A method, system, device and medium for detecting converter slag overflow

Abstract

The application discloses a converter slag overflow detection method, system, device and medium, relates to the technical field of steel smelting process control, and the method comprises the steps of obtaining a first video stream and a second video stream of a converter, the first video stream is obtained by a first image acquisition device arranged at the mouth of the converter, and the second video stream is obtained by a second image acquisition device arranged under the converter; slag layer area features are extracted based on the first video stream, and slag particle motion features of the under-furnace area are extracted based on the second video stream; the slag layer area features are the proportion of the total area of slag blocks in the mouth area to the area of the mouth area; the slag layer area features and the slag particle motion features are weighted and fused to calculate a slag overflow index; and the slag overflow index is compared with a preset threshold interval to determine the slag overflow grade of the converter.

Description

Technical Field

[0001] This application relates to the field of steel smelting process control technology, and in particular to a converter slag overflow detection method, system, equipment and medium. Background Technology

[0002] Currently, during the blowing process in converter steelmaking, the molten pool inside the furnace boils violently, and slag easily overflows from the furnace opening, a phenomenon known as "slag overflow" or "slag runoff." Slag overflow can damage equipment such as the furnace opening and fume hood, causing significant economic losses and safety accidents. Therefore, real-time and accurate detection of slag overflow and timely early warning are crucial. Converter slag overflow detection mainly relies on a monitoring scheme using a single camera (usually a furnace opening camera).

[0003] However, changes in the brightness of the high-temperature flame at the furnace opening and the obstruction of smoke and dust can easily interfere with the image analysis of a single camera, making it difficult to distinguish slag overflow from normal flame splashes and sparks, thus leading to false alarms or missed alarms. Summary of the Invention

[0004] The summary section introduces a series of simplified concepts, which will be further explained in detail in the detailed description section. This summary section is not intended to limit the key and essential technical features of the claimed technical solution, nor is it intended to determine the scope of protection of the claimed technical solution.

[0005] In a first aspect, embodiments of this application provide a method for detecting converter slag overflow, the method comprising: The system acquires a first video stream and a second video stream of the converter. The first video stream is acquired by a first image acquisition device deployed at the converter opening, and the second video stream is acquired by a second image acquisition device deployed under the converter. Based on the first video stream, the slag layer area feature is extracted, and based on the second video stream, the slag particle movement feature in the area below the furnace is extracted. The slag layer area feature is the ratio of the total area of ​​slag blocks in the furnace opening area to the area of ​​the furnace opening area. The slag layer area characteristics and the slag particle movement characteristics are weighted and fused to generate the slag overflow index. The overflow index is compared with a preset threshold range to determine the overflow level of the converter.

[0006] In one embodiment of the present invention, the step of extracting the slag layer area features based on the first video stream includes: Extract the furnace opening area image from the first video stream and determine the average grayscale value of the furnace opening area image; The furnace opening area image is segmented based on an adaptive threshold segmentation algorithm to generate a segmented image. The block parameters and constant parameters of the adaptive threshold segmentation algorithm are determined by the average gray value of the furnace opening area image. Perform a closing operation on the segmented image to obtain the first slag block region image; An opening operation is performed on the first processed image to obtain the second slag block region image; The image of the second slag block region is processed based on the contour search algorithm to determine the slag layer area characteristics.

[0007] In one embodiment of the present invention, the slag particle movement characteristics include the number of slag particles and the density of slag particles, and the extraction of slag particle movement characteristics in the under-furnace area based on the second video stream includes: The foreground is extracted from the second video stream based on the background subtraction algorithm to generate a foreground moving target region; Motion vector analysis of pixels between adjacent frames of the second video stream is performed based on optical flow method to generate trajectory information and velocity information of moving targets in the second video stream; Based on a preset brightness threshold, each frame in the second video stream is converted to a preset color space to filter out the bright areas. Based on the intersection of the foreground moving target area, the trajectory information, the velocity information, and the highlighted area, the number and density of slag particles are determined.

[0008] In one embodiment of the present invention, the step of weightedly fusing the slag layer area characteristics and the slag particle movement characteristics to generate a slag overflow index includes: Determine whether the proportion of the total area of ​​the slag block to the area of ​​the furnace opening exceeds a preset proportion threshold. If the proportion of the total area of ​​the slag block to the area of ​​the furnace mouth exceeds a preset threshold, the preset index threshold is used as the slag overflow index.

[0009] In one embodiment of the present invention, the step of weightedly fusing the slag layer area characteristics and the slag particle movement characteristics to generate a slag overflow index includes: If the proportion of the total area of ​​the slag block to the area of ​​the furnace opening does not exceed a preset proportion threshold, it is determined whether the number of slag particles exceeds a preset quantity threshold. If the number of slag particles exceeds a preset threshold, the ratio of the total area of ​​the slag block to the area of ​​the furnace opening is weighted and calculated with the density of the slag particles to generate an overflow slag index.

