Method and system for positioning abnormal state of chain of coal mine scraper conveyor

Abstract

The invention relates to the technical field of coal mine intellectualization, and discloses a coal mine scraper conveyor chain abnormal state positioning method and system, and the method comprises the following steps: 1, constructing a scraper conveyor abnormal state chain detection system; 2, carrying out abnormal state chain tracking based on a chain detection and recognition result; step 3, constructing a fault database containing abnormal state chain information; and 4, calculating the real-time position of the abnormal chain by combining the size of the scraper conveyor and the parameters of the encoder. According to the method, feature extraction and detection are achieved through a computer system of a specific model for a to-be-detected coal mine specific scene, accurate positioning of the abnormal state of the chain of the coal mine scraper conveyor in real time is achieved, safe and stable operation of the coal mine scraper conveyor is improved, and intelligent and unmanned construction of an underground coal mine transportation system is promoted.

Description

Technical Field

[0001] This invention belongs to the technical field of point cloud data processing and vibration measurement in computer vision, and particularly relates to a method and system for locating abnormal states of a coal mine scraper conveyor chain. Background Technology

[0002] As a key piece of equipment in fully mechanized mining faces, the operating status of scraper conveyors significantly impacts the safe and efficient production of coal mine transportation systems. The chain, as a major component of the scraper conveyor's sprocket drive system, is susceptible to severe wear and even chain breakage due to factors such as its own materials, human error, and uneven load distribution. Scraper chain failures account for approximately 40% of all scraper conveyor failures, ranging from minor disruptions to production efficiency and even serious threats to personnel safety. Therefore, researching methods for identifying normal and faulty scraper conveyor chain operation is crucial for ensuring the safe and stable operation and maintenance of scraper conveyors.

[0003] Currently, based on different usage methods of scraper conveyor chain operation data, fault identification methods can be divided into traditional fault identification methods and deep learning-based artificial intelligence fault identification methods. Traditional fault identification methods require the deployment of multiple sensors, and threshold settings require a large amount of prior knowledge and subjective intervention from technical personnel, thus presenting problems such as practical implementation difficulties and high costs. Artificial intelligence fault diagnosis methods can extract and identify fault features of coal mine scraper conveyor chains, but still suffer from problems such as large identification delays and increased communication costs due to the deepening of deep learning networks and the amount of data. For example, Chinese patent CN116835228A discloses a scraper conveyor scraper and chain link positioning system and method. When the scraper conveyor is operating normally, the controller encodes and positions all scrapers and chain links of the scraper chain based on the scraper positioning information obtained by the scraper identification system and the scraper and chain link discrimination information obtained by the scraper chain information acquisition and analysis system. When a fault occurs in the scraper or chain of the scraper conveyor, the fault point of the scraper and chain is accurately identified and located.

[0004] Most fault location methods, such as those described above, rely on electronic tags and their corresponding tag readers. These methods are prone to signal obstruction and recognition failures, making them ill-suited for the high-dust, high-vibration, and high-impact environments of underground mines. Furthermore, electronic tags lack semantic-level analysis capabilities for fault locations, and the location mechanism depends on mechanical interval division. Therefore, existing methods struggle to achieve accurate and intelligent fault diagnosis, hindering the automation level of underground equipment operation and maintenance. Summary of the Invention

[0005] To address the problems existing in the aforementioned scraper conveyor chain fault detection methods, and to accurately locate abnormal states of scraper conveyor chains in underground coal mines, this invention proposes a method for locating abnormal states of scraper conveyor chains. This method accurately identifies abnormal chain states based on a lightweight model, tracks abnormal chains within the camera's field of view using the Bytetrack algorithm, constructs an abnormal chain database indexed by ID information, and calculates the chain running speed by combining scraper conveyor parameters and encoder feedback. The real-time position information of the abnormal chain is calculated by indexing the abnormal chain ID and combining it with the chain's real-time speed. The technical solution provided by this invention effectively solves the problems mentioned in the background, ensuring the safe and stable operation of underground coal mine transportation systems, providing new technical support for detecting abnormal states of scraper conveyor chains, and is of great significance for the safety, efficiency, and intelligence of underground coal mine transportation systems.

