protocol. At the same time, it is output to the user controller for abnormal alarm and opening control of the ladle sliding nozzle opening.

[0060] The automatic detection system for the opening of the sliding gate of the continuous casting ladle described in this invention provides real-time feedback of the real-time opening of the sliding gate to the control system. This system is used for the automatic optimization and control of the opening of the sliding gate before the final pouring of the continuous casting ladle, thereby improving the accuracy of the opening control and the control effect of the remaining steel quantity in the continuous casting ladle.

[0061] The automatic detection system for the opening of the sliding gate of the continuous casting ladle, as described in this invention, accurately feeds back the real-time opening of the sliding gate to the control system. This system is used for automatic alarm and interlock control of abnormal opening of the sliding gate in the continuous casting ladle, preventing the risk of abnormal casting interruption accidents caused by the excessively small opening of the sliding gate during casting, and improving the accuracy and automation of production process monitoring.

[0062] Compared with the prior art, the advantages of the present invention are:

[0063] 1. The technical solution of the present invention directly acquires and analyzes the image of the follow-up target connected to the sliding gate nozzle of the ladle through a non-contact detection mode of image recognition, and obtains the detection quantity reflecting the opening degree of the sliding gate nozzle of the ladle in real time; it realizes the automatic detection of the opening degree of the sliding gate nozzle of the continuous casting ladle, and has the advantages of accuracy, remoteness, automation and reliability, which can effectively solve the problems of easy damage and short life of displacement sensors, and large error and poor reliability of other detection methods;

[0064] 2. The technical solution of this invention, by accurately feeding back the real-time opening of the ladle sliding gate to the control system, can not only be used for the automatic optimization control of the opening of the sliding gate before the final pouring of the continuous casting ladle, improving the opening control accuracy and ensuring that the ladle sliding gate can be closed to an optimal opening position before the final pouring, so as to delay the slag discharge time caused by the residual steel vortex, which is beneficial to improving the control effect of the residual steel amount in the continuous casting ladle; it can also be used for automatic alarm and interlock control of abnormal opening of the continuous casting ladle sliding gate, preventing the risk of abnormal pouring accidents caused by the ladle sliding gate opening being too small during casting, and improving the monitoring accuracy and automation level of the production process;

[0065] 3. The technical solution of this invention, when processing the acquired image to obtain the pixel coordinates reflecting the location of the detected target, achieves linear smoothing filtering of the image by calling the GaussianBlur function of the OpenCV library and selecting a 5×5 pixel convolution kernel. The smoothing effect is good, the computational speed is not high, and it helps to reduce the demand on the hardware system. The erosion and dilation algorithms are used to perform secondary denoising processing on the contour region of the long water inlet target, which can effectively eliminate noise and meaningless areas in the image. Attached Figure Description

[0066] Figure 1 This is a block flowchart illustrating the automatic detection method for the sliding gate opening of a continuous casting ladle according to the present invention.

[0067] Figure 2 This is a schematic diagram of the automatic detection system for the sliding gate opening of the continuous casting ladle of the present invention.

[0068] Figure 3 This is a block flowchart illustrating the method for measuring the opening degree of the sliding gate nozzle in continuous casting ladle according to the present invention.

[0069] Figure 4 This is a schematic diagram of the target detection and recognition process of the present invention;

[0070] Figure 5 This is a schematic diagram of template image matching according to the present invention;

[0071] Figure 6 This is a schematic diagram of the process for determining the detection target location of the present invention;

[0072] Figure 7 This is a schematic diagram of the center coordinates of the target contour of the present invention;

[0073] Figure 8 This is a flowchart illustrating the process for determining the opening value of the sliding gate nozzle in a steel ladle according to the present invention.

[0074] In the diagram, 1 is the long sprue, 2 is the lower slide gate sprue, 3 is the upper slide gate sprue, 4 is the ladle, 5 is the tundish, 6 is the hydraulic cylinder, 7 is the pneumatic cylinder, 8 is the robotic arm, 11 is the image acquisition unit, 12 is the signal transmission unit, 13 is the image processing unit, 131 is the target recognition module, 132 is the target detection module, 133 is the opening conversion module, and 14 is the signal output unit. Detailed Implementation

[0075] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0076] like Figure 1 As shown, the first aspect of the technical solution of the present invention provides an automatic detection method for the opening degree of the sliding gate nozzle in a continuous casting ladle, comprising the following steps:

[0077] 1) The image acquisition unit acquires real-time images of the servo target connected to the slide gate nozzle of the ladle and generates digital signal output;

[0078] 2) The acquired digital image signal of the servo target is transmitted to the image processing unit;

[0079] 3) The image processing unit automatically analyzes and processes the obtained image of the moving target connected to the ladle slide gate nozzle to obtain the detection quantity that reflects the opening degree of the ladle slide gate nozzle in real time;

[0080] 4) The detected opening value of the ladle sliding gate is output to the user system through the signal output unit for display, alarm and control.

