Fully mechanized coal mining face large coal pre-crushing method and device and fully mechanized coal mining face video monitoring system

Abstract

The invention provides a fully mechanized coal mining face lump coal pre-crushing method and device and a fully mechanized coal mining face video monitoring system, and relates to the technical field of computers, and the method comprises the steps: carrying out the semantic segmentation of each image frame in a real-time video stream, and obtaining a coal flow region image corresponding to each image frame; the real-time video stream is used for monitoring the coal flow of the fully mechanized coal mining face; generating a coal flow area image queue based on the coal flow area image corresponding to each image frame; performing image recognition on the coal flow area image queue, and determining a large coal recognition result of the fully mechanized coal mining face; and starting a pre-crusher corresponding to the fully mechanized coal face to pre-crush the large coal in the coal flow based on the large coal identification result. According to the method, the device and the system provided by the invention, fine monitoring on the coal flow area is realized, the coal piling phenomenon is prevented, and the comprehensive efficiency of mine mining is improved.

Description

Technical Field

[0001] This invention relates to the field of computer technology, and in particular to a method, apparatus, and video monitoring system for pre-crushing large coal blocks in a fully mechanized mining face. Background Technology

[0002] With coal mining entering the era of intelligent technology, the construction of intelligent coal mines is widely recognized as an inevitable trend in the industry's development, a necessary path for the coal industry's growth, and a crucial cornerstone and technological guarantee for the future transformation of coal production methods. While improving coal production efficiency and reducing production costs, intelligent coal mine construction can also enhance coal mine safety, reduce the labor intensity of coal miners, and contribute to the high-quality development of the coal industry.

[0003] In daily underground production, large chunks of coal from the longwall mining face often get stuck at the feed inlet of the transfer conveyor, leading to frequent coal pile-ups. This continuous accumulation of coal can overload or even crush the scraper conveyor, potentially causing equipment malfunctions and safety accidents. Typically, a pre-crusher is installed between the scraper conveyor and the transfer conveyor to pre-crush large coal chunks. A supervisor is required at the head of the conveyor to monitor the operation of the head section and the pre-crusher, ensuring their proper functioning. However, the existence of safety hazards and the possibility of the pre-crusher running idle negatively impacts mine safety and increases production costs.

[0004] Therefore, how to identify large coal pieces collected from fully mechanized mining faces, prevent coal pile-up, and improve the overall efficiency of mine mining has become a technical problem that the industry urgently needs to solve. Summary of the Invention

[0005] This invention provides a method, apparatus, and video monitoring system for pre-crushing large coal blocks in a fully mechanized mining face, which solves the technical problem of how to identify large coal blocks collected in a fully mechanized mining face, prevent coal pile-up, and improve the overall efficiency of mine mining.

[0006] This invention provides a method for pre-crushing large coal blocks in a fully mechanized longwall mining face, comprising: Semantic segmentation is performed on each image frame in the real-time video stream to obtain the coal flow area image corresponding to each image frame; the real-time video stream is used to monitor the coal flow in the fully mechanized mining face; A coal flow region image queue is generated based on the coal flow region image corresponding to each image frame. Image recognition is performed on the coal flow area image queue to determine the identification result of large coal blocks in the fully mechanized mining face; Based on the identification results of the large coal pieces, the pre-crusher corresponding to the fully mechanized mining face is started to pre-crush the large coal pieces in the coal flow.

[0007] In some embodiments, after performing semantic segmentation on each image frame in the real-time video stream to obtain the coal flow region image corresponding to each image frame, the method further includes: When the semantic segmentation results of each image frame include coal flow area images, the first adjustment posture of the camera corresponding to the real-time video stream is determined based on the region position and / or region size of the coal flow area images in each image frame. Based on the first adjustment posture, the pan-tilt head of the camera is controlled to adjust the posture of the camera; In the case where the semantic segmentation results of each image frame do not include coal flow area images, the second adjustment posture of the camera corresponding to the real-time video stream is determined based on the coal mining equipment area images in the semantic segmentation results and the relative position of the coal mining equipment and the coal flow area. Based on the second adjustment posture, the pan-tilt head of the camera is controlled to adjust the posture of the camera.

[0008] In some embodiments, the step of performing image recognition on the coal flow area image queue to determine the identification result of large coal blocks in the fully mechanized mining face includes: The coal flow area image queue is input into the large coal identification model to obtain the large coal prediction results and large coal prediction confidence values ​​of each coal flow area image output by the large coal identification model. If the prediction result of large coal pieces in the coal flow area images of a preset number of images is true and the prediction confidence value of large coal pieces in each coal flow area image is greater than the preset confidence threshold, it is determined that there are large coal pieces in the coal flow of the fully mechanized mining face. Based on the prediction results of large coal blocks in the images of various coal flow regions, the identification results of large coal blocks in the longwall mining face are determined.

[0009] In some embodiments, the step of activating the pre-crusher corresponding to the fully mechanized mining face to pre-crush the large coal pieces in the coal flow based on the large coal piece identification result includes: Based on the large coal identification results, the number of large coal pieces contained in the coal flow is determined; If the number of large coal pieces exceeds a preset quantity, start the pre-crusher corresponding to the fully mechanized mining face and reduce the speed of the coal mining machine and scraper conveyor corresponding to the fully mechanized mining face.

[0010] In some embodiments, activating the pre-crusher corresponding to the fully mechanized mining face includes: Using the identification results of large coal pieces at historical moments as the state, the control parameters of the pre-crusher at historical moments as the actions, and the crushing results of large coal pieces at historical moments as the rewards, reinforcement learning is performed to generate the control parameters of the pre-crusher at the current moment. Based on the control parameters of the pre-crusher at the current moment, the pre-crusher corresponding to the fully mechanized mining face is started.