[0010] In one embodiment of the present invention, after weighted and fused calculation of the slag layer area characteristics and the slag particle movement characteristics to generate the overflow slag index, the following steps are included: The overflow slag index is smoothed over time based on a time window sliding filter algorithm.

[0011] In one embodiment of the present invention, after acquiring the first video stream and the second video stream of the converter, the process includes: Monitor the first network connection status of the first image acquisition device and the second network connection status of the second image acquisition device; If either the first network connection state or the second network connection state is detected as interrupted, a network reconnection operation is performed based on a backoff algorithm until both the first network connection state and the second network connection state are detected as connected.

[0012] Secondly, this application proposes a converter slag overflow detection system, the system comprising: a data acquisition module, a feature extraction module, and a slag overflow determination module; The data acquisition module is configured to acquire a first video stream and a second video stream of the converter, wherein the first video stream is acquired by a first image acquisition device deployed at the converter opening, and the second video stream is acquired by a second image acquisition device deployed under the converter. The feature extraction module is configured to: extract slag layer area features based on the first video stream, and extract slag particle movement features in the under-furnace area based on the second video stream, wherein the slag layer area feature is the ratio of the total area of ​​slag blocks in the furnace mouth area to the area of ​​the furnace mouth area. The slag overflow determination module is configured to: perform weighted fusion calculation on the slag layer area characteristics and the slag particle movement characteristics to generate a slag overflow index; compare the slag overflow index with a preset threshold range to determine the slag overflow level of the converter.

[0013] Thirdly, an electronic device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program stored in the memory to implement the steps of a converter slag overflow detection method as described in any of the first aspects above.

[0014] Fourthly, this application also proposes a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of a converter slag overflow detection method according to any one of the first aspects.

[0015] In summary, the converter slag overflow detection method of this application embodiment acquires a first video stream by deploying a first image acquisition device at the converter opening and a second image acquisition device deployed under the converter to acquire a second video stream. This solves the problem of false alarms that are easily caused by monitoring with a single camera in the prior art. Features are extracted from the first video stream and the second video stream respectively, and then the extracted features are weighted and fused to determine the current slag overflow level of the converter. This distinguishes slag overflow from normal flame splash and sparks, making the slag overflow judgment result more accurate.

[0016] The converter slag overflow detection method proposed in this application, along with other advantages, objectives, and features of this application, will be partly apparent from the following description and partly understood by those skilled in the art through study and practice of this application. Attached Figure Description

[0017] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit this specification. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 This is a schematic flowchart of a converter slag overflow detection method provided in an embodiment of this application; Figure 2 This is a schematic diagram of a converter slag overflow detection system provided in an embodiment of this application; Figure 3 This is a schematic diagram of an electronic device for detecting converter slag overflow provided in an embodiment of this application. Detailed Implementation

[0018] To better understand the technical solutions provided in the embodiments of this specification, the technical solutions of the embodiments of this specification will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the embodiments of this specification and the specific features in the embodiments are detailed descriptions of the technical solutions of the embodiments of this specification, rather than limitations on the technical solutions of this specification. In the absence of conflict, the embodiments of this specification and the technical features in the embodiments can be combined with each other.

[0019] In this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, without necessarily requiring or implying any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element. The term "two or more" includes two or more cases.

[0020] Please see Figure 1 This is a flowchart illustrating a converter slag overflow detection method provided in an embodiment of this application, specifically including: S110. Acquire a first video stream and a second video stream of the converter. The first video stream is acquired by a first image acquisition device deployed at the converter opening, and the second video stream is acquired by a second image acquisition device deployed under the converter. For example, an image acquisition device (camera) is installed at the furnace mouth of the converter, and another image acquisition device (camera) is installed below the furnace. Both devices simultaneously capture video footage of the converter during operation, obtaining a first video stream from the furnace mouth perspective and a second video stream from the below perspective. Specifically, an 8-megapixel high-temperature HD network camera is installed on the furnace-front operating platform of the 210-ton converter as the first image acquisition device at the furnace mouth, directly facing the furnace mouth and covering the entire furnace mouth area. A high-speed dome network camera with a wide-angle lens is installed on the safety island below the furnace as the second image acquisition device, covering the slag ladle car and molten steel ladle car areas. The two cameras simultaneously acquire video streams, providing raw visual data for subsequent inspection.

[0021] By acquiring video from both the furnace opening and the area below the furnace, spatial complementary monitoring of slag overflow phenomena is achieved. The furnace opening can capture the macroscopic slag layer distribution, while the area below the furnace can detect microscopic slag particle splashing, thus solving the problems of limited field of view of a single camera and the risk of missed detection due to obstruction by flames or smoke.