[0081] Specifically, such as Figure 2 As shown, the image acquisition unit 11 acquires images of the following target connected to the ladle slide gate 2 in real time, wherein the following target is the long gate 1 that is tightly connected to the ladle slide gate 2.

[0082] In this example, the long sprue 1 is connected to the lower slide sprue 2 of the ladle 4 by a mechanical arm 8 driven by cylinder 7. During the opening and closing of the lower slide sprue 2 of the ladle, the long sprue 1 will move synchronously.

[0083] like Figure 3 As shown, the image processing unit automatically analyzes and processes the obtained image of the moving target connected to the ladle's sliding gate nozzle to obtain a detection quantity that reflects the real-time opening degree of the ladle's sliding gate nozzle. This process specifically includes the following steps:

[0084] 3.1) Identify the detection target from the acquired image of the moving target of the sliding gate nozzle under the connected steel ladle;

[0085] 3.2) Process the acquired image containing the target to obtain the pixel coordinates reflecting the location of the target;

[0086] 3.3) Convert the pixel coordinates of the target into a detection quantity that reflects the opening of the ladle sliding gate.

[0087] Furthermore, such as Figure 4 As shown, identifying the target to be detected from the acquired image of the servo target connected to the ladle's lower slide gate nozzle includes:

[0088] 3.1.1) From the acquired images of the moving target at the sprue nozzle of the connecting ladle, select a portion of image samples as template images for target detection and recognition, such as... Figure 5 As shown;

[0089] In this example, the selected image samples are the middle images taken from two different long nozzle photos collected during production, one in a cold state and the other in a hot state.

[0090] 3.1.2) From the real-time acquired images, traverse and search for similar sub-images according to the size of the template image to identify the detection target, such as... Figure 5 As shown;

[0091] In this example, a template matching algorithm is used to search for whether there is a target in the acquired image that is similar to the template image. When the similarity is greater than 85%, it is determined that the target of the long water outlet to be detected exists.

[0092] 3.1.3) If the target exists, distinguish between a cold-state sprue and a hot-state sprue based on the brightness of the target image; if the target does not exist, end the detection process.

[0093] In this example, if the brightness of the target image of the long water inlet is within the threshold range of the cold water inlet, it is determined to be a cold water inlet; if the brightness of the target image of the long water inlet is within the threshold range of the hot water inlet, it is determined to be a hot water inlet.

[0094] In this example, the brightness threshold range of the hot water inlet is between 100 and 120, and the brightness threshold range of the cold water inlet is between 60 and 80.

[0095] like Figure 6 As shown, the process of processing the acquired image containing the target to obtain the pixel coordinates reflecting the target's location specifically includes:

[0096] 3.2.1) Extract the effective detection selection region from the acquired image where the target is to be detected, such as... Figure 7 As shown;

[0097] The effective detection selection area is pre-selected according to the displacement change range of the detection target; the size of the effective detection selection area extracted in this example is 350×140 pixels;

[0098] 3.2.2) Perform preliminary denoising on the image of the effective detection area to initially filter out the influence of Martian smoke on the image;

[0099] This example uses a Gaussian filtering algorithm to perform preliminary denoising on the image of the effective detection selection area.

[0100] Gaussian filtering algorithms use a weighted average of the value of each pixel in an image and the values ​​of its neighboring pixels to achieve linear smoothing. This algorithm can effectively eliminate Gaussian noise in images, i.e., noise that conforms to a normal distribution. The specific implementation steps are as follows: First, select a convolution kernel template of a specified size (5×5 pixels in this example); slide the template across the image, operating on each pixel; for the pixel at the center of the template, multiply it by the gray values ​​of other pixels within the template, and then sum them; the weight is inversely proportional to the distance between the pixel and the center of the template, with closer pixels having higher weights; finally, replace the value of the center pixel with the weighted average gray value of the pixels in the neighborhood determined by the template. These steps are repeated until the entire image has been processed.