[0011] In some embodiments, after generating a coal flow region image queue based on the coal flow region images corresponding to each image frame, the method further includes: Image recognition is performed on the coal flow area image queue to determine the coal flow rate of the fully mechanized mining face; When the coal flow rate is greater than the preset flow rate, the pre-crusher corresponding to the fully mechanized mining face is started, and the speed of the coal mining machine and scraper conveyor corresponding to the fully mechanized mining face is reduced.

[0012] This invention provides a pre-crushing device for large coal blocks in a fully mechanized longwall mining face, comprising: The semantic segmentation module is used to perform semantic segmentation on each image frame in the real-time video stream to obtain the coal flow area image corresponding to each image frame; the real-time video stream is used to monitor the coal flow in the fully mechanized mining face. The queue generation module is used to generate a coal flow region image queue based on the coal flow region image corresponding to each image frame. The image recognition module is used to perform image recognition on the coal flow area image queue to determine the identification result of large coal blocks in the fully mechanized mining face; The pre-crushing module is used to start the pre-crusher corresponding to the fully mechanized mining face to pre-crush the large coal pieces in the coal flow based on the identification results of the large coal pieces.

[0013] This invention provides a video monitoring system for a fully mechanized mining face, including a camera, a video stream processing module, a video stream analysis module, and a fully mechanized mining face control module; The camera is mounted on a support along the scraper conveyor of the fully mechanized mining face and is used to monitor the coal flow of the fully mechanized mining face and generate a real-time video stream. The video stream processing module is connected to the camera and is used to send the real-time video stream to the video stream analysis module based on streaming media communication. The video stream analysis module is connected to the video stream processing module and is used to execute the pre-crushing method for large coal blocks in the fully mechanized mining face and generate control commands for the pre-crusher, coal mining machine and scraper conveyor corresponding to the fully mechanized mining face. The fully mechanized mining face control module is connected to the video stream analysis module and is used to control the pre-crusher, the coal mining machine and the scraper conveyor respectively based on the control commands of the pre-crusher, the coal mining machine and the scraper conveyor corresponding to the fully mechanized mining face.

[0014] The present invention provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the pre-crushing method for large coal blocks in a fully mechanized longwall face.

[0015] The present invention provides a non-transitory computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the method for pre-crushing large coal blocks in a fully mechanized mining face.

[0016] The present invention provides a method, apparatus, and video monitoring system for pre-crushing large coal pieces in a fully mechanized mining face. The method involves semantic segmentation of each image frame in a real-time video stream to obtain the coal flow area image corresponding to each image frame. The real-time video stream is used to monitor the coal flow in the fully mechanized mining face. Based on the coal flow area images corresponding to each image frame, a coal flow area image queue is generated. Image recognition is performed on the coal flow area image queue to determine the identification result of large coal pieces in the fully mechanized mining face. Based on the large coal identification result, the pre-crusher corresponding to the fully mechanized mining face is activated to pre-crush the large coal pieces in the coal flow. By monitoring the coal flow in the fully mechanized mining face and performing semantic segmentation on the real-time video stream to obtain the coal flow area image queue, refined monitoring of the coal flow area is achieved. Image recognition of the coal flow area image queue accurately identifies large coal pieces in the coal flow, allowing the pre-crusher to be activated before the large coal pieces enter the transfer machine to crush them, preventing coal accumulation. This eliminates the need for manual supervision, reducing manpower and increasing efficiency, thus improving the overall efficiency of mine mining. Attached Figure Description

[0017] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0018] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0019] Figure 1 This is a flowchart illustrating the method for pre-crushing large coal blocks in a fully mechanized mining face provided by the present invention.

[0020] Figure 2 This is the control logic diagram of the pre-crushing method for large coal blocks in a fully mechanized mining face provided by the present invention.

[0021] Figure 3 This is a schematic diagram of the structure of the pre-crushing device for large coal blocks in a fully mechanized mining face provided by the present invention.

[0022] Figure 4 This is a schematic diagram of the structure of the video monitoring system for the fully mechanized mining face provided by the present invention.

[0023] Figure 5 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0024] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0025] It should be noted that the terms "first," "second," etc., used in this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or device that comprises a series of steps, units, or modules is not necessarily limited to those explicitly listed, but may include other steps, units, or modules not explicitly listed or inherent to such processes, methods, products, or devices.

[0026] In the technical solution of this invention, the collection, storage, use, processing, transmission, provision and disclosure of personal information all comply with the provisions of relevant laws and regulations, necessary confidentiality measures have been taken, and they do not violate public order and good morals.

[0027] Figure 1 This is a schematic flowchart of the pre-crushing method for large coal blocks in a fully mechanized mining face provided by the present invention, as shown below. Figure 1 As shown, the method includes steps 110, 120, 130 and 140.

[0028] Step 110: Perform semantic segmentation on each image frame in the real-time video stream to obtain the coal flow area image corresponding to each image frame; the real-time video stream is used to monitor the coal flow in the fully mechanized mining face.

[0029] Specifically, the main body executing the pre-crushing method for large coal blocks in fully mechanized mining faces provided in this embodiment of the invention is a pre-crushing device for large coal blocks in fully mechanized mining faces. This device can be implemented by software, such as a pre-crushing program for large coal blocks in fully mechanized mining faces running on a server; it can also be implemented by hardware, such as a computer or server that executes the pre-crushing method for large coal blocks in fully mechanized mining faces.

[0030] A fully mechanized longwall face refers to an underground coal mining area in a coal mine that employs comprehensive mechanized mining methods. Fully mechanized longwall faces utilize fully mechanized equipment, including coal mining machines, scraper conveyors, and supports, achieving automation and mechanization of coal mining, transportation, and support. Supports are used to support the roof of the longwall face, ensuring its safety. The coal mining machine typically moves along the strike of the coal seam, cutting the seam and unloading coal blocks onto the scraper conveyor. The scraper conveyor transports the coal blocks cut by the coal mining machine to a transfer conveyor. The transfer conveyor transports the coal blocks to a crusher, and after crushing, they are transported out via a belt conveyor.