[0006] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows: In a first aspect, this application provides a method for locating abnormal conditions of a scraper conveyor chain in a coal mine, the method comprising the following steps: Step 1: Construct a chain detection system for coal mine scraper conveyors, using a lightweight model to detect and identify chains in abnormal states; Step 2: Within the field of view of the chain detection system, the algorithm is used to track abnormal chains in real time. After obtaining information about the abnormal chains, IDs are assigned to multiple abnormal chains. Step 3: Construct a database of multiple types of abnormal state chains using IDs as indexes; Step 4: Calculate the real-time position information of the abnormal state chain by indexing the abnormal state chain ID and combining the chain's real-time movement speed.

[0007] According to one embodiment of this application, the real-time position information calculation process of the abnormal state chain in step four includes: Step 401: Extract relevant data of the abnormal state chain from the database based on the chain's ID; Step 402: Based on the real-time feedback of the sprocket speed from the encoder of the scraper conveyor, and combined with the parameter information of the scraper conveyor, calculate the chain speed in real time. Step 403: Integrate the abnormal state chain database information and the real-time linear velocity of the chain's movement direction to establish an abnormal state chain real-time position calculation model that includes multiple faults and calculate the accurate position of the abnormal state chain.

[0008] According to one embodiment of this application, the process of establishing a real-time position calculation model for the abnormal state chain and calculating the accurate position of the abnormal state chain in step 403 includes: Step 4041: During the process of identifying and continuously tracking the abnormal state chain, the system assigns a unique fixed ID to each abnormal chain and updates and saves its trajectory information at a set time interval; when the chain passes through the center line of the camera's field of view, record this moment as ; meanwhile, define the distance from the initial position of the chain along the running direction to the upper and lower chain turning points as d0, the length of the chain as L, and the system synchronously reads the real-time linear velocity v of the scraper conveyor speed encoder; Step 4042: Calculate the relationship between the movement duration t of the current abnormal state chain starting from the initial position and the period T required for the scraper conveyor chain to run one week. If t < T, it indicates that the chain has not completed a full cycle, and the path displacement S of the chain along the running direction from the initial position can be calculated by Equation (1): (Equation 1); If t > T, then adopt a periodic positioning strategy and substitute it into Equation (2) to calculate its equivalent path displacement relative to the initial position: (Equation 2); Step 4043: By comparing the currently calculated displacement S with the distance d0, determine the chain segment where the faulty link is located: If S < d0, it means that the abnormal link has not reached the turning point and is located in the upper chain segment. Calculate the distance d between the link and the head of the scraper conveyor according to Equation (3): (Equation 3); If d0 < S < L / 2 + d0, it means that the abnormal link has reached the turning point and has reached the lower chain segment. Calculate the distance d between the link and the head of the scraper conveyor according to Equation (4): (Equation 4); Otherwise, it can be determined that the abnormal link has returned to the upper chain segment, and the distance d between the link and the head of the scraper conveyor is calculated according to Equation (5): (Equation 5).

[0009] According to an embodiment of the present application, the construction process of the chain detection system for the coal mine scraper conveyor in the first step includes: Step 101: Determine the installation position of the explosion-proof camera according to the external dimension parameters of the coal mine scraper conveyor and in combination with the encoder installation position; Step 102: Determine the field of view range of the explosion-proof camera on the scraper conveyor and determine the projection point of the center of the field of view as the initial node of the detection system.

[0010] According to an embodiment of the present application, the process of determining the installation position of the explosion-proof camera in Step 101 includes: Step 1011: Determine the installation location information of the explosion-proof camera, including the installation height h and the angle θ with the plane of the scraper conveyor; Step 1012: Determine the camera's field of view based on the camera's installation height and angle, defining the direction along the scraper conveyor as the x-axis and the direction perpendicular to the scraper conveyor as the y-axis.