[0101] This example implements linear smoothing filtering of images by calling the GaussianBlur function of the OpenCV library and selecting a 5×5 pixel convolution kernel. Its advantages are good smoothing effect and low computational pressure. This can reduce the computational pressure on existing hardware control systems and reduce the requirements for the computing power of hardware systems, which helps to reduce the operation and maintenance costs of the entire control system.

[0102] 3.2.3) Binarize the image of the effective detection selection area to obtain the target contour, such as... Figure 7 As shown;

[0103] The step of binarizing the image of the effective detection area to obtain the target contour includes: selecting different binarization thresholds based on whether the sprue is cold or hot.

[0104] In this example, the threshold parameter for the cold inlet is 80, and the threshold parameter for the hot inlet is 230.

[0105] 3.2.4) Perform secondary denoising on the target contour image to further filter out the influence of external burrs, small sparks, and other bright light on the contour;

[0106] This example uses erosion and dilation algorithms to perform secondary denoising on the target contour region of the long water inlet.

[0107] Image erosion is a morphological processing technique whose main purpose is to alter or simplify the shape of an image by eliminating small, meaningless regions.

[0108] The implementation steps of the erosion algorithm in this example are as follows:

[0109] a) Select a structuring element. In this example, the structuring element is a 3×3 two-dimensional matrix that can be used to erode any 3×3 pixel-sized region in the image.

[0110] b) Slide the structuring element to every position on the image;

[0111] c) After each move, perform an erosion operation on the neighborhood of the current position; the erosion operation is accomplished by finding the minimum value in the neighborhood covered by the structuring element. This minimum value is used as the new position value, thereby changing the shape of the original image.

[0112] Repeat the above steps until the entire image has been processed.

[0113] A key feature of this method is its ability to effectively eliminate noise and meaningless regions in images. For example, if an image contains isolated small dots, these dots may interfere with subsequent analysis or processing. Through erosion, these dots can be "eroded away," resulting in a cleaner and easier-to-analyze image.

[0114] The erosion calculation formula is as follows: Assuming the structuring element is A and the original image is f(x,y), the eroded image g(x,y) can be calculated using the following formula: g(x,y)=min{f(xa,yb),a∈A,b∈A}, where min{} represents the minimum value operation. This means that for each position covered by the structuring element, the minimum value in its neighborhood is selected as the new position value.

[0115] Since erosion eliminates noise and meaningless areas in the image, it also changes the shape of the original image, making the target area smaller. In essence, it causes the image boundary to shrink. Therefore, it is necessary to perform a dilation operation to enlarge the target area to restore the boundary. This can be used to fill holes or defects in the image.

[0116] The dilation operation can be viewed as the inverse of the erosion operation, accomplished by finding the maximum value within the neighborhood covered by the structuring element. This maximum value is used as the new location value, thereby altering the shape of the original image and thus obtaining the region contained within the boundary when the boundary is known.

[0117] 3.2.5) Based on the position of the target contour after secondary denoising, generate the corresponding bounding rectangle of the target contour and determine its center coordinates, such as... Figure 7 As shown, the pixel coordinates P of the detected target are represented by the horizontal pixel coordinates of the center coordinate point.

[0118] like Figure 8 As shown, the step of converting the pixel coordinates of the detected target into a detection quantity that reflects the opening degree of the ladle sliding gate specifically includes:

[0119] 3.3.1) After the ladle reaches the casting position and the long nozzle is installed, zero-position calibration is performed with the sliding gate fully closed. The target pixel coordinates P at this time are extracted as the current zero-position pixel coordinates P0 of the ladle. See [link to relevant documentation]. Figure 7 As shown;

[0120] 3.3.2) After the ladle is opened for casting, the fully open position is calibrated while the sliding gate is fully open. The pixel coordinates P of the detected target at this time are extracted as the pixel coordinates P1 of the current fully open position of the ladle. See [link to relevant documentation]. Figure 7 As shown;

[0121] 3.3.3) Based on the real-time change of the target pixel coordinate P of the ladle as the sliding gate opening changes during casting, the ladle sliding gate opening detection quantity SG is determined: SG=[(P-P0) / (P1-P0)]×100%;

[0122] In this example, the zero-position pixel coordinates of the ladle are P0 = 762, the fully open position pixel coordinates are P1 = 1502, and the target pixel coordinates of the long gate detection after the ladle opens during casting are P = 1405. Therefore, the ladle sliding gate opening detection amount SG = [(1405-762) / (1502-762)] × 100% = 86.89%.