[0031] Coal flow refers to the continuous movement of coal in a flowing state on conveying equipment (such as scraper conveyors, belt conveyors, etc.) during coal mining and transportation. The coal flow area refers to the spatial range involved in the entire process of coal being cut from the coal mining machine and transported by scraper conveyors in a fully mechanized mining face.

[0032] Real-time video streaming refers to video content that is transmitted and played in real time as a continuous stream of video data over a network or transmission medium. An image frame is a still image within a video stream at a specific point in time. Video is essentially composed of a series of consecutive image frames.

[0033] Intrinsically safe mining cameras can be installed on the supports along the scraper conveyor line in the fully mechanized mining face. These cameras, equipped with pan-tilt control systems, are intrinsically safe and can be used safely in environments containing explosive gases or other hazardous conditions. The equipment integrates advanced artificial intelligence (AI) algorithms, enabling precise detection, identification, analysis, and real-time alarm processing of target objects. Its main functions include, but are not limited to, automatic tracking, target detection, and abnormal behavior analysis. The cameras monitor the coal flow, generating real-time video streams.

[0034] Semantic segmentation refers to the use of computer vision technology to classify each pixel in an image frame and assign each pixel a semantic category label. Individual image frames in a real-time video stream may contain coal flow areas, coal mining equipment areas, and other background areas.

[0035] A neural network model can be trained using sample image frames. Each pixel in the sample image frame has a semantic category label (coal flow region, coal mining equipment region, and background region), resulting in a coal flow region segmentation model capable of semantic segmentation of image frames. The neural network model can be a convolutional neural network, etc. The coal flow region segmentation model is used to perform semantic segmentation on each image frame, labeling each pixel in the image frame as a coal flow region, coal mining equipment region, and background region, etc. Finally, the coal flow region is segmented, resulting in a coal flow region image. The semantic segmentation result can be a coal flow region image, a coal mining equipment region image, and a background region image within the image frame.

[0036] Step 120: Generate a coal flow region image queue based on the coal flow region images corresponding to each image frame.

[0037] Specifically, each image frame has a timestamp indicating the acquisition time. The coal flow area images corresponding to each image frame can be arranged in chronological order to obtain a coal flow area image queue.

[0038] The coal flow area image queue preserves the temporal and change information of the coal flow, which is helpful for subsequent dynamic analysis.

[0039] Step 130: Perform image recognition on the coal flow area image queue to determine the identification results of large coal blocks in the fully mechanized mining face.

[0040] Specifically, in the coal mining and transportation process, large coal lumps refer to those coal lumps that are significantly larger in size than regular coal lumps. Large coal lumps are typically much larger than ordinary coal lumps, usually have higher hardness, and may contain sharp edges or protruding parts, which increases the risk of jamming or clogging equipment during transportation.

[0041] It can identify each coal flow region image in a coal flow region image queue and determine the prediction results of large coal pieces in each coal flow region image. The prediction results of large coal pieces can include the number of pixels occupied by large coal pieces in the coal flow region image. Based on the number of pixels, the characteristics of large coal pieces can be determined, including area, perimeter, and aspect ratio, and can also be used to estimate the volume and flow rate of large coal pieces.

[0042] Based on the large coal prediction results from various coal flow area images, the identification result of large coal in the fully mechanized mining face can be determined. While false positives may occur in the large coal prediction results from a single coal flow area image, if the large coal prediction results from multiple coal flow area images are consistent or similar, the obtained large coal prediction result is reliable. Therefore, it can be determined that large coal exists in the coal flow of the fully mechanized mining face.

[0043] Step 140: Based on the large coal identification results, start the pre-crusher corresponding to the fully mechanized mining face to pre-crush the large coal in the coal flow.

[0044] Specifically, a pre-crusher is installed between the scraper conveyor and the transfer conveyor to pre-crush large pieces of coal.

[0045] If the presence of large coal pieces in the coal flow of the fully mechanized mining face is confirmed based on the large coal piece identification results, a start signal can be sent to the pre-crusher to control it to begin operation. The rotational speed and crushing force of the pre-crusher are dynamically adjusted according to the size and quantity of the large coal pieces to achieve the best crushing effect.

[0046] The pre-crushing method for large coal in a fully mechanized mining face provided in this invention involves semantic segmentation of each image frame in a real-time video stream to obtain coal flow area images corresponding to each image frame. The real-time video stream is used to monitor the coal flow in the fully mechanized mining face. Based on the coal flow area images corresponding to each image frame, a coal flow area image queue is generated. Image recognition is performed on the coal flow area image queue to determine the identification result of large coal in the fully mechanized mining face. Based on the large coal identification result, the pre-crusher corresponding to the fully mechanized mining face is started to pre-crush the large coal in the coal flow. By monitoring the coal flow in the fully mechanized mining face and performing semantic segmentation on the real-time video stream to obtain the coal flow area image queue, refined monitoring of the coal flow area is achieved. By performing image recognition on the coal flow area image queue, large coal in the coal flow can be accurately identified. The pre-crusher is started in advance to crush the large coal before it enters the transfer machine, preventing coal accumulation. No manual supervision is required, which reduces manpower and increases efficiency, thereby improving the overall efficiency of mine mining.

[0047] It should be noted that each embodiment of the present invention can be freely combined, rearranged, or executed individually, and does not need to rely on or depend on a fixed execution order.