[0022] S120. Extract the ratio of the total area of ​​slag blocks in the furnace mouth region to the area of ​​the furnace mouth region based on the first video stream, and extract the slag particle movement characteristics in the furnace lower region based on the second video stream. For example, image processing is performed on the first video stream from the furnace opening to calculate the proportion of the total area of ​​slag blocks in the furnace opening area to the entire monitored area of ​​the furnace opening, i.e., the slag layer area feature; targeted image processing is performed on the second video stream from below the furnace to extract the motion-related features of slag particles in the area below the furnace, i.e., slag particle motion features. Specifically: for the first video stream captured by the furnace opening camera, the ROI (Region of Interest) region [200, 1500, 1200, 400] at the furnace opening is extracted and processed to calculate the proportion of the total area of ​​slag blocks in this region; for the second video stream captured by the below-furnace camera, the ROI region [850, 10, 500, 1000] is extracted, and the motion features such as the trajectory, quantity, and density of slag particles in this region are extracted.

[0023] Based on the differences in the manifestation of slag overflow at the furnace opening and below the furnace, different core features are extracted. At the furnace opening, the focus is on the area ratio of static or slow-moving slag layers, and below the furnace, the focus is on the movement characteristics of dynamic slag particles. This achieves accurate extraction of multimodal features and avoids the one-sidedness of judging slag overflow based on a single feature.

[0024] S130. The slag layer area characteristics and the slag particle movement characteristics are weighted and fused to generate the overflow slag index. For example, according to preset fusion rules, different weights are assigned to the slag layer area characteristics at the furnace mouth and the slag particle movement characteristics under the furnace. A weighted calculation method is then used to obtain a quantitative index that comprehensively reflects the degree of converter slag overflow, namely the slag overflow index. Specifically, the preset weights for furnace mouth features are 60% and for under-furnace features are 40%. When the slag block area ratio at the furnace mouth is 0.8% and the slag particle density under the furnace is 20 particles / cm², the specific slag overflow index is calculated according to the weighted formula. This transforms the qualitative description of the two-path features into a quantitative value, achieving a quantitative assessment of the degree of slag overflow. By combining the two independent visual features into a comprehensive index through weighted fusion, a scientific judgment logic based on furnace mouth dominance and under-furnace verification is realized, avoiding false alarms caused by single-feature judgments and making slag overflow judgment more objective and accurate.

[0025] S140. The overflow index is compared with a preset threshold range to determine the overflow level of the converter.

[0026] For example, a threshold range for the overflow index is pre-defined. The calculated overflow index is compared with these ranges. Based on the index's range, the current overflow level of the converter is determined to be normal, slight, moderate, or severe, and the result is output. Specifically, the preset threshold ranges are: overflow index less than 1 is normal, 1-5 is slight, 5-15 is moderate, and greater than or equal to 15 is severe. When the calculated overflow index is 3, it is judged as slight overflow; when the overflow index is 18, it is judged as severe overflow. The operator can directly take corresponding measures according to the level. Converting the quantified overflow index into a graded overflow level meets the actual operational needs of industrial sites. Operators do not need to interpret complex values ​​and can quickly take emergency measures according to the level, improving the practicality of the detection results and operational efficiency. At the same time, the graded judgment also provides a basis for multi-level alarms and automated control.

[0027] In summary, the converter slag overflow detection method proposed in this application solves the problem of false alarms caused by monitoring with a single camera in the prior art by acquiring a first video stream through a first image acquisition device deployed at the converter opening and a second image acquisition device deployed under the converter. By extracting features from the first and second video streams respectively and then performing weighted fusion calculation on the extracted features, the current slag overflow level of the converter can be determined, thereby distinguishing slag overflow from normal flame splashes and sparks, making the slag overflow judgment result more accurate.

[0028] In some examples, the extraction of slag layer area features based on the first video stream includes: Extract the furnace opening area image from the first video stream and determine the average grayscale value of the furnace opening area image; The furnace opening area image is segmented based on an adaptive threshold segmentation algorithm to generate a segmented image. The block parameters and constant parameters of the adaptive threshold segmentation algorithm are determined by the average gray value of the furnace opening area image. Perform a closing operation on the segmented image to obtain the first slag block region image; An opening operation is performed on the first processed image to obtain the second slag block region image; The image of the second slag block region is processed based on the contour search algorithm to determine the slag layer area characteristics.

[0029] For example, when extracting slag layer area features from the first video stream, the furnace opening area image is first extracted from the first video stream, and the average grayscale value of the furnace opening area image is determined. Since the brightness of the furnace opening flame changes drastically during converter blowing, fixed-threshold segmentation methods are difficult to adapt to slag block identification under different lighting conditions. Therefore, this invention employs an adaptive threshold segmentation algorithm, where the block parameters and constant parameters are not preset fixed values, but are dynamically determined by the average grayscale value of the current furnace opening area image. Specifically, when the average grayscale value is low, it indicates that the furnace opening area is generally dark. In this case, a smaller block parameter is used for local threshold calculation to enhance the contrast between slag blocks and the background in dark areas, such as 151x151 when the brightness is low. When the average grayscale value is high, it indicates that the furnace opening area is generally bright. In this case, a larger block parameter is used for threshold calculation to avoid over-segmentation of slag blocks in bright areas, such as 201x201 when the brightness is high, to enhance the contrast of slag blocks under different lighting conditions. Based on the dynamically adjusted block parameters and constant parameters, adaptive threshold segmentation is performed on the furnace opening area image to generate a segmented image containing slag block candidate areas.