[0011] According to one embodiment of this application, the process of determining the initial node of the detection system in step 102 includes: Step 1021: Establish the spatial relationship between the camera's field of view size and the actual location area of ​​the scraper conveyor; Step 1022: Using the midpoint of the x-axis and y-axis in the field of view as the origin of the field of view, determine the pixel coordinates of the origin of the field of view, and obtain the initial node position coordinates on the actual scraper conveyor through spatial transformation.

[0012] According to one embodiment of this application, step two, which involves real-time tracking and ID allocation of multiple abnormal state chains using an algorithm, includes: Step 2011: Obtain the detection results, including the location, category, and confidence level of the chain target, and input the above information into the Betytrack tracking algorithm; Step 2012: Use a Kalman filter to predict the position of the existing trajectory in the current frame, and calculate the IOU matching degree between the detection box and the predicted box; Step 2013: Use the Hungarian algorithm to match the best trajectory and determine the target ID.

[0013] According to one embodiment of this application, in the process of constructing the database of multiple types of abnormal states in step three, the relevant data of the abnormal state chain is constructed using ID as an index, and the data storage period is divided based on the initial detection node.

[0014] According to one embodiment of this application, the abnormal state chain database construction process in step three includes: Step 3011: When an abnormal chain is detected, the chain database automatically synchronizes the detected chain information using the ID as an index. In the dynamic database, the time information and fault type information experienced after the chain is identified are recorded in real time. Step 3012: Based on the actual dimensions of the scraper conveyor and the sprocket speed fed back by the encoder, dynamically estimate the cycle T of one revolution of the chain, and use this as the update cycle of the abnormal state chain database.

[0015] Secondly, this application provides a coal mine scraper conveyor chain abnormality detection system, used to implement the coal mine scraper conveyor chain abnormality location method provided in the first aspect, the system comprising: The detection and identification unit is used to build a coal mine scraper conveyor chain detection system, which uses a lightweight model to detect and identify abnormal chain conditions. The real-time tracking unit is used within the field of view of the chain detection system to track abnormal chains in real time using algorithms. After obtaining information about abnormal chains, it assigns IDs to multiple abnormal chains. Database building unit, used to construct a database of various types of abnormal state chains using ID as an index; The position calculation unit calculates the real-time position information of the abnormal state chain by indexing the abnormal state chain ID and combining the real-time movement speed of the chain.

[0016] Compared with the prior art, the main advantages of the present invention are:

[0017] 1. This invention employs a lightweight visual model (such as RT-DETR) to directly perform image recognition on the chain. This approach does not rely on easily obscured physical tags and can robustly determine the abnormal state of the chain solely through visual features, fundamentally improving the system's adaptability to harsh environments with high dust and strong vibrations in coal mines.

[0018] 2. This invention addresses the difficulty in locating multiple abnormal chain states during the inspection of scraper conveyors in traditional coal mines. Based on the Betytrack algorithm, it tracks the identified abnormal chain states within the field of view and assigns them IDs, constructing an abnormal chain state database indexed by IDs. Combined with encoder information, it achieves accurate location of multiple abnormal chain states, providing important reference information for the maintenance of scraper conveyors in coal mines.

[0019] 3. Based on the actual size of the scraper conveyor and the sprocket speed fed back by the encoder, this invention dynamically estimates the cycle T of one revolution of the chain, and uses this as the update cycle of the abnormal state chain database, which significantly reduces the computational load of the algorithm system and achieves the effect of avoiding high-frequency polling and redundant calculation.

[0020] 4. The experimental results analysis of the present invention at two time points of chain operation show that the calculated position error of the positioning method on the scraper conveyor chain is within 7%, and the overall error of less than 10% can meet the requirements for determining the position range of the scraper conveyor chain under abnormal conditions. Attached Figure Description

[0021] Figure 1 A schematic diagram of the overall scheme for fault tracking and location of scraper conveyor chain; Figure 2 A flowchart illustrating the algorithm framework for tracking the chain in abnormal states of a scraper conveyor within the field of view; Figure 3The following are the results of tracking the abnormal state chain under different environmental conditions: (a) tracking the abnormal state chain under normal conditions; (b) tracking the abnormal state chain under smoke and dust conditions; and (c) tracking the abnormal state chain under uneven lighting conditions. Figure 4 A diagram illustrating the effect of a database with multiple types of abnormal state chains; Figure 5 Flowchart for chain positioning in abnormal conditions of scraper conveyor; Figure 6 This is an application effect diagram of the chain detection system for abnormal conditions of scraper conveyors. Detailed Implementation