[0048] In some embodiments, after semantic segmentation of each image frame in the real-time video stream to obtain the coal flow region image corresponding to each image frame, the method further includes: When the semantic segmentation results of each image frame include coal flow area images, the first adjustment posture of the camera corresponding to the real-time video stream is determined based on the region position and / or region size of the coal flow area images in each image frame. Based on the first adjustment posture, control the camera's pan-tilt head to adjust the camera's posture; In the absence of coal flow area images in the semantic segmentation results of each image frame, the second adjustment posture of the camera corresponding to the real-time video stream is determined based on the coal mining equipment area images in the semantic segmentation results and the relative position of the coal mining equipment and the coal flow area. Based on the second adjustment posture, the camera's pan-tilt head is controlled to adjust the camera's posture.

[0049] Specifically, in a fully mechanized longwall mining face, the main coal mining equipment includes the coal mining machine, scraper conveyor, and supports. These devices all possess a certain degree of mobility to adapt to the advancement of the mining area. The movement of the coal mining machine, scraper conveyor, and supports is coordinated to ensure the continuity and efficiency of the coal mining process.

[0050] High-definition cameras are installed at regular intervals on the supports along the scraper conveyor line to ensure coverage of the entire coal flow area. The movement of the coal mining equipment will affect the orientation of the cameras mounted on the supports. For example, when the coal mining machine moves on the working face, its position and orientation changes will affect the camera's viewing angle and shooting range; the scraper conveyor moves forward through its pushing device, which may cause the camera's viewing angle to shift; and the orientation adjustments of the supports during movement (such as lifting, pushing, etc.) will directly affect the orientation of the cameras mounted on them.

[0051] Therefore, during the acquisition of real-time video streams, it is necessary to dynamically adjust the camera's orientation. The camera's orientation refers to its position and direction in three-dimensional space.

[0052] If the semantic segmentation results of each image frame include coal flow area images, it indicates that the camera can monitor at least part of the coal flow area, but it cannot be completely determined that the camera can cover the coal flow area and achieve the desired monitoring effect. The location of the coal flow area image in each image frame can be determined based on the position of pixels in the coal flow area image within each image frame; the size (relative value) of the coal flow area image in each image frame can be determined based on the ratio (pixel percentage) of the number of pixels in the coal flow area image to the total number of pixels in each image frame.

[0053] If the coal flow area image is centered within its frame and its size is relatively large (pixel percentage greater than a preset ratio), then the camera can be considered to cover the coal flow area, and no attitude adjustment is needed. If the coal flow area image is located near the edge of its frame and its size is relatively small (pixel percentage less than a preset ratio), then the camera cannot be considered to cover the coal flow area, and attitude adjustment is required.

[0054] At this point, the first adjustment posture of the camera can be determined based on the difference and deviation direction between the center of the coal flow area image and the center of the image frame. Alternatively, the first adjustment posture of the camera can be determined by comparing the size of the coal flow area image with a preset area size, so that the size of the coal flow area image is within a reasonable range.

[0055] Based on the first adjustment posture, control the camera's pan-tilt head to rotate freely up, down, left, and right, thereby adjusting the camera's posture.

[0056] If the semantic segmentation results of each image frame do not include images of the coal flow area, it means that the camera cannot monitor the coal flow area. In this case, the semantic segmentation results may include images of the coal mining equipment area. In a fully mechanized mining face, the relative position of the coal mining equipment (including the coal mining machine, scraper conveyor, or other auxiliary equipment) to the coal flow area is generally fixed. The position of the coal mining equipment can be determined based on the image of the coal mining equipment area, and then the second adjustment posture of the camera can be generated based on the relative position of the coal mining equipment to the coal flow area. For example, if the coal flow area is to the right of the coal mining equipment, and the semantic segmentation results of the image frame only include images of the coal mining equipment area, the second adjustment posture of the camera can be determined to be a rightward adjustment, with an angle range of 0 to 90 degrees.

[0057] Based on the second adjustment posture, the camera's pan-tilt head is controlled to rotate freely up, down, left, and right, thereby adjusting the camera's posture.

[0058] The pre-crushing method for large coal pieces in fully mechanized mining faces provided in this invention dynamically adjusts the camera's posture based on the semantic segmentation results of each image frame. This effectively addresses the impact of equipment movement on the camera's posture, ensuring the stability and reliability of video monitoring and improving the accuracy of identifying large coal pieces in the coal flow.

[0059] In some embodiments, image recognition is performed on a coal flow area image queue to determine the identification result of large coal blocks in the fully mechanized mining face, including: Input the coal flow area image queue into the large coal identification model to obtain the large coal prediction results and large coal prediction confidence values ​​of each coal flow area image output by the large coal identification model. If the prediction result of large coal pieces in the coal flow area images of a preset number of images is true and the prediction confidence value of large coal pieces in each coal flow area image is greater than the preset confidence threshold, it is determined that there are large coal pieces in the coal flow of the fully mechanized mining face. Based on the prediction results of large coal blocks in the images of various coal flow regions, the identification results of large coal blocks in the longwall mining face are determined.

[0060] Specifically, a neural network model can be trained using sample coal flow region images. Each pixel in the sample coal flow region image is labeled with a "large coal" tag (whether it belongs to a large coal lump), resulting in a large coal identification model capable of recognizing coal lump images in the coal flow region. The neural network model can be a convolutional neural network, etc. The large coal identification model is then used to perform image recognition on each coal flow region image, labeling each pixel in the coal flow region image as either a large coal tag or a non-large coal tag, thus obtaining a large coal prediction result. The large coal prediction result can be true (belongs to a large coal lump) or false (does not belong to a large coal lump). Simultaneously, the large coal identification model can also output a large coal prediction confidence value, used to represent the reliability of the model's prediction results. The large coal prediction confidence value is a value between 0 and 1; the higher the value, the higher the reliability of the prediction result.

