Anti-interference method for roller and hydraulic support of coal mining machine

Abstract

The invention belongs to the technical field of mine anti-interference, and discloses a coal mining machine roller and hydraulic support anti-interference method, which comprises the following steps: adding red marks on a coal mining machine roller and a hydraulic support in a coal mining fully mechanized coal mining face, the red marks comprising a red LED lamp group and a red coating; running images of a coal mining machine roller and a hydraulic support are collected through a monitoring camera preset in a coal mining fully mechanized coal mining face; and calculating the minimum distance between the outermost end of the roller of the coal mining machine and the edge of the hydraulic support by combining a target detection algorithm on the basis of the acquired operation image, and executing multi-stage early warning and executing a preset control strategy on the basis of a comparison result of the minimum distance and a preset safety threshold value. According to the technical scheme, the problems that a traditional identifier is prone to being interfered by the environment and low in recognition precision are solved, interference early warning can be achieved, an equipment control system can be directly linked to execute emergency shutdown / pause actions, and compared with a traditional mode only depending on early warning, the anti-interference effectiveness is greatly improved.

Description

Technical Field

[0001] This invention belongs to the field of anti-interference technology in mining, and in particular relates to a method for preventing interference between the drum and hydraulic support of a coal mining machine. Background Technology

[0002] The coal mining machine drum and hydraulic support are two key pieces of equipment in fully mechanized coal mining faces. The coal mining machine is responsible for cutting coal, while the hydraulic support is responsible for support and movement. However, in actual operation, interference problems can easily occur between the two. This mainly happens when the coal mining machine is cutting the coal seam, and the side guards of the corresponding hydraulic support are not fully retracted or not retracted at all, causing the side guards to overlap with the drum, thus leading to interference. In existing technologies, cameras and image recognition solutions are commonly used to solve the interference problem, but the following issues still exist: The working environment has a significant impact: During the coal cutting process, a large amount of dust and water mist are generated around the drum. Existing cameras have poor dust penetration capabilities, which will blur the captured images and make it difficult to meet the requirements for image recognition accuracy. Complex lighting conditions: The lighting layout in underground fully mechanized mining faces is unique, with problems of insufficient overall lighting and localized glare. This will affect the camera's ability to clearly image the coal mining machine drum and hydraulic supports, thus affecting the accuracy of image recognition.

[0003] Limitations of image recognition technology itself: Recognizing complex shapes and postures is difficult. The shapes of coal mining machine drums and hydraulic supports are complex, and their postures change under different working conditions, increasing the difficulty of image recognition and potentially leading to inaccuracies. For example, the cutting teeth on the edges of a high-speed rotating coal mining machine drum are difficult to accurately identify, thus affecting subsequent anti-interference judgments. Simultaneously, image processing involves a large amount of computation. Processing and analyzing images captured by cameras requires extensive calculations, such as feature extraction, target recognition, and distance calculation. This can result in slow system response speeds, failing to meet real-time anti-interference requirements.

[0004] In summary, existing camera recognition solutions are susceptible to the effects of coal dust and lighting conditions, and require high-precision camera deployment, resulting in high costs. To address these issues, this invention proposes a method for preventing interference between the coal mining machine drum and the hydraulic support. Summary of the Invention

[0005] The purpose of this invention is to provide a method for preventing interference between the coal mining machine drum and the hydraulic support, so as to solve the problems existing in the prior art.

[0006] To achieve the above objectives, the present invention provides a method for preventing interference between a coal mining machine drum and a hydraulic support, comprising: S1: Add red markings to the coal mining machine drum and hydraulic support in the fully mechanized coal mining face. The red markings include red LED lights and red paint. S2: The operation images of the coal mining machine drum and hydraulic support are collected by a monitoring camera preset in the coal mining face; wherein, the front end of the lens of the monitoring camera is provided with a filter, which is used to filter out stray light and receive the red light corresponding to the red mark; S3: Based on the acquired operating images, the minimum distance between the outermost end of the coal mining machine drum and the edge of the hydraulic support is calculated using a target detection algorithm. Based on the comparison between the minimum distance and the preset safety threshold, multi-level early warnings are executed and preset control strategies are implemented.