[0022] To more clearly demonstrate how to achieve the objectives described in this invention, the technical solutions of this invention will be clearly and thoroughly described below in conjunction with the accompanying drawings and specific embodiments. The following examples will help those skilled in the art to further understand this invention, but do not limit the invention in any way. It should be noted that those skilled in the art can make several modifications and improvements without departing from the concept of this invention, and these all fall within the protection scope of this invention.

[0023] Before describing this application, the working process of the scraper conveyor is summarized as follows: The scraper conveyor drives the sprocket to rotate through the drive device. The sprocket meshes with the chain, pulling the scraper to circulate in the trough, thereby promoting the material conveying; the encoder installed on the sprocket shaft provides real-time feedback of speed and position signals, which is used to accurately calculate the chain running speed and rotation cycle, and realize the synchronization of the operation status monitoring and control system.

[0024] Continuing from the above description of the process of this application: A method for locating abnormal states of a coal mine scraper conveyor chain, comprising: Reference Figure 1 The diagram shown illustrates the overall scheme for fault tracking and location of a scraper conveyor chain according to an embodiment of this application. The steps of this method are described below: Step 1: Construction of a coal mine scraper conveyor chain detection system. A lightweight model is used to detect and identify abnormal chain conditions. The preferred lightweight model is the lightweight real-time end-to-end detection model RT-DETR. Specifically, this includes: Step 101: Determine the installation location of the explosion-proof camera based on the external dimensions of the coal mine scraper conveyor and the encoder installation location.

[0025] Step 102: Determine the field of view of the explosion-proof camera on the scraper conveyor, and determine the projection point of the center of the field of view as the initial node of the detection system.

[0026] In step 101, the installation location of the explosion-proof camera is determined: Step 1011: The explosion-proof camera is installed at the end of the scraper conveyor, specifically involving the camera's installation height h and the angle θ between the camera and the plane of the scraper conveyor.

[0027] Step 1012: Based on the camera's installation height and shooting angle, determine the camera's field of view, which is the distance along the x-axis of the scraper conveyor and the distance along the y-axis perpendicular to the scraper conveyor. Its field of view should cover the entire chain running plane.

[0028] In step 102, the initial node of the detection system is determined: Step 1021: Construct the spatial relationship between the camera's field of view size and the actual location area of ​​the scraper conveyor.

[0029] Step 1022: Using the midpoint of the x-axis and y-axis in the viewport as the origin, determine the pixel coordinates of the viewport origin. Through spatial transformation, obtain the initial node position coordinates located on the actual scraper conveyor.

[0030] Step 2: Real-time tracking of abnormal state chains within the field of view: Reference Figure 2 The diagram shown illustrates the framework of an algorithm for tracking the chain in an abnormal state of a scraper conveyor within the field of view, as provided in this embodiment of the application. The steps are described below: Step 201: Based on the intelligent recognition algorithm, the intelligent algorithm is used to track the identified abnormal state chains in real time, obtain relevant information of the abnormal state chains within the field of view, and assign corresponding IDs to the multiple abnormal state chains within the field of view.

[0031] In step 201, the intelligent algorithm tracks multiple abnormal state chains in real time: Step 2011: Obtain the results of the recognition algorithm, including the location, category, and confidence level of the chain target, and pass the above information into the Bytetrack tracking algorithm.

[0032] Step 2012: Use a Kalman filter to predict the position of the existing trajectory in the current frame, and calculate the IOU matching degree between the detection box and the predicted box.

[0033] Step 2013: Use the Hungarian algorithm to match the best trajectory and determine the target ID.

[0034] Reference Figure 3 The image shows the effect of the tracking and recognition algorithm under different environments.