[0061] If, in the large coal prediction results of each coal flow region image, a predetermined number of coal flow region images show true predictions for large coal pieces, and the confidence value of the large coal predictions for each coal flow region image is greater than a predetermined confidence threshold, then it indicates that a sufficient number of coal flow region images have detected large coal pieces with high reliability. Therefore, it can be determined that large coal pieces exist in the coal flow of the fully mechanized mining face. The predetermined number of images and the predetermined confidence threshold can be set according to actual needs.

[0062] Furthermore, the location of large coal pieces can be determined based on the pixel positions corresponding to large coal pieces in the large coal piece prediction results of each coal flow region image. The size of the large coal piece can be determined based on the ratio of the number of pixels corresponding to the large coal piece to the total number of pixels in the coal flow region image or image frame. The location and size of the large coal piece are then determined as the large coal piece recognition result.

[0063] The pre-crushing method for large coal in fully mechanized mining faces provided in this invention determines the identification result of large coal in fully mechanized mining faces based on the prediction results of large coal in each coal flow area image and the prediction confidence value of large coal, thereby improving the accuracy of identifying the presence of large coal in the coal flow.

[0064] In some embodiments, based on the large coal lump identification results, the pre-crusher corresponding to the fully mechanized mining face is activated to pre-crush the large coal lump in the coal flow, including: Based on the results of large coal identification, the quantity of large coal pieces contained in the coal flow is determined; If the quantity of large coal pieces exceeds the preset quantity, start the pre-crusher corresponding to the fully mechanized mining face and reduce the speed of the coal mining machine and scraper conveyor corresponding to the fully mechanized mining face.

[0065] Specifically, Figure 2 This is the control logic diagram of the pre-crushing method for large coal blocks in a fully mechanized mining face provided by the present invention, as shown below. Figure 2As shown, multiple cameras can be installed at regular intervals on the supports of the fully mechanized mining face. These cameras are used to acquire real-time video streams. The real-time video streams can be used to identify large coal pieces and coal flow congestion.

[0066] Based on the large coal piece identification results, the quantity of large coal pieces in the coal flow can be determined. If the quantity of large coal pieces exceeds the preset quantity, it can be considered that there are too many large coal pieces in the coal flow, which may damage the equipment or cause congestion. In this case, the pre-crusher corresponding to the fully mechanized mining face can be activated to pre-crush the large coal pieces on the scraper conveyor. At the same time, the speed of the coal mining machine and scraper conveyor corresponding to the fully mechanized mining face can be reduced to ensure that the large coal pieces are fully pre-crushed. The preset quantity can be set according to actual needs.

[0067] After starting the pre-crusher and adjusting the speeds of the coal mining machine and scraper conveyor, continue to monitor the number of large coal pieces in the coal flow to ensure stable system operation. For example, a continuous monitoring period can be set. After starting the pre-crusher, the timer begins. If no large coal pieces are detected in the coal flow after the continuous monitoring period ends, the pre-crusher is stopped, and the speeds of the coal mining machine and scraper conveyor are restored to improve coal mining efficiency. If large coal pieces are still detected in the coal flow, the pre-crusher continues to be started, and the speeds of the coal mining machine and scraper conveyor are adjusted.

[0068] The pre-crushing method for large coal in fully mechanized mining faces provided in this embodiment of the invention starts the pre-crusher corresponding to the fully mechanized mining face when the number of large coal pieces exceeds a preset number, and reduces the speed of the coal mining machine and scraper conveyor corresponding to the fully mechanized mining face, so as to ensure that large coal pieces in the coal flow of the fully mechanized mining face are effectively processed, thereby improving production efficiency and equipment operation safety.

[0069] In some embodiments, activating the pre-crusher corresponding to the fully mechanized mining face includes: Using the identification results of large coal pieces at historical moments as the state, the control parameters of the pre-crusher at historical moments as the actions, and the crushing results of large coal pieces at historical moments as the rewards, reinforcement learning is performed to generate the control parameters of the pre-crusher at the current moment. Based on the control parameters of the pre-crusher at the current moment, start the pre-crusher corresponding to the fully mechanized mining face.

[0070] Specifically, reinforcement learning can be used to achieve automatic control of the pre-crusher.

[0071] It can collect historical data on the identification of large coal pieces, the control parameters of the pre-crusher, and the crushing results of large coal pieces.

[0072] The system uses historical data on large coal pieces as its status, including information such as quantity, location, and size. It uses historical data on the control parameters of the pre-crusher as its actions, including the pre-crusher's speed and crushing force. Finally, it uses historical data on the crushing results of large coal pieces as its reward, determining different reward values ​​based on these results.

[0073] The reinforcement learning model is trained using historical states, actions, and rewards. Through trial and error, the model learns the optimal control strategy to maximize cumulative rewards. When given the current large coal chunk identification result, the reinforcement learning model can output the control parameters of the pre-crusher for that current moment.

[0074] After determining the crushing result of large coal pieces at the current moment, the current large coal piece identification result, the control parameters of the pre-crusher, and the large coal piece crushing result are used as learning samples to train the reinforcement learning model and predict the control parameters of the pre-crusher at the next moment.

[0075] The pre-crushing method for large coal blocks in fully mechanized mining faces provided in this invention achieves automated control of the pre-crusher through reinforcement learning, dynamically adjusts the control parameters of the pre-crusher to cope with changes in coal flow, and improves the production efficiency and equipment operation safety of the fully mechanized mining face.

[0076] In some embodiments, after generating a coal flow region image queue based on the coal flow region images corresponding to each image frame, the method further includes: Image recognition is performed on the coal flow area image queue to determine the coal flow rate of the fully mechanized mining face; When the coal flow rate is greater than the preset flow rate, start the pre-crusher corresponding to the fully mechanized mining face and reduce the speed of the coal mining machine and scraper conveyor corresponding to the fully mechanized mining face.