[0030] After generating the segmented image, it is further optimized through morphological operations. First, a closing operation is performed on the segmented image. This operation, through dilation followed by erosion, bridges the small holes created by threshold segmentation within the slag block region and connects adjacent slag block regions into a continuous whole, thus generating the first slag block region image, in which the connectivity of the slag blocks is enhanced. Next, an opening operation is performed on the first slag block region image. This operation, through erosion followed by dilation, eliminates isolated noise points and burrs on the slag block edges, preserving the main slag block region while removing false interference, thus generating the second slag block region image. In this second slag block region image, the slag block shape is more regular and the noise is significantly reduced. Finally, a contour-finding algorithm is used to process the second slag block region image, extracting the contours of all slag block regions in the image. The proportion of the total slag block area to the furnace mouth area is calculated based on the pixel area of ​​each contour, thereby determining the slag layer area characteristics.

[0031] The adaptive threshold parameter is dynamically adjusted based on the average gray value of the furnace mouth image, allowing the segmentation algorithm to adapt to the complex environment of varying flame brightness at the furnace mouth and improve the accuracy of slag block segmentation. The combination of morphological closing and opening operations effectively connects scattered slag blocks and removes noise, avoiding the problems of missed slag blocks and miscounting noise.

[0032] In some examples, the slag particle movement characteristics include the number and density of slag particles, and the extraction of slag particle movement characteristics in the under-furnace region based on the second video stream includes: The foreground is extracted from the second video stream based on the background subtraction algorithm to generate a foreground moving target region; Motion vector analysis of pixels between adjacent frames of the second video stream is performed based on optical flow method to generate trajectory information and velocity information of moving targets in the second video stream; Based on a preset brightness threshold, each frame in the second video stream is converted to a preset color space to filter out the bright areas. Based on the intersection of the foreground moving target area, the trajectory information, the velocity information, and the highlighted area, the number and density of slag particles are determined.

[0033] For example, when extracting slag particle motion features from the under-furnace area based on the second video stream, a background subtraction algorithm is first applied to extract the foreground of the second video stream, generating foreground moving target regions. The core of the background subtraction algorithm lies in establishing and dynamically updating the background model. For example, a Gaussian mixture model is used to perform multimodal modeling of the grayscale distribution of each pixel in the under-furnace scene. The current frame image is compared pixel by pixel with the background model. Pixels with deviations exceeding a preset threshold are determined as foreground moving targets, thereby separating all dynamically changing regions from the complex under-furnace scene. These regions may contain splashed slag particles, drifting smoke and dust, or moving equipment shadows, providing initial motion candidate regions for subsequent fine screening. On this basis, optical flow is further used to perform motion vector analysis on the pixels between adjacent frames of the second video stream. By calculating the changes of pixels in the time domain and the correlation between adjacent frames, the instantaneous motion direction and velocity of each pixel are determined, thereby generating trajectory information and velocity information of moving targets in the second video stream. Optical flow can effectively capture the continuous displacement characteristics of moving targets. For example, the Farneback dense optical flow algorithm is used to calculate the two-dimensional motion vector of each pixel. Based on this, it is possible to distinguish between real moving targets with continuous motion trajectories and instantaneous pseudo-motion regions caused by light fluctuations or camera shake, and further distinguish between real moving particles and static bright spots.

[0034] While acquiring the foreground moving target area and the trajectory and speed information of the moving target, each frame of the second video stream is converted to a preset color space based on a preset brightness threshold to filter out bright areas. Specifically, the image in the original RGB (Red-Green-Blue) color space is converted to the HSV (Hue-Saturation-Value) color space. The HSV color space decomposes color into three independent components: hue, saturation, and brightness, which can more accurately characterize the high-temperature luminescence characteristics of slag particles. The preset brightness threshold is used to filter pixels whose brightness component is higher than the threshold. Combined with the constraint condition of the hue component in the range of white to bright yellow, bright areas that match the optical characteristics of high-temperature slag particles are separated from the image. These areas contain potential slag particle candidate points, but may also contain non-slag particle interference such as static metal reflections or high-temperature equipment surfaces. Finally, a logical AND operation is performed on the foreground moving target area, the trajectory and velocity information verified by the optical flow method, and the highlighted area. That is, only pixels that simultaneously meet the following conditions are retained: they belong to the foreground moving target area, have valid trajectory and velocity information that conforms to the motion characteristics, and are located in the highlighted area filtered by the preset brightness threshold. Through this strict intersection judgment of multimodal information, static bright reflections, non-slag moving objects, and various noise interferences are effectively removed, generating a pure slag mask area. Then, based on the number of connected components in this slag mask area, the number and density of slag particles are determined, which serve as the core motion feature characterizing the intensity of slag overflow under the furnace.