[0035] Step 3: Constructing a database of multiple types of abnormal state chains: Reference Figure 4 The diagram shown illustrates the effect of a database of various abnormal state chains provided in this application embodiment. The steps are described below: Step 301: Construct relevant data for the abnormal state chain using ID as the index, and divide the data storage period based on the initial detection node.

[0036] In step 301, the abnormal state chain database is constructed:

[0037] Step 3011: When an abnormal chain is identified, the chain database automatically synchronizes the detected chain information using the ID as an index. In the dynamic database, information such as the time and fault type experienced after the chain is identified is recorded in real time.

[0038] Step 3012: Based on the actual dimensions of the scraper conveyor and the sprocket speed fed back by the encoder, dynamically estimate the cycle T of one revolution of the chain, and use this as the update cycle of the abnormal state chain database.

[0039] Step 4: Real-time position calculation of the abnormal state chain

[0040] Reference Figure 5 The diagram shown illustrates the chain positioning process for an abnormal state of a scraper conveyor, as described in this embodiment of the application. The steps are as follows:

[0041] Step 401: Extract relevant data of the abnormal state chain from the database based on the chain ID.

[0042] Step 402: Based on the real-time feedback of the sprocket speed from the encoder of the scraper conveyor, and combined with the parameter information of the scraper conveyor, calculate the chain speed in real time.

[0043] Step 403: Integrate the abnormal state chain database information and the real-time linear velocity of the movement direction to establish an abnormal state chain real-time position calculation model that includes various faults such as chain breakage and wear, and calculate the accurate position of the abnormal state chain.

[0044] In step 403, the accurate location of the abnormal state chain is calculated as follows:

[0045] Step 4041: During the identification and continuous tracking of abnormal chains, the system assigns a unique fixed ID to each abnormal chain and updates and saves its trajectory information at set time intervals. When the chain passes the center line of the camera's field of view, that moment is recorded as... Simultaneously, the distance from the initial position of the chain (i.e., the position first detected) along the chain running direction to the upper and lower chain reversal point (i.e., the sprocket turning position) is defined as d0, and the chain length is L. At the same time, the system synchronously reads the real-time linear velocity v of the scraper conveyor speed encoder.

[0046] Step 4042: Calculate the relationship between the movement duration t of the current abnormal state chain from the initial position and the period T required for the scraper conveyor chain to run one week. If t < T, it indicates that the chain has not completed a full cycle, and the path displacement S from the initial position along the running direction can be calculated by Equation (1): (Equation 1); If t > T, then adopt a periodic positioning strategy and substitute it into Equation (2) to calculate its equivalent path displacement relative to the initial position: (Equation 2); Step 4043: By comparing the currently calculated displacement S with the initial reference distance d0 (i.e., the path length from the initial detection position to the upper and lower chain turning points), determine the chain segment where the faulty link is located: If S < d0, it means that the abnormal link has not reached the turning point and is located in the upper chain segment. At this time, calculate the distance d between the link and the head of the scraper conveyor according to Equation (3): (Equation 3); If d0 < S < L / 2 + d0, it means that the abnormal link has reached the turning point and reached the lower chain segment. At this time, calculate the distance d between the link and the head of the scraper conveyor according to Equation (4): (Equation 4); Otherwise, it can be determined that the abnormal link returns to the upper chain segment, and the distance d between the link and the head of the scraper conveyor can be calculated according to Equation (5): (Equation 5); Refer to Figure 6 As shown, it is the application effect diagram of the scraper conveyor abnormal state chain detection system provided by the embodiment of the present application.

[0047] In order to verify the technical effects that the present invention can produce, tests on the method of the present invention were carried out.