[0007] Optionally, the red LED light group is evenly arranged circumferentially along the edge of the end plate of the coal mining machine drum to form a bright red outline marking around the drum.

[0008] Optionally, when the coal mining machine drum starts to rotate, the red LED lights automatically turn on, and when the coal mining machine drum stops rotating, the red LED lights turn off after a delay.

[0009] Optionally, red paint can be applied to the interference-prone areas of the hydraulic support to form a red marking strip.

[0010] Optionally, step S3 specifically includes: The acquired running images are subjected to denoising and image enhancement processing to obtain preprocessed running images; The red target region of the preprocessed running image is extracted based on the threshold segmentation algorithm, and the candidate regions of the coal mining machine drum and hydraulic support are selected. The filtered images are input into the target detection model for target detection, and the edge detection results of the coal mining machine drum and hydraulic support are output; wherein, the target detection model is built based on a deep learning model; The pixel distance between the coal mining machine drum and the hydraulic support is determined based on the detection results; the minimum distance between the coal mining machine drum and the hydraulic support in actual space is determined based on the pixel distance and the camera calibration parameters. A three-level safety threshold is set, and a corresponding level of early warning control strategy is executed based on the comparison result between the minimum distance and the three-level safety threshold. The three-level safety threshold includes a first-level early warning threshold, a second-level early warning threshold, and a third-level early warning threshold that decrease step by step.

[0011] Optionally, the training process of the target detection model specifically includes: Acquire training data, which includes operational training images of the coal mining machine drum and hydraulic support, and the corresponding edge detection results; An initial object detection model is constructed. The training data is input into the initial object detection model to perform object detection. The training is carried out with the goal of minimizing the loss between the initial training result after object detection and the edge detection result corresponding to the running training image, so as to obtain the trained object detection model.

[0012] Optionally, the implementation of the corresponding level of early warning control strategy specifically includes: When the minimum distance is less than the first-level warning threshold, a yellow audible and visual warning will be issued to alert the operator to pay attention. When the minimum distance is less than the level 2 warning threshold, an orange audible and visual warning is issued, and the coal mining machine is controlled to reduce its speed to 50% of its rated speed; When the minimum distance is less than the level 3 warning threshold, a red audible and visual warning is issued, triggering the coal mining machine shutdown protection, and an alarm message is sent to the control center of the fully mechanized mining face.

[0013] The technical effects of this invention are as follows: This invention uses visual enhancement, intelligent recognition, and real-time early warning as its core technical approaches, achieving anti-interference functionality through three key components: 1. Visual target enhancement: Red LED lights are installed around the coal mining machine drum to provide a directional red light source for the drum; high-saturation red paint is applied to the edges of the hydraulic supports to enhance the visual identification of the support edges; at the same time, a red filter is installed in front of the image acquisition camera to filter stray light from underground, highlight the red target area, and improve image contrast.

[0014] 2. Real-time image acquisition: Explosion-proof cameras are deployed at key locations in the fully mechanized mining face to capture real-time images of the coal mining machine drum and support. The image data is then transmitted to the local control unit via an underground industrial Ethernet network.

[0015] 3. Intelligent Algorithm Analysis and Early Warning: The local control unit is equipped with target detection and distance calculation algorithms to process the acquired images in real time, identify the position coordinates of the drum and the support, and calculate the shortest distance between them. When the distance is less than a preset safety threshold, an audible and visual warning is immediately triggered, and the coal mining machine control system can be linked to implement speed reduction or shutdown protection as needed. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a schematic diagram of the arrangement structure of the red LED lights on the coal mining machine drum in an embodiment of the present invention; Figure 2 This is a schematic diagram of the treatment of the red paint on the edge of the hydraulic support in an embodiment of the present invention; Figure 3 This is a schematic diagram of the camera and filter in an embodiment of the present invention; Figure 4 This is a flowchart illustrating the implementation of an embodiment of the present invention. Detailed Implementation

[0018] Various exemplary embodiments of the present invention will now be described in detail. This detailed description should not be considered as a limitation of the present invention, but rather as a more detailed description of certain aspects, features, and embodiments of the present invention.