[0035] Background subtraction algorithm effectively separates the foreground and background, avoiding the impact of background interference on slag particle detection; motion vector analysis using optical flow method accurately identifies the motion characteristics of slag particles, effectively distinguishing between real slag splashes and static bright spots and irrelevant moving debris; brightness threshold screening in color space utilizes the high brightness characteristics of high-temperature slag particles to eliminate interference sources such as flames and sparks; intersection screening with multiple conditions achieves accurate identification of slag particles under the furnace, significantly reducing the false alarm rate of under-furnace feature extraction and making the statistical results of slag particle quantity and density more accurate.

[0036] In some examples, the weighted fusion calculation of the slag layer area characteristics and the slag particle movement characteristics to generate the slag overflow index includes: Determine whether the proportion of the total area of ​​the slag block to the area of ​​the furnace opening exceeds a preset proportion threshold. If the proportion of the total area of ​​the slag block to the area of ​​the furnace mouth exceeds a preset threshold, the preset index threshold is used as the slag overflow index.

[0037] For example, when the proportion of the total slag area to the furnace mouth area exceeds a preset threshold, it is determined that the current slag overflow status of the converter has reached a significant level of visibility at the furnace mouth. The preset threshold is an empirically determined value based on the critical visual characteristics of slag overflow in converter steelmaking. When the slag coverage rate at the furnace mouth exceeds this threshold, it indicates that a large area of ​​slag has accumulated or overflowed at the furnace mouth. At this point, the slag overflow phenomenon is very clear and does not require auxiliary verification from under-furnace features. Based on this determination logic, a preset index threshold is directly output as the slag overflow index. This preset index threshold corresponds to the lower limit of the medium or severe level in the slag overflow classification system, ensuring that the system can immediately output a high slag overflow index that matches the significant slag overflow characteristics at the furnace mouth, avoiding potential response delays caused by waiting for under-furnace feature calculations or weighted fusion processes.

[0038] Specifically, the preset threshold for the proportion of slag block area at the furnace mouth is 0.01 (1%), and the preset index threshold is 15 (corresponding to a severe slag overflow level). When the slag block area proportion at the furnace mouth is 1.2% in the later stage of the 210-ton converter blowing process, which exceeds the 1% threshold, it indicates that there is obvious slag overflow at the furnace mouth. The system directly sets the slag overflow index to 15, without needing to calculate the weight of the slag particle characteristics under the furnace, and directly enters the severe slag overflow judgment stage.

[0039] An excessive proportion of slag block area at the furnace mouth is a direct and obvious manifestation of slag overflow. Rapid judgment in this case can improve the system response speed; it avoids redundant calculation of features under the furnace, reduces the system's computational load, and improves the real-time performance of the detection algorithm; direct judgment of obvious slag overflow at the furnace mouth can effectively avoid judgment delays caused by weighted calculations, and buy time for operators to handle emergencies.

[0040] In some examples, the weighted fusion calculation of the slag layer area characteristics and the slag particle movement characteristics to generate the slag overflow index includes: If the proportion of the total area of ​​the slag block to the area of ​​the furnace opening does not exceed a preset proportion threshold, it is determined whether the number of slag particles exceeds a preset quantity threshold. If the number of slag particles exceeds a preset threshold, the ratio of the total area of ​​the slag block to the area of ​​the furnace opening is weighted and calculated with the density of the slag particles to generate an overflow slag index.

[0041] For example, if the proportion of the total area of ​​the slag block to the area of ​​the furnace mouth region does not exceed a preset proportion threshold, it is determined whether the number of slag particles exceeds a preset number threshold. Here, the preset proportion threshold is an empirical value pre-calibrated based on the critical visual characteristics of slag overflow at the furnace mouth in converter steelmaking processes, used to define whether a directly observable significant slag overflow state has occurred at the furnace mouth; the preset number threshold is an empirical value pre-set based on the correlation between slag particle splashing under the furnace and the early stage of slag overflow, used to define whether there is sufficient activity of slag particles under the furnace to characterize the initial signs of slag overflow. When the proportion of the slag block area at the furnace mouth does not exceed the preset proportion threshold, it indicates that the current slag overflow state has not yet formed significant characteristics at the furnace mouth level, and a clear judgment cannot be made solely based on furnace mouth characteristics; if the number of slag particles under the furnace exceeds the preset number threshold, it indicates that a relatively active slag particle splashing phenomenon has occurred in the under-furnace area, and slag overflow may be in the initial splashing stage, requiring further comprehensive judgment based on under-furnace characteristics.