[0123] like Figure 2 As shown, in a second aspect of the technical solution of the present invention, an automatic detection system for the opening degree of the sliding gate nozzle of a continuous casting ladle is provided, comprising an image acquisition unit 11, a signal transmission unit 12, an image processing unit 13, and a signal output unit 14.

[0124] The image acquisition unit 11 includes an industrial camera, a protective cover, and a pan-tilt bracket;

[0125] The industrial camera is a network camera with a high-definition lens, used to acquire real-time images of the moving target connected to the slide gate nozzle of the ladle, and generate a digital signal output suitable for network transmission.

[0126] The protective cover is a water-air composite cooling protective cover, whose outer shell can simultaneously pass water and air, to protect industrial cameras from the effects of dust, splashes and high-temperature environments.

[0127] The pan-tilt bracket adopts a three-dimensional adjustable pan-tilt head, which is used to easily adjust the angle and position of the industrial camera to capture images;

[0128] This example uses an 8-megapixel starlight-level bullet network camera with an Ethernet signal interface;

[0129] The signal transmission unit 14 is connected to the image acquisition unit 11 and is used to transmit the digital image signal emitted by the image acquisition unit 11 to the image processing unit 13.

[0130] The signal transmission unit 12 in this example includes an optoelectronic switch, a network cable, and an optical fiber cable;

[0131] The image processing unit 13 is connected to the signal transmission unit 12 and is used to receive image information sent by the image acquisition unit 11 in real time, and automatically analyze and process the target image that moves with the ladle slide gate 2 to form a detection quantity reflecting the opening degree of the ladle slide gate.

[0132] The image processing unit 13 in this example consists of an industrial control computer, which has a central processing unit, memory, and network interface.

[0133] The image processing unit 13 includes a target recognition module 131, a target detection module 132, and an aperture conversion module 133;

[0134] The target recognition module 131 is used to identify the target to be detected from the acquired image of the moving target of the connecting ladle slide gate nozzle;

[0135] The target detection module 132 is used to process the image of the target to obtain pixel coordinates reflecting the position of the target;

[0136] The opening conversion module 133 is used to convert the pixel coordinates of the detected target into a detection quantity that reflects the opening of the ladle sliding gate.

[0137] The signal output unit 14 is used to output the ladle sliding gate opening detection quantity generated by the image processing unit 13 to the user system through the signal output interface for display, alarm and control.

[0138] In this example, the signal output interface uses an Ethernet communication interface. Through the OPC communication protocol, the detection data of the ladle sliding gate opening is output to the remote monitoring screen of the user system for display. At the same time, it is output to the user controller for abnormal alarm and opening control of the ladle sliding gate opening.

[0139] The technical solution of this invention directly acquires and analyzes the image of the target connected to the sliding gate nozzle of the ladle through a non-contact detection mode of image recognition, and obtains the detection quantity reflecting the opening degree of the sliding gate nozzle of the ladle in real time; thus realizing the automatic detection of the opening degree of the sliding gate nozzle of the continuous casting ladle.

[0140] Furthermore, the technical solution of this invention, by accurately feeding back the real-time opening of the ladle sliding gate to the control system, can not only be used for the automatic optimization control of the opening of the sliding gate before the final pouring of the continuous casting ladle, improving the opening control accuracy and ensuring that the ladle sliding gate can be closed to an optimal opening position before the final pouring, so as to delay the slag discharge time caused by the residual steel vortex, which is beneficial to improving the control effect of the residual steel amount in the continuous casting ladle; it can also be used for automatic alarm and interlock control of abnormal opening of the continuous casting ladle sliding gate, preventing the risk of abnormal pouring accidents caused by the ladle sliding gate opening being too small during casting, and improving the monitoring accuracy and automation level of the production process.

[0141] This invention can be widely used in the fields of process detection and automatic process control in continuous casting production.

Claims

1. An automatic detection method for the opening degree of the sliding gate nozzle in a continuous casting ladle, characterized by comprising the following steps: 1) The image acquisition unit acquires real-time images of the servo target connected to the slide gate nozzle of the ladle and generates digital signal output; 2) The acquired digital image signal of the servo target is transmitted to the image processing unit; 3) The image processing unit automatically analyzes and processes the obtained image of the moving target connected to the ladle slide gate nozzle to obtain the detection quantity that reflects the opening degree of the ladle slide gate nozzle in real time; 4) The detected opening value of the ladle sliding gate is output to the user system through the signal output unit for display, alarm and control.