[0048] To verify the effectiveness of the vision- and encoder-based fault tracking and localization method for scraper conveyor chains, a tracking and localization experiment was conducted on a scraper conveyor in the laboratory, with a camera mounted on it, and the results were analyzed. The system mainly included a camera, a scraper conveyor, and a built-in encoder. The distance from the camera's installation position to the chain's direction of movement was L, the total chain length was 2x, and the total time required for one complete chain rotation was 94 seconds. To verify the effectiveness of the position calculation method at different time intervals, and considering the dimensions of the scraper conveyor and the requirement that the final position be at the top for easy measurement of the actual position, three sets of experiments were conducted with stop times of 16s, 110s, 204s, 91s, 185s, and 279s. Three time sets (16s, 110s, 204s) show the positioning position as above, located at the initial positioning point along the chain direction, with 1, 2, and 3 loops respectively. Three time sets (91s, 185s, 279s) show the positioning position as above, located at the initial positioning point against the chain direction, with 1, 2, and 3 loops respectively. These results validated the position outputs of methods ② and ③ in the positioning method. Based on the experimental results from the laboratory at two time points during chain operation, the calculated position error of the positioning method on the scraper conveyor chain is within 7%. Overall, an error of less than 10% is sufficient for determining the chain position range in abnormal states of the scraper conveyor.

[0049] In summary, the method of the present invention is simple in steps, reasonable in design, and has good point identification and detection effect. By monitoring the working status of the coal mine scraper conveyor chain in real time, the health status of the chain can be grasped in a timely manner, and targeted maintenance measures can be taken to ensure the safety and service life of the chain, and provide reliable technical support for the safe operation of the coal mine transportation system.

[0050] The above description is merely a preferred embodiment of the present invention and does not constitute any limitation on the present invention. Any simple modifications, alterations, or equivalent structural changes made to the above embodiments based on the technical essence of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A method for locating abnormal conditions of a scraper conveyor chain in a coal mine, characterized in that, The method includes the following steps: Step 1, construct a chain detection system for a coal mine scraper conveyor, and detect and identify abnormal state chains with a lightweight model; Step 2, within the visual field range of the chain detection system, use a positioning algorithm to track the abnormal state chains in real time. After obtaining the information of the abnormal state chains, assign IDs to multiple abnormal state chains; Step 3, construct a database of various types of abnormal state chains with the ID as the index; Step 4, by indexing the ID of the abnormal state chain and combining the real-time movement speed of the chain, calculate the real-time position information of the abnormal state chain.

2. The method for locating abnormal states of a coal mine scraper conveyor chain according to claim 1, characterized in that, The process of calculating the real-time position information of the abnormal state chain in Step 4 includes: Step 401, extract the relevant data of the chain from the database according to the ID of the abnormal state chain; Step 402, based on the real-time feedback of the sprocket wheel speed by the encoder of the scraper conveyor and combined with the parameter information of the scraper conveyor, calculate the real-time movement speed of the chain in real time; Step 403, comprehensively consider the information in the abnormal state chain database and the real-time linear speed of the chain movement direction, establish a real-time position calculation model for the abnormal state chain containing multiple faults, and calculate the accurate position of the abnormal state chain.

3. The method for locating abnormal states of a coal mine scraper conveyor chain according to claim 2, characterized in that, The process of establishing a real-time position calculation model for the abnormal state chain in Step 403 to calculate the accurate position of the abnormal state chain includes: Step 4041: During the identification and continuous tracking of abnormal state chains, the system assigns a unique fixed ID to each abnormal chain and updates and saves its trajectory information at set time intervals; when the chain passes the center line of the camera's field of view, that moment is recorded as... Meanwhile, the distance from the initial position of the chain along the chain running direction to the reversal point of the upper and lower chains is defined as d0, the chain length is L, and the system synchronously reads the real-time linear velocity of the scraper conveyor speed encoder as v. Step 4042, calculate the relationship between the movement duration t of the current abnormal state chain starting from the initial position and the period T required for the scraper conveyor chain to run one week. If t < T, it indicates that the chain has not completed a complete cycle, and the path displacement S of it along the running direction from the initial position can be calculated by Equation (1): (Formula 1); If t > T, then adopt a periodic positioning strategy and substitute it into Equation (2) to calculate its equivalent path displacement relative to the initial position: (Equation 2); Step 4043, by comparing the currently calculated displacement S with the distance d0, judge the chain segment where the faulty link is located: If S < d0, it means that the abnormal link has not reached the turning point and is located in the upper chain segment. Calculate the distance d between the link and the head of the scraper conveyor according to Equation (3): (Equation 3); If d0 < S < L / 2 + d0, it means that the abnormal link has reached the turning point and reached the lower chain segment. Calculate the distance d between the link and the head of the scraper conveyor according to Equation (4): (Equation 4); Otherwise, it can be determined that the abnormal link returns to the upper chain segment, and the distance d between the link and the head of the scraper conveyor can be calculated according to Equation (5): (Equation 5).