[0042] When the number of slag particles exceeds a preset threshold, a weighted fusion calculation is performed using the ratio of the total area of ​​the slag lumps to the area of ​​the furnace mouth region and the density of the slag particles to generate a slag overflow index. Specifically, firstly, the ratio of the total area of ​​the slag lumps to the area of ​​the furnace mouth region is normalized to generate a first normalized feature value, which characterizes the relative degree of slag layer coverage at the furnace mouth. Simultaneously, the density of the slag particles is normalized to generate a second normalized feature value, which characterizes the relative intensity of slag particle splashing below the furnace. Subsequently, the first normalized feature value is multiplied by a first weighting coefficient to obtain a first weighted value, and the second normalized feature value is multiplied by a second weighting coefficient to obtain a second weighted value, wherein the value range of the first weighting coefficient is 50% to 70%, and the value range of the second weighting coefficient is 30% to 50%. Finally, the first weighted value and the second weighted value are added together to obtain the overflow index. This overflow index serves as a quantitative indicator that comprehensively reflects the current degree of slag overflow in the converter and is used for subsequent slag overflow level determination.

[0043] Specifically, the system sets a threshold of 1% for the area of ​​slag blocks at the furnace opening and a threshold of 12 for the number of slag particles below the furnace. The furnace opening feature has a weight of 60%, and the below-furnace feature has a weight of 40%. During the mid-stage of the 210-ton converter blowing process, if the area of ​​slag blocks at the furnace opening is 0.7% (not exceeding 1%), but the number of slag particles detected below the furnace is 15 (more than 12), it indicates that slag overflow is in the initial splashing stage. At this time, the system performs a weighted summation of the furnace opening slag block area ratio (0.7%) and the below-furnace slag particle density (e.g., 18 particles / cm²) to obtain a specific slag overflow index (e.g., 0.7 × 60% + 18 × 40% = 7.62). Based on the calculated slag overflow index, it is mapped to different slag overflow levels. For example, an overflow index < 1 is "Normal," 1 ≤ overflow index < 5 is "Slight," 5 ≤ overflow index < 15 is "Medium," and an overflow index ≥ 15 is "Serious." The calculated slag overflow index is 7.62, therefore the slag overflow level at this time is medium.

[0044] In response to the situation where the initial characteristics of slag overflow are not obvious at the furnace mouth but slag particles have already splashed under the furnace, early warning of slag overflow is achieved through auxiliary verification of characteristics under the furnace, solving the problem of missed detection in the early stage of slag overflow by traditional detection methods. The weight of furnace mouth and under the furnace is reasonably allocated, which retains the dominance of furnace mouth characteristics while giving full play to the early monitoring advantages of under the furnace characteristics, making the calculation of the slag overflow index more in line with the actual development process of slag overflow. The quantitative calculation of the initial stage of slag overflow allows operators to discover potential slag overflow hazards in advance and take timely minor intervention measures to prevent the slag overflow from escalating further.

[0045] In some examples, after weighted and fused calculation of the slag layer area characteristics and the slag particle movement characteristics to generate the slag overflow index, the following steps are included: The overflow slag index is smoothed over time based on a time window sliding filter algorithm.

[0046] For example, after weighted fusion calculation of the slag layer area features and slag particle movement features to generate the overflow index, the overflow index is further smoothed temporally based on a time window sliding filter algorithm. Specifically, a fixed-length first-in-first-out queue is constructed as a sliding time window. The length of the sliding time window is preset according to the dynamic characteristics of the converter blowing process and the frame rate of the image acquisition device, for example, set to the time span corresponding to 10 frames of images. The continuously calculated overflow indices are stored sequentially into the sliding time window in chronological order. Whenever a new overflow index is generated and stored in the window, the overflow index stored first in the window is removed, thereby maintaining the overflow index sequence of the most recent time span within the window. Subsequently, a moving average filter or median filter operation is performed on the overflow index sequence currently contained in the sliding window. The moving average filter calculates the arithmetic mean of all overflow indices in the window as the smoothed overflow index at the current moment, while the median filter sorts all overflow indices in the window by value and takes the median value as the smoothed overflow index at the current moment. The overflow index, after time-series smoothing, replaces the original calculated value and serves as the basis for subsequent overflow level determination, thereby ensuring that the final output overflow level is based on the comprehensive analysis results of multiple consecutive frames of images, rather than the instantaneous detection results of a single frame of images.

[0047] It effectively eliminates the abnormal fluctuations in the slag overflow index caused by instantaneous interference in single-frame images (such as sudden flashes of the furnace flame or accidental sparks splashing from below the furnace), significantly reducing the false alarm rate of the system; it makes the output results of the slag overflow index more stable and smooth, avoiding frequent jumps in the slag overflow level and improving the reliability of the detection results; it meets the stability requirements of industrial-grade detection systems, ensuring the accuracy of alarm signals and avoiding operator fatigue caused by frequent false alarms.

[0048] In some examples, after acquiring the first and second video streams from the converter, the process includes: Monitor the first network connection status of the first image acquisition device and the second network connection status of the second image acquisition device; If either the first network connection state or the second network connection state is detected as interrupted, a network reconnection operation is performed based on a backoff algorithm until both the first network connection state and the second network connection state are detected as connected.