2. The automatic detection method for the opening degree of the sliding gate nozzle of a continuous casting ladle according to claim 1, characterized in that: The automatic detection method for the opening of the sliding gate of the continuous casting ladle uses a non-contact detection mode based on image recognition to directly collect and analyze the image of the moving target connected to the sliding gate of the ladle, thereby obtaining the detection quantity reflecting the opening of the sliding gate in real time and realizing the automatic detection of the opening of the sliding gate of the continuous casting ladle.

3. The automatic detection method for the opening degree of the sliding gate nozzle in continuous casting ladle according to claim 1, characterized in that: The follower target connected to the ladle slide gate nozzle includes a long nozzle that is tightly connected to the ladle slide gate nozzle.

4. The automatic detection method for the opening degree of the sliding gate nozzle of a continuous casting ladle according to claim 1, characterized in that... In step 3), the image processing unit automatically analyzes and processes the obtained image of the moving target connected to the ladle sliding gate nozzle to obtain a detection quantity that reflects the opening degree of the ladle sliding gate nozzle in real time, specifically including: 3.1) Identify the detection target from the acquired image of the moving target of the sliding gate nozzle under the connected steel ladle; 3.2) Process the acquired image containing the target to obtain the pixel coordinates reflecting the location of the target; 3.3) Convert the pixel coordinates of the target into a detection quantity that reflects the opening of the ladle sliding gate.

5. The automatic detection method for the opening degree of the sliding gate nozzle in continuous casting ladle according to claim 4, characterized in that: Step 3.1) involves identifying the detection target from the acquired image of the moving target of the connecting ladle's lower slide gate nozzle, specifically including: 3.1.1) Select a portion of the image samples from the collected images of the moving target at the bottom slide gate of the connected steel ladle as template images for target detection and recognition; 3.1.2) Traverse and search similar sub-images from the real-time acquired images according to the size of the template image to identify the target to be detected; 3.1.3) If the target exists, distinguish between a cold water inlet and a hot water inlet based on the brightness of the target image; if the target does not exist, the detection ends.

6. The automatic detection method for the opening degree of the sliding gate nozzle of a continuous casting ladle according to claim 4, characterized in that... In step 3.2), the acquired image containing the target is processed to obtain the pixel coordinates reflecting the location of the target, specifically including: 3.2.1) Extract the effective detection selection area from the acquired image where the target is to be detected; 3.2.2) Perform preliminary denoising on the image of the effective detection area to initially filter out the influence of Martian smoke on the image; 3.2.3) Binarize the image of the effective detection selection area to obtain the target contour; 3.2.4) Perform secondary denoising on the target contour image to further filter out the influence of external burrs, small sparks, and other bright light on the contour; 3.2.5) Based on the position of the target contour after the secondary denoising process, generate the corresponding target contour bounding rectangle and determine its center coordinate point; use the horizontal pixel coordinates of the center coordinate point to represent the pixel coordinates P of the detected target.

7. The automatic detection method for the opening degree of the sliding gate nozzle in continuous casting ladle according to claim 6, characterized in that: The effective detection area is pre-selected according to the displacement change range of the detection target; The preliminary denoising process for the effective detection selection area image is performed using a Gaussian filtering algorithm. The binarization processing of the effective detection area image includes selecting different binarization thresholds based on whether the water inlet is cold or hot. The target contour image undergoes secondary denoising processing using erosion and dilation algorithms.

8. The automatic detection method for the opening degree of the sliding gate nozzle of a continuous casting ladle according to claim 4, characterized in that... In step 3.3), the pixel coordinates of the detected target are converted into a detection quantity that reflects the opening degree of the ladle sliding gate, specifically including: 3.3.1) After the ladle reaches the casting position and the long nozzle is installed, zero-position calibration is performed with the sliding nozzle fully closed. The target pixel coordinates P at this time are extracted as the zero-position pixel coordinates P0 of the current ladle. 3.3.2) After the ladle is opened for casting, the fully open position calibration is performed with the sliding gate fully open, and the target pixel coordinate P at this time is extracted as the fully open position pixel coordinate P1 of the current ladle. 3.3.3) Based on the real-time change of the target pixel coordinate P of the ladle as the sliding gate opening changes during casting, the ladle sliding gate opening detection quantity SG is determined: The measured value of the sliding gate opening of the ladle is SG = [(P-P0) / (P1-P0)] × 100%; Where P is the target pixel coordinate, P1 is the fully open pixel coordinate of the ladle, and P0 is the zero pixel coordinate of the ladle.