[0019] It should be understood that the terminology used in this invention is merely for describing particular embodiments and is not intended to limit the invention. Furthermore, with respect to numerical ranges in this invention, it should be understood that each intermediate value between the upper and lower limits of the range is also specifically disclosed. Every smaller range between any stated value or intermediate value within a stated range, and any other stated value or intermediate value within said range, is also included in this invention. The upper and lower limits of these smaller ranges may be independently included or excluded from the range.

[0020] Various modifications and variations can be made to the specific embodiments described in this specification without departing from the scope or spirit of the invention, as will be apparent to those skilled in the art. Other embodiments derived from this specification will also be obvious to those skilled in the art. This application specification and embodiments are merely exemplary.

[0021] The terms “include,” “including,” “have,” “contain,” etc., used in this article are all open-ended terms, meaning that they include but are not limited to.

[0022] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0023] like Figure 1 - Figure 4 As shown, this embodiment provides a method for preventing interference between the coal mining machine drum and the hydraulic support, including: S1: Add red markings to the coal mining machine drum and hydraulic support in the fully mechanized coal mining face. The red markings include red LED lights and red paint. S2: The operation images of the coal mining machine drum and hydraulic support are collected by a monitoring camera preset in the coal mining face; wherein, the front end of the lens of the monitoring camera is provided with a filter, which is used to filter out stray light and receive the red light corresponding to the red mark; S3: Based on the acquired operating images, the minimum distance between the outermost end of the coal mining machine drum and the edge of the hydraulic support is calculated using a target detection algorithm. Based on the comparison between the minimum distance and the preset safety threshold, multi-level early warnings are executed and preset control strategies are implemented.

[0024] The technical solution of this embodiment solves the problems of traditional signs being easily interfered with by the environment, having low recognition accuracy, and the limitations of image recognition technology itself. It can not only realize interference warning, but also directly link the equipment control system to execute emergency shutdown / pause actions. Compared with the traditional method of relying solely on warnings, it greatly improves the effectiveness of anti-interference.

[0025] 1. Design of a visual target enhancement system (1) Arrangement of red LED lights on the coal mining machine drum Lighting selection: Explosion-proof red LED lights for underground coal mines are selected, with a rated voltage of DC 12V / 24V, which is compatible with the power supply system of the coal mining machine; ensuring good penetration in underground dusty environments; and adapting to high humidity and high dust environments.

[0026] Arrangement method: LED lights are evenly arranged along the edge of the drum end plate in a ring array to ensure that the lights can fully cover the drum blades and end plate area and avoid blind spots in lighting; the lights are fixed to the drum cutting part housing by brackets, maintaining a safe distance from the drum to prevent collision during rotation, and at the same time facilitating cable laying and fixing.

[0027] Control logic: The LED lights are linked to the start-up of the coal mining machine drum. When the drum starts to rotate, the LED lights automatically turn on; when the drum stops, the LED lights turn off after a delay, making it easier to observe the drum status after the machine stops.

[0028] (2) Red paint treatment on the edges of hydraulic supports Coating selection: Use a wear-resistant, corrosion-resistant, and highly saturated red coating specifically for underground coal mines. It has good adhesion and impact resistance, and can withstand the friction of underground coal, the erosion of hydraulic oil, and the mechanical wear caused by the extension and retraction of the support. It ensures a strong visual contrast with the underground environment.

[0029] Application area: Focus on applying paint to the top beam edge, shield beam edge, and front edge of the base of the hydraulic support, forming a continuous red marking band; before painting, the support surface must be derusted and degreased to ensure that the paint adheres tightly to the support surface.