[0077] Specifically, such as Figure 2 As shown, the coal flow rate of the fully mechanized mining face can be determined as the congestion index.

[0078] Image recognition is performed on the coal flow region image queue. The size of the coal block region is determined based on the number of pixels occupied by coal blocks in the coal flow region image. Based on the size of the coal block region, the cross-sectional area of ​​the coal flow is determined. Combined with the belt speed of the scraper conveyor and the density of the coal, the flow rate of the coal flow is calculated.

[0079] If the coal flow rate exceeds the preset flow rate, indicating a potential congestion, the pre-crusher corresponding to the fully mechanized mining face can be activated to pre-crush the coal blocks on the scraper conveyor. Simultaneously, the speeds of the coal mining machine and scraper conveyor at the fully mechanized mining face can be reduced to ensure thorough pre-crushing of the coal blocks. The preset flow rate can be set according to actual needs.

[0080] After starting the pre-crusher and adjusting the speeds of the coal mining machine and scraper conveyor, continue to monitor the coal flow rate to ensure stable system operation. For example, a continuous monitoring period can be set. After starting the pre-crusher, the timer begins. If, after the continuous monitoring period ends, the coal flow rate is not found to be lower than the preset flow rate, the pre-crusher is stopped, and the speeds of the coal mining machine and scraper conveyor are restored to improve coal mining efficiency. If the coal flow rate continues to be monitored to be higher than the preset flow rate, the pre-crusher continues to be started, and the speeds of the coal mining machine and scraper conveyor are adjusted.

[0081] The pre-crushing method for large coal blocks in fully mechanized mining faces provided in this embodiment of the invention starts the pre-crusher corresponding to the fully mechanized mining face when the coal flow rate is greater than the preset flow rate, and reduces the speed of the coal mining machine and scraper conveyor corresponding to the fully mechanized mining face, so as to ensure that the coal flow congestion in the fully mechanized mining face is effectively handled, thereby improving production efficiency and equipment operation safety.

[0082] In some embodiments, lidar and infrared thermal imagers can also be installed on supports along the scraper conveyor to ensure coverage of the entire coal flow area. Lidar is used to acquire three-dimensional point cloud data of the coal flow, while the infrared thermal imager is used to detect the temperature distribution of the coal flow. Combining lidar point cloud data and infrared thermal imaging data supplements visual information and improves the accuracy of coal flow detection.

[0083] It can simultaneously acquire real-time video streams from cameras, point cloud data from LiDAR, and thermal imaging data from infrared thermal imagers. The cameras, LiDAR, and infrared thermal imagers are set to have the same orientation and field of view. Timestamp and spatial calibration techniques are used to ensure temporal and spatial alignment of the multimodal data (image frames, point cloud data, and thermal imaging data).

[0084] The image frames acquired by the camera are subjected to noise reduction and contrast enhancement to adapt to low-light and coal dust interference environments. The point cloud data acquired by the lidar is filtered and downsampled to remove noise points and extract the 3D contour and shape features of the coal flow. The thermal imaging data acquired by the infrared thermal imager is temperature calibrated to extract the temperature distribution characteristics of the coal flow.

[0085] Preprocessed image frames, point cloud data, and thermal imaging data are fused to generate a multimodal feature map. For example, height information from the point cloud data and temperature information from the thermal imaging data are mapped onto the image frames to enhance the feature representation of the coal flow region. This multimodal feature map is then used as input to the model for subsequent semantic segmentation.

[0086] Point cloud data is used to calculate the velocity and load of the coal flow. By analyzing the trajectory and accumulation height of the coal flow in the point cloud data, the dynamic changes of the coal flow are estimated. Thermal imaging data can be used to detect temperature anomalies in large coal pieces, helping to determine their hardness and potential risks. For example, large coal pieces typically have high hardness and large size, and their temperature may be low.

[0087] Real-time monitoring of environmental parameters at the longwall mining face (such as light intensity, coal dust concentration, and temperature). Dynamically adjusting the hyperparameters of the semantic segmentation model (such as learning rate and regularization coefficient) based on changes in these environmental parameters. For example, increasing the weight of thermal imaging data in low-light conditions and increasing the weight of point cloud data when coal dust concentration is high.

[0088] In addition, point cloud data and thermal imaging data can be combined to assist the camera in accurate positioning and ensure accurate identification of coal flow areas.

[0089] In some embodiments, the audio signals extracted from the real-time video stream are useful for identifying large coal pieces, especially when combined with other sensor data, which can significantly improve the accuracy and reliability of the identification.

[0090] By extracting audio signals from real-time video streams, the sound signals generated when large pieces of coal collide or rub against equipment in the coal flow can be captured. After noise reduction processing of these sound signals, spectral features can be extracted. These spectral features are then input into a neural network model for training, resulting in a trained sound recognition model for large coal pieces.

[0091] The real-time collected sound signals are input into the trained large coal sound recognition model, which outputs the sound recognition result for large coal. This sound recognition result, along with the large coal recognition result output by the large coal recognition model in the above embodiment, is used as the basis for determining whether large coal pieces exist. Based on the result, the operating parameters of the coal mining equipment are adjusted in real time, such as starting the pre-crusher or reducing the speed of the coal mining machine and scraper conveyor.

[0092] Vibration sensors and other sensors can also be installed on scraper conveyors. By fusing sound signals with data from other sensors (such as vibration signals), the physical characteristics of the coal flow can be reflected more comprehensively.

[0093] The apparatus provided in the embodiments of the present invention will be described below. The apparatus described below can be referred to in correspondence with the method described above.