9. An automatic detection system for the opening degree of the sliding gate nozzle in a continuous casting ladle, characterized in that: It consists of an image acquisition unit, a signal transmission unit, an image processing unit, and a signal output unit; The image acquisition unit is used to acquire images of the moving target connected to the sprue nozzle of the ladle in real time and generate digital signal output. The signal transmission unit is connected to the image acquisition unit and is used to transmit the digital image signal emitted by the image acquisition unit to the image processing unit. The image processing unit is connected to the signal transmission unit and is used to receive image information from the image acquisition unit in real time, automatically analyze and process the target image that moves with the ladle slide gate nozzle, and form a detection quantity reflecting the opening degree of the ladle slide gate nozzle. The signal output unit is used to output the ladle sliding gate opening detection value generated by the image processing unit to the user system through the signal output interface for display, alarm and control; The aforementioned automatic detection system for the opening of the sliding gate of the continuous casting ladle directly acquires and analyzes images of the moving target connected to the sliding gate of the ladle through a non-contact detection mode of image recognition, thereby obtaining the detection quantity reflecting the opening of the sliding gate of the ladle in real time and realizing the automatic detection of the opening of the sliding gate of the continuous casting ladle.

10. The automatic detection system for the opening degree of the sliding gate nozzle of a continuous casting ladle according to claim 9, characterized in that: The image acquisition unit includes an industrial camera, a protective cover, and a pan-tilt bracket; The industrial camera is used to acquire real-time images of the moving target connected to the sprue nozzle of the ladle and generate a digital signal output suitable for network transmission. The protective cover is used to protect industrial cameras from dust, splashes, and high-temperature environments. The pan-tilt bracket is used to easily adjust the angle and position of the industrial camera capturing images.

11. The automatic detection system for the opening degree of the sliding gate of a continuous casting ladle according to claim 9, characterized in that: The signal transmission unit includes an optoelectronic switch, a network cable, and an optical fiber cable.

12. The automatic detection system for the opening degree of the sliding gate nozzle of a continuous casting ladle according to claim 9, characterized in that: The image processing unit consists of an industrial control computer, which has a central processing unit, a memory, and a network interface.

13. The automatic detection system for the opening degree of the sliding gate of a continuous casting ladle according to claim 9, characterized in that: The image processing unit includes a target recognition module, a target detection module, and an aperture conversion module; The target recognition module is used to identify the target to be detected from the acquired image of the moving target of the connecting ladle slide gate nozzle; The target detection module is used to process the image of the target to obtain pixel coordinates reflecting the location of the target; The opening conversion module is used to convert the pixel coordinates of the detected target into a detection quantity that reflects the opening of the ladle sliding gate.

14. The automatic detection system for the opening degree of the sliding gate of a continuous casting ladle according to claim 9, characterized in that: The signal output interface adopts an Ethernet communication interface and outputs the ladle sliding nozzle opening detection data to the remote monitoring screen of the user system via the OPC communication protocol. At the same time, it is output to the user controller for abnormal alarm and opening control of the ladle sliding nozzle opening.

15. The automatic detection system for the opening degree of the sliding gate of a continuous casting ladle according to claim 9, characterized in that: The aforementioned automatic detection system for the opening of the sliding gate nozzle in continuous casting ladles provides real-time, accurate feedback of the sliding gate nozzle opening to the control system. This system is used for the automatic optimization and control of the sliding gate nozzle opening before final pouring in continuous casting ladles, thereby improving the accuracy of opening control and enhancing the control effect of residual steel quantity in continuous casting ladles.

16. The automatic detection system for the opening degree of the sliding gate of a continuous casting ladle according to claim 9, characterized in that: The aforementioned automatic detection system for the opening of the sliding gate of the continuous casting ladle accurately feeds back the real-time opening of the sliding gate to the control system. This system is used for automatic alarm and interlock control of abnormal opening of the sliding gate of the continuous casting ladle, preventing the risk of abnormal casting interruption accidents caused by the sliding gate opening being too small during casting, and improving the accuracy and automation of production process monitoring.

Images