[0030] (3) Configuration of red filter for camera Filter parameters: A narrow-band red filter is selected, with the center wavelength matching the LED light wavelength, effectively filtering incandescent lamps, other equipment light sources and ambient stray light in the well.

[0031] Installation method: The filter is fixed to the front of the explosion-proof camera lens by thread or snap, and is coaxial with the lens to ensure that the image acquisition is distortion-free; the surface of the filter is coated with an anti-reflection film to reduce reflection, and a dustproof protective cover is also installed.

[0032] In this embodiment, red LED lights are installed around the drum of the coal mining machine to provide a directional red light source for the drum; high-saturation red paint is applied to the edge of the hydraulic support to enhance the visual identification of the support edge; at the same time, a red filter is installed in front of the image acquisition camera to filter stray light from underground, highlight the red target area, and improve image contrast.

[0033] 2. Image acquisition system deployment: Camera selection: Explosion-proof high-definition cameras for underground coal mines are selected, with a resolution of no less than 1920×1080 and a frame rate of no less than 25fps to ensure clear and smooth images; the lens focal length is selected as a variable zoom lens, which can adjust the viewing angle according to the width of the working face to cover the main operating areas of the drum and support; it has a wide dynamic range function to adapt to the contrast between strong light and shadow areas under LED lighting.

[0034] Deployment location: The camera is positioned at the fully mechanized mining face, with the lens facing the running trajectory of the coal mining machine drum to ensure a clear capture of the relative position of the drum and the front and rear supports; the camera is fixed to the top beam or column of the support using explosion-proof brackets to prevent collisions when the support extends or retracts.

[0035] Data transmission: In existing technologies, images captured by cameras need to be transmitted to the control center for processing. During data transmission, factors such as network bandwidth and signal interference can cause transmission delays, thus affecting the real-time performance of the anti-interference system. To address this issue, this embodiment uses underground industrial Ethernet to transmit image data. The camera is connected to the nearest support controller via an explosion-proof network cable, and then the data is aggregated to the local control unit of the coal mining machine via a ring network switch, ensuring real-time, delay-free image data transmission.

[0036] This embodiment uses explosion-proof cameras deployed at key locations in the fully mechanized mining face to capture real-time images of the coal mining machine drum and support. Simultaneously, the image data is transmitted to the local control unit via underground industrial Ethernet, enhancing the real-time performance of the anti-interference system.

[0037] 3. Intelligent Algorithm Design and Implementation: The local control unit is equipped with target detection and distance calculation algorithms to process the acquired images in real time, identify the position coordinates of the roller and the support, and calculate the shortest distance between them. Specifically, this includes: (1) Image preprocessing: The acquired images are denoised and enhanced: the median filtering algorithm is used to remove image noise caused by underground dust; histogram equalization is used to enhance the contrast between the red target area and the background; the threshold segmentation algorithm is used to extract the red target area, and the candidate areas of the drum (red LED light illumination area) and the support (red paint area) are initially screened. This stage aims to eliminate underground environmental interference, enhance the features of the red target, and lay the foundation for subsequent identification. The specific steps are as follows: Noise Removal: An adaptive median filtering algorithm is adopted to dynamically adjust the size of the filtering window to remove noise while preserving edge details, thereby improving the signal-to-noise ratio of the filtered image.

[0038] Color space conversion: Convert the RGB image to the HSV color space, utilize the feature range of red targets in the H, S, and V channels, extract red candidate regions through threshold segmentation, generate a binary mask, and filter out non-red backgrounds (such as coal walls and equipment shells).

[0039] Image enhancement: Contrast-limited adaptive histogram equalization is performed on the red candidate region to improve local contrast; small-area noise in the mask is eliminated through morphological opening and closing operations, and holes in the target region are filled to ensure the continuous and complete outline of the red target.