4. The method for locating abnormal states of a coal mine scraper conveyor chain according to claim 1, characterized in that, The process of constructing the chain detection system for the coal mine scraper conveyor in Step 1 includes: Step 101, according to the external dimension parameters of the coal mine scraper conveyor and combined with the installation position of the encoder, determine the installation position of the explosion-proof camera; Step 102, determine the visual field range of the explosion-proof camera on the scraper conveyor, and determine the projection point of the visual field center as the initial node of the detection system.

5. The method for locating abnormal states of a coal mine scraper conveyor chain according to claim 4, characterized in that, The process of determining the installation position of the explosion-proof camera in Step 101 includes: Step 1011: Determine the installation position information of the explosion-proof camera, and the installation position information includes the installation height h and the angle θ with the plane of the scraper conveyor; Step 1012: According to the installation height and angle of the camera, determine the visual field range of the camera, and define the direction along the scraper conveyor as the x-axis and the direction perpendicular to the scraper conveyor as the y-axis.

6. The method for locating abnormal states of a coal mine scraper conveyor chain according to claim 4, characterized in that, The process of determining the initial node of the detection system in Step 102 includes: Step 1021: Establish the spatial relationship between the camera's field of view size and the actual location area of ​​the scraper conveyor; Step 1022: Using the midpoint of the x-axis and y-axis in the field of view as the origin of the field of view, determine the pixel coordinates of the origin of the field of view, and obtain the initial node position coordinates on the actual scraper conveyor through spatial transformation.

7. The method for locating abnormal states of a coal mine scraper conveyor chain according to claim 1, characterized in that, Step two, which uses an algorithm to track multiple abnormal state chains in real time and perform ID allocation, includes: Step 2011: Obtain the detection results, including the location, category, and confidence level of the chain target, and input the above information into the Betytrack tracking algorithm; Step 2012: Use a Kalman filter to predict the position of the existing trajectory in the current frame, and calculate the IOU matching degree between the detection box and the predicted box; Step 2013: Use the Hungarian algorithm to match the best trajectory and determine the target ID.

8. The method for locating abnormal states of a coal mine scraper conveyor chain according to claim 1, characterized in that, In step three, during the construction of the database of multiple types of abnormal state chains, the relevant data of the abnormal state chain is constructed using ID as an index, and the data storage period is divided based on the initial detection node.

9. A method for locating abnormal states of a coal mine scraper conveyor chain according to claim 1, characterized in that, The abnormal state chain database construction process in step three includes: Step 3011: When an abnormal chain is detected, the chain database automatically synchronizes the detected chain information using the ID as an index. In the dynamic database, the time information and fault type information experienced after the chain is identified are recorded in real time. Step 3012: Based on the actual dimensions of the scraper conveyor and the sprocket speed fed back by the encoder, dynamically estimate the cycle T of one revolution of the chain, and use this as the update cycle of the abnormal state chain database.

10. A coal mine scraper conveyor chain abnormality detection system, used to implement the coal mine scraper conveyor chain abnormality positioning method according to any one of claims 1-9, characterized in that... The system includes: The detection and identification unit is used to build a coal mine scraper conveyor chain detection system, which uses a lightweight model to detect and identify abnormal chain conditions. The real-time tracking unit is used within the field of view of the chain detection system to track abnormal chains in real time using algorithms. After obtaining information about abnormal chains, it assigns IDs to multiple abnormal chains. Database building unit, used to construct a database of various types of abnormal state chains using ID as an index; The position calculation unit calculates the real-time position information of the abnormal state chain by indexing the abnormal state chain ID and combining the real-time movement speed of the chain.

Images