[0094] Figure 3 This is a schematic diagram of the structure of the pre-crushing device for large coal blocks in a fully mechanized mining face provided by the present invention, as shown below. Figure 3 As shown, the device includes: The semantic segmentation module 310 is used to perform semantic segmentation on each image frame in the real-time video stream to obtain the coal flow area image corresponding to each image frame; the real-time video stream is used to monitor the coal flow in the fully mechanized mining face. The queue generation module 320 is used to generate a coal flow region image queue based on the coal flow region image corresponding to each image frame. Image recognition module 330 is used to perform image recognition on the coal flow area image queue to determine the identification result of large coal pieces in the fully mechanized mining face; The pre-crushing module 340 is used to start the pre-crusher corresponding to the fully mechanized mining face to pre-crush the large coal pieces in the coal flow based on the identification results of large coal pieces.

[0095] The pre-crushing device for large coal in a fully mechanized mining face provided in this invention performs semantic segmentation on each image frame in a real-time video stream to obtain the coal flow area image corresponding to each image frame. The real-time video stream is used to monitor the coal flow in the fully mechanized mining face. Based on the coal flow area images corresponding to each image frame, a coal flow area image queue is generated. Image recognition is performed on the coal flow area image queue to determine the identification result of large coal in the fully mechanized mining face. Based on the identification result of large coal, the pre-crusher corresponding to the fully mechanized mining face is started to pre-crush the large coal in the coal flow. Because the coal flow in the fully mechanized mining face is monitored and the real-time video stream is semantically segmented to obtain the coal flow area image queue, refined monitoring of the coal flow area is achieved. Because image recognition is performed on the coal flow area image queue, large coal in the coal flow can be accurately identified. The pre-crusher is started in advance to crush the large coal before it enters the transfer machine, preventing coal piling. No manual supervision is required, which reduces manpower and increases efficiency, thereby improving the overall efficiency of mine mining.

[0096] Figure 4 This is a schematic diagram of the structure of the video monitoring system for the fully mechanized mining face provided by the present invention, as shown below. Figure 4 As shown, the system includes a camera 410, a video stream processing module 420, a video stream analysis module 430, and a fully mechanized mining face control module 440.

[0097] The camera is installed on the support along the scraper conveyor of the fully mechanized mining face to monitor the coal flow of the fully mechanized mining face and generate a real-time video stream; The video stream processing module, connected to the camera, is used to send real-time video streams to the video stream analysis module based on streaming media communication. The video stream analysis module, connected to the video stream processing module, is used to execute the pre-crushing method for large coal blocks in the fully mechanized mining face in the above embodiments, and generate control commands for the pre-crusher, coal mining machine and scraper conveyor corresponding to the fully mechanized mining face. The fully mechanized mining face control module, connected to the video stream analysis module, is used to control the pre-crusher, coal mining machine, and scraper conveyor respectively based on the control commands of the pre-crusher, coal mining machine, and scraper conveyor corresponding to the fully mechanized mining face.

[0098] Specifically, cameras are deployed at the longwall mining face. One pan-tilt-zoom (PTZ) camera is fixedly installed on a support directly above the coal flow of the scraper conveyors at regular intervals. Lighting is provided to ensure effective visual monitoring of the coal flow area of ​​the scraper conveyors. Multiple cameras can be used. These cameras can be network cameras. The real-time video stream captured by the cameras can be obtained through their network addresses (IP addresses).

[0099] The video stream processing module is used to receive real-time video streams from cameras, process the video streams, and push the processed video stream data to the video stream analysis module for management, providing synchronous, stable, and reliable real-time video streams for subsequent intelligent video image analysis services.

[0100] The video stream analysis module is used to execute the pre-crushing method for large coal blocks in the fully mechanized mining face described in the above embodiments, and to generate control commands for the pre-crusher, coal mining machine, and scraper conveyor.

[0101] The fully mechanized mining face control module automatically controls the pre-crusher, coal mining machine, and scraper conveyor based on the control commands of the pre-crusher, coal mining machine, and scraper conveyor.

[0102] The video monitoring system for fully mechanized mining faces provided in this invention enables refined monitoring of the coal flow area. It can accurately identify large pieces of coal in the coal flow and start the pre-crusher to crush the large pieces of coal before they enter the transfer machine, preventing coal piling. It eliminates the need for manual supervision, thereby reducing manpower and increasing efficiency, and improving the overall efficiency of mine mining.

[0103] Figure 5 This is a schematic diagram of the structure of the electronic device provided by the present invention, such as... Figure 5 As shown, the electronic device may include: a processor 510, a communications interface 520, a memory 530, and a communications bus 540, wherein the processor 510, the communications interface 520, and the memory 530 communicate with each other via the communications bus 540. The processor 510 can call logical commands stored in the memory 530 to execute the methods described in the above embodiments, for example: Semantic segmentation is performed on each image frame in the real-time video stream to obtain the coal flow region image corresponding to each image frame; the real-time video stream is used to monitor the coal flow in the fully mechanized mining face; a coal flow region image queue is generated based on the coal flow region images corresponding to each image frame; image recognition is performed on the coal flow region image queue to determine the identification result of large coal pieces in the fully mechanized mining face; based on the identification result of large coal pieces, the pre-crusher corresponding to the fully mechanized mining face is started to pre-crush the large coal pieces in the coal flow.

[0104] Furthermore, when the logical commands in the aforementioned memory can be implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a 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 commands 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 described in the various embodiments of this invention. 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.

[0105] The processor in the electronic device provided in this embodiment of the invention can call logical instructions in the memory to implement the above method. Its specific implementation method is the same as the aforementioned method implementation method and can achieve the same beneficial effects, which will not be repeated here.

[0106] This invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to perform the methods provided in the above embodiments.

[0107] The specific implementation method is the same as the aforementioned method implementation method and can achieve the same beneficial effects, so it will not be repeated here.

[0108] This invention provides a computer program product, including a computer program that, when executed by a processor, implements the method described above.

[0109] The device embodiments described above are merely illustrative. 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 modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0110] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0111] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention 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; and these 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 the present invention.