[0040] (2) Target detection and recognition: Based on a deep learning model, a target detection algorithm is trained. A dedicated dataset is constructed to target the morphological features of the downhole drum and support. The model is then trained and optimized to achieve accurate recognition of the drum blades, end plates, and support edges. The coordinates of the bounding rectangle of the target and the confidence score are output. Targets with low confidence scores are then re-verified to ensure recognition accuracy. Model selection: The network structure was optimized for the downhole scene, the input layer was adapted to the camera resolution (1920×1080), and the robustness of the model to changes in downhole lighting was improved through Mosaic data augmentation (random cropping, scaling, color gamut transformation); the number of convolutional layer parameters was reduced to reduce the amount of computation. The detection head now features a dedicated anchor frame for "roller blades" and "support edges".

[0041] Dataset training: Construct a dataset containing real-world images from underground mines, label the targets as "roller" and "support edge", and divide the dataset into training, validation, and test sets in a 7:2:1 ratio; use transfer learning to initialize model weights (pre-trained COCO dataset).

[0042] Target verification: The detection results are screened for confidence level. Low-confidence targets are verified by combining shape features: the roller target must meet the requirements of a circular outline and red LED light spot distribution, and the support edge must meet the requirements of a straight / broken line outline and a continuous red marking band, excluding interfering targets such as coal blocks and pipelines.

[0043] (3) Distance calculation: Using a monocular vision ranging method, combined with camera calibration parameters (focal length, pixel size, etc.), the shortest distance between the roller and the edge of the bracket obtained by target detection is calculated in actual space; in order to eliminate measurement errors, a filtering algorithm is introduced to smooth the distance data and improve the stability of ranging.

[0044] Camera calibration: The camera intrinsic parameter matrix (focal length f, principal point coordinates cx, cy) and distortion coefficients are obtained in advance through calibration methods and stored in the local control unit; recalibration is performed every 30 days or after the camera is moved to ensure the accuracy of the ranging benchmark.

[0045] Pixel distance conversion: Based on the pixel coordinates of the roller and the support edge obtained from target detection, calculate the shortest pixel distance between them in the image; combine the camera installation parameters (height H above the ground, tilt angle θ) to calculate the actual spatial distance, which serves as the conversion benchmark between pixels and actual size.

[0046] Data smoothing: Kalman filtering is introduced to reduce noise in the ranging results of consecutive frames. The state equation is set as a uniform motion model, and the observation equation is based on the monocular ranging value. The ranging fluctuation caused by vibration and occlusion is reduced through prediction-update iteration.

[0047] (4) Early warning decision-making: Three safety thresholds are preset, which are set according to the operating speed of the coal mining machine and the support spacing in the fully mechanized mining face: The first-level warning threshold triggers a yellow audible and visual warning, alerting operators to pay attention. The level 2 warning threshold triggers an orange audible and visual warning, causing the coal mining machine to reduce its speed to 50% of its rated speed. The Level 3 warning threshold triggers a red audible and visual warning, immediately activating the coal mining machine shutdown protection and simultaneously sending alarm information to the working face control center.

[0048] The ranging results are compared with the safety threshold, and instructions are sent to the coal mining machine control system via the CAN bus; at the same time, the ranging data is stored locally (10 data points per second) for fault backtracking and threshold optimization.

[0049] 4. Early warning and control execution system Early warning device: Audible and visual alarms are installed in the coal mining machine control room and the working face control center. Different warning levels correspond to different frequencies of audible and visual signals (Level 1: low frequency audible and visual, 2s interval; Level 2: medium frequency audible and visual, 1s interval; Level 3: high frequency audible and visual, continuous alarm). At the same time, the relative position and distance data between the drum and the support are displayed in real time on the control room display screen.

[0050] Control execution: The early warning system communicates with the coal mining machine control system via CAN bus. When a level 2 or higher early warning is triggered, it automatically sends a speed reduction or shutdown command to the coal mining machine controller. The system has a manual / automatic switching function, and operators can manually intervene in the control logic according to the actual situation.

[0051] If feasible, this embodiment also includes data storage and backtracking: the algorithm synchronously records real-time monitoring data (including images, distance, and early warning information) and stores it to the downhole data server to support subsequent accident tracing, equipment status analysis, and safety threshold optimization.