[0049] For example, after acquiring the first and second video streams from the converter, the system continuously monitors the first network connection status of the first image acquisition device and the second network connection status of the second image acquisition device. The first network connection status represents the connectivity of the communication link between the first image acquisition device and the processing unit, and the second network connection status represents the connectivity of the communication link between the second image acquisition device and the processing unit. The system monitors the two network connection statuses in real time by periodically sending heartbeat detection packets or monitoring the status of the underlying network interface. When the system detects that either the first or second network connection status is interrupted, it determines that the video stream transmission of the corresponding image acquisition device is abnormal. At this time, a network reconnection operation is immediately triggered to ensure that the dual video streams can resume stable acquisition as soon as possible, avoiding the failure of the overflow detection function due to the interruption of a single video stream.

[0050] If either the first or second network connection state is detected as interrupted, the system performs a network reconnection operation based on a backoff algorithm. The core mechanism of this backoff algorithm is that after each failed reconnection attempt, the waiting time for the next reconnection increases according to a preset pattern, thereby avoiding exacerbating network congestion or consuming system resources due to frequent reconnection requests during network fluctuations or recovery. Specifically, the system first attempts to re-establish the connection with the interrupted image acquisition device at an initial reconnection interval. If the reconnection fails, the reconnection interval is multiplied by a preset growth factor, for example, setting the reconnection interval sequentially to 1 second, 2 seconds, 4 seconds, 8 seconds, 16 seconds, 32 seconds, or increasing the reconnection interval by a factor of 1.5, until a preset maximum reconnection delay threshold is reached, such as 60 seconds. After reaching the maximum reconnection delay threshold, the system will maintain this maximum interval and continue reconnection attempts until both the first and second network connection states are detected as connected, meaning that both video streams have fully resumed acquisition. This adaptive reconnection mechanism ensures that the system can quickly restore the connection when the network fluctuates briefly, and continue to wait for recovery with low resource consumption when the network is interrupted for a long time, thereby ensuring the continuous availability of the system in harsh industrial environments.

[0051] like Figure 2 As shown, this application proposes a converter slag overflow detection system, which includes: a data acquisition module 21, a feature extraction module 22, and a slag overflow determination module 23; The data acquisition module 21 is configured to acquire a first video stream and a second video stream of the converter, wherein the first video stream is acquired by a first image acquisition device deployed at the converter opening, and the second video stream is acquired by a second image acquisition device deployed under the converter. The feature extraction module 22 is configured to: extract slag layer area features based on the first video stream, and extract slag particle movement features in the furnace lower region based on the second video stream, wherein the slag layer area feature is the ratio of the total area of ​​slag blocks in the furnace mouth region to the area of ​​the furnace mouth region. The slag overflow determination module 23 is configured to: perform weighted fusion calculation on the slag layer area characteristics and the slag particle movement characteristics to generate a slag overflow index; compare the slag overflow index with a preset threshold range to determine the slag overflow level of the converter.

[0052] The effects of applying the aforementioned method in the above system can be found in the description of the aforementioned method embodiments, and will not be repeated here.

[0053] like Figure 3 As shown, this application embodiment also provides an electronic device 300, including a memory 310, a processor 320, and a computer program 311 stored in the memory 310 and executable on the processor. When the processor 320 executes the computer program 311, it implements the steps of any of the above-mentioned converter overflow detection methods.

[0054] Since the electronic device described in this embodiment is the device used to implement a converter slag overflow detection device in the embodiments of this application, those skilled in the art can understand the specific implementation method and various variations of the electronic device in this embodiment based on the method described in the embodiments of this application. Therefore, how the electronic device implements the method in the embodiments of this application will not be described in detail here. Any device used by those skilled in the art to implement the method in the embodiments of this application falls within the scope of protection of this application.

[0055] In practical implementation, when the computer program 311 is executed by the processor, it can achieve the following: Figure 1 Any of the corresponding implementation methods in the embodiments.

[0056] It should be noted that the descriptions of each embodiment in the above embodiments have different focuses. For parts that are not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0057] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-readable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-readable program code.

[0058] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create a machine for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0059] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0060] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0061] This application also provides a computer program product, which includes computer software instructions that, when executed on a processing device, cause the processing device to execute the LDPC decoding method of a solid-state drive controller.

[0062] A computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the flow or function according to the embodiments of this application is generated. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium may be any available medium that a computer can store or a data storage device such as a server or data center that integrates one or more available media. The available medium may be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid-state disk (SSD)).

[0063] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0064] In the several embodiments provided in this application, it should be understood that the disclosed devices, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces, or indirect coupling or communication connection between devices or units, and may be electrical, mechanical, or other forms.

[0065] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0066] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0067] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0068] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

[0069] Although preferred embodiments have been described in this specification, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this specification.