Claims

1. A method for pre-crushing large coal blocks in a fully mechanized longwall mining face, characterized in that, include: Semantic segmentation is performed on each image frame in the real-time video stream to obtain the coal flow region image corresponding to each image frame; The real-time video stream is used to monitor the coal flow at the fully mechanized mining face; A coal flow region image queue is generated based on the coal flow region image corresponding to each image frame. Image recognition is performed on the coal flow area image queue to determine the identification result of large coal blocks in the fully mechanized mining face; Based on the identification results of the large coal pieces, the pre-crusher corresponding to the fully mechanized mining face is started to pre-crush the large coal pieces in the coal flow.

2. The method for pre-crushing large coal blocks in a fully mechanized mining face according to claim 1, characterized in that, After performing semantic segmentation on each image frame in the real-time video stream to obtain the coal flow region image corresponding to each image frame, the method further includes: When the semantic segmentation results of each image frame include coal flow area images, the first adjustment posture of the camera corresponding to the real-time video stream is determined based on the region position and / or region size of the coal flow area images in each image frame. Based on the first adjustment posture, the pan-tilt head of the camera is controlled to adjust the posture of the camera; In the case where the semantic segmentation results of each image frame do not include coal flow area images, the second adjustment posture of the camera corresponding to the real-time video stream is determined based on the coal mining equipment area images in the semantic segmentation results and the relative position of the coal mining equipment and the coal flow area. Based on the second adjustment posture, the pan-tilt head of the camera is controlled to adjust the posture of the camera.

3. The method for pre-crushing large coal blocks in a fully mechanized mining face according to claim 1, characterized in that, The step of performing image recognition on the coal flow area image queue to determine the identification result of large coal blocks in the fully mechanized mining face includes: The coal flow area image queue is input into the large coal identification model to obtain the large coal prediction results and large coal prediction confidence values ​​of each coal flow area image output by the large coal identification model. If the prediction result of large coal pieces in the coal flow area images of a preset number of images is true and the prediction confidence value of large coal pieces in each coal flow area image is greater than the preset confidence threshold, it is determined that there are large coal pieces in the coal flow of the fully mechanized mining face. Based on the prediction results of large coal blocks in the images of various coal flow regions, the identification results of large coal blocks in the longwall mining face are determined.

4. The method for pre-crushing large coal blocks in a fully mechanized mining face according to claim 1, characterized in that, Based on the identification results of the large coal pieces, the pre-crusher corresponding to the fully mechanized mining face is activated to pre-crush the large coal pieces in the coal flow, including: Based on the large coal identification results, the number of large coal pieces contained in the coal flow is determined; If the number of large coal pieces exceeds a preset quantity, start the pre-crusher corresponding to the fully mechanized mining face and reduce the speed of the coal mining machine and scraper conveyor corresponding to the fully mechanized mining face.

5. The method for pre-crushing large coal blocks in a fully mechanized mining face according to claim 4, characterized in that, The activation of the pre-crusher corresponding to the fully mechanized mining face includes: Using the identification results of large coal pieces at historical moments as the state, the control parameters of the pre-crusher at historical moments as the actions, and the crushing results of large coal pieces at historical moments as the rewards, reinforcement learning is performed to generate the control parameters of the pre-crusher at the current moment. Based on the control parameters of the pre-crusher at the current moment, the pre-crusher corresponding to the fully mechanized mining face is started.

6. The method for pre-crushing large coal blocks in a fully mechanized mining face according to claim 1, characterized in that, After generating a coal flow region image queue based on the coal flow region images corresponding to each image frame, the method further includes: Image recognition is performed on the coal flow area image queue to determine the coal flow rate of the fully mechanized mining face; When the coal flow rate is greater than the preset flow rate, the pre-crusher corresponding to the fully mechanized mining face is started, and the speed of the coal mining machine and scraper conveyor corresponding to the fully mechanized mining face is reduced.

7. A pre-crushing device for large coal blocks in a fully mechanized longwall mining face, characterized in that, include: The semantic segmentation module is used to perform semantic segmentation on each image frame in the real-time video stream to obtain the coal flow region image corresponding to each image frame. The real-time video stream is used to monitor the coal flow at the fully mechanized mining face; The queue generation module is used to generate a coal flow region image queue based on the coal flow region image corresponding to each image frame. The image recognition module is used to perform image recognition on the coal flow area image queue to determine the identification result of large coal blocks in the fully mechanized mining face; The pre-crushing module is used to start the pre-crusher corresponding to the fully mechanized mining face to pre-crush the large coal pieces in the coal flow based on the identification results of the large coal pieces.

8. A video monitoring system for a fully mechanized mining face, characterized in that, Includes a camera, a video stream processing module, a video stream analysis module, and a fully mechanized mining face control module; The camera is mounted on a support along the scraper conveyor of the fully mechanized mining face and is used to monitor the coal flow of the fully mechanized mining face and generate a real-time video stream. The video stream processing module is connected to the camera and is used to send the real-time video stream to the video stream analysis module based on streaming media communication. The video stream analysis module is connected to the video stream processing module and is used to execute the pre-crushing method for large coal blocks in a fully mechanized mining face as described in any one of claims 1 to 6, and generate control commands for the pre-crusher, coal mining machine and scraper conveyor corresponding to the fully mechanized mining face. The fully mechanized mining face control module is connected to the video stream analysis module and is used to control the pre-crusher, the coal mining machine and the scraper conveyor respectively based on the control commands of the pre-crusher, the coal mining machine and the scraper conveyor corresponding to the fully mechanized mining face.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method for pre-crushing large coal blocks in a fully mechanized mining face as described in any one of claims 1 to 6.

10. A non-transitory 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 method for pre-crushing large coal blocks in a fully mechanized mining face as described in any one of claims 1 to 6.

Images