[0052] In summary, this embodiment constructs a closed-loop system encompassing identification, data collection, analysis, early warning, and control. It not only enables interference early warning but also directly links the equipment control system to execute emergency shutdown / pause actions. Compared to traditional methods that rely solely on early warning, this significantly improves the effectiveness of interference prevention.

[0053] The above description is merely a preferred embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for preventing interference between a shearer drum and a hydraulic support, characterized in that, The application relates to a red identification method for a coal mining machine and a hydraulic support in a fully-mechanized coal mining face. S1: red identification is added to a coal mining machine drum and a hydraulic support in a fully-mechanized coal mining face, and the red identification comprises a red LED lamp group and red paint; S2: an operation image of the coal mining machine drum and the hydraulic support is collected through a preset monitoring camera in the fully-mechanized coal mining face; a filter is arranged at the front end of a lens of the monitoring camera, and the filter is used for filtering stray light and receiving red light corresponding to the red identification; S3: based on the collected operation image, a minimum distance between the outermost end of the coal mining machine drum and the edge of the hydraulic support is calculated by combining a target detection algorithm, and a multi-stage early warning is executed and a preset control strategy is executed based on a comparison result of the minimum distance and a preset safety threshold.

2. The method for preventing interference between a shearer drum and a hydraulic support according to claim 1, characterized in that, The red LED lamp group is uniformly arranged along the edge of an end disc of the coal mining machine drum in a circumferential direction, and a red highlight contour mark surrounding the drum is formed.

3. The method according to claim 2, characterized in that, When the coal mining machine drum starts to rotate, the red LED lamp group is automatically turned on, and when the coal mining machine drum stops rotating, the red LED lamp group is delayed to be turned off.

4. The method for preventing interference between a shearer drum and a hydraulic support according to claim 1, characterized in that, Red paint is coated at an interference position of the hydraulic support to form a red identification belt.

5. The method for preventing interference between a shearer drum and a hydraulic support according to claim 1, characterized in that, The step S3 specifically comprises the following steps. The collected operation image is subjected to denoising treatment and image enhancement treatment to obtain a pretreated operation image; a red target region of the pretreated operation image is extracted according to a threshold segmentation algorithm, and a candidate region of the coal mining machine drum and the hydraulic support is screened out; the screened image is input into a target detection model to perform target detection, and an edge detection result of the coal mining machine drum and the hydraulic support is output; the target detection model is constructed based on a deep learning model; a pixel distance between the coal mining machine drum and the hydraulic support is determined based on the detection result; and a minimum distance between the coal mining machine drum and the hydraulic support in an actual space is determined based on the pixel distance and camera calibration parameters; three-stage safety threshold values are set, and corresponding grade early warning control strategies are executed based on a comparison result of the minimum distance and the three-stage safety threshold values; the three-stage safety threshold values comprise a first-stage early warning threshold value, a second-stage early warning threshold value and a third-stage early warning threshold value which are gradually decreased.

6. The method according to claim 5, characterized in that, The training process of the target detection model specifically comprises the following steps. training data is obtained, and the training data comprises operation training images of the coal mining machine drum and the hydraulic support and corresponding edge detection results; an initial target detection model is constructed, the training data is input into the initial target detection model to perform target detection, and the initial target detection model is trained to obtain a trained target detection model, with the minimum loss between the initial training result after target detection and the edge detection result corresponding to the operation training image as the target.

7. The method according to claim 5, characterized in that, The corresponding grade early warning control strategies are specifically as follows. when the minimum distance is smaller than the first-stage early warning threshold value, a yellow sound and light early warning is given to prompt an operator to pay attention to observation; when the minimum distance is smaller than the second-stage early warning threshold value, an orange sound and light early warning is given to control the coal mining machine to reduce speed to 50% of a rated speed; when the minimum distance is smaller than the third-stage early warning threshold value, a red sound and light early warning is given to trigger a coal mining machine shutdown protection, and alarm information is sent to a fully-mechanized coal mining face centralized control center.

Images