[0070] Obviously, those skilled in the art can make various modifications and variations to this specification without departing from its spirit and scope. Therefore, if such modifications and variations fall within the scope of the claims and their equivalents, this specification is also intended to include such modifications and variations.

Claims

1. A method for detecting converter slag overflow, characterized in that, The method includes: The system acquires a first video stream and a second video stream of the converter. The first video stream is acquired by a first image acquisition device deployed at the converter opening, and the second video stream is acquired by a second image acquisition device deployed under the converter. Based on the first video stream, the slag layer area feature is extracted, and based on the second video stream, the slag particle movement feature in the area below the furnace is extracted. The slag layer area feature is the ratio of the total area of ​​slag blocks in the furnace opening area to the area of ​​the furnace opening area. The slag layer area characteristics and the slag particle movement characteristics are weighted and fused to generate the slag overflow index. The overflow index is compared with a preset threshold range to determine the overflow level of the converter.

2. The converter slag overflow detection method according to claim 1, characterized in that, The step of extracting slag layer area features based on the first video stream includes: Extract the furnace opening area image from the first video stream and determine the average grayscale value of the furnace opening area image; The furnace opening area image is segmented based on an adaptive threshold segmentation algorithm to generate a segmented image. The block parameters and constant parameters of the adaptive threshold segmentation algorithm are determined by the average gray value of the furnace opening area image. Perform a closing operation on the segmented image to obtain the first slag block region image; An opening operation is performed on the first processed image to obtain the second slag block region image; The image of the second slag block region is processed based on the contour search algorithm to determine the slag layer area characteristics.

3. The converter slag overflow detection method according to claim 1, characterized in that, The slag particle movement characteristics include the number and density of slag particles. The extraction of slag particle movement characteristics in the under-furnace region based on the second video stream includes: The foreground is extracted from the second video stream based on the background subtraction algorithm to generate a foreground moving target region; Motion vector analysis of pixels between adjacent frames of the second video stream is performed based on optical flow method to generate trajectory information and velocity information of moving targets in the second video stream; Based on a preset brightness threshold, each frame in the second video stream is converted to a preset color space to filter out the bright areas. Based on the intersection of the foreground moving target area, the trajectory information, the velocity information, and the highlighted area, the number and density of slag particles are determined.

4. The converter slag overflow detection method according to claim 3, characterized in that, The step of weightedly fusing the slag layer area characteristics and the slag particle movement characteristics to generate the overflow slag index includes: Determine whether the proportion of the total area of ​​the slag block to the area of ​​the furnace opening exceeds a preset proportion threshold. If the proportion of the total area of ​​the slag block to the area of ​​the furnace mouth exceeds a preset threshold, the preset index threshold is used as the slag overflow index.

5. The converter slag overflow detection method according to claim 4, characterized in that, The step of weightedly fusing the slag layer area characteristics and the slag particle movement characteristics to generate the overflow slag index includes: If the proportion of the total area of ​​the slag block to the area of ​​the furnace opening does not exceed a preset proportion threshold, it is determined whether the number of slag particles exceeds a preset quantity threshold. If the number of slag particles exceeds a preset threshold, the ratio of the total area of ​​the slag block to the area of ​​the furnace opening is weighted and calculated with the density of the slag particles to generate an overflow slag index.

6. The converter slag overflow detection method according to claim 1, characterized in that, After weighted and fused calculation of the slag layer area characteristics and the slag particle movement characteristics to generate the overflow slag index, the following steps are included: The overflow slag index is smoothed over time based on a time window sliding filter algorithm.

7. The converter slag overflow detection method according to claim 1, characterized in that, After acquiring the first and second video streams from the converter, the process includes: Monitor the first network connection status of the first image acquisition device and the second network connection status of the second image acquisition device; If either the first network connection state or the second network connection state is detected as interrupted, a network reconnection operation is performed based on a backoff algorithm until both the first network connection state and the second network connection state are detected as connected.

8. A converter slag overflow detection system, characterized in that, The system includes: a data acquisition module, a feature extraction module, and an overflow determination module; The data acquisition module is configured to acquire a first video stream and a second video stream of the converter, wherein the first video stream is acquired by a first image acquisition device deployed at the converter opening, and the second video stream is acquired by a second image acquisition device deployed under the converter. The feature extraction module is configured to: extract slag layer area features based on the first video stream, and extract slag particle movement features in the under-furnace area based on the second video stream, wherein the slag layer area feature is the ratio of the total area of ​​slag blocks in the furnace mouth area to the area of ​​the furnace mouth area. The slag overflow determination module is configured to: perform weighted fusion calculation on the slag layer area characteristics and the slag particle movement characteristics to generate a slag overflow index; compare the slag overflow index with a preset threshold range to determine the slag overflow level of the converter.

9. An electronic device, comprising: The memory and processor are characterized in that the processor is used to execute a computer program stored in the memory to implement the steps of a converter slag overflow detection method as described in any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of a converter slag overflow detection method as described in any one of claims 1-7.

Images