Coal face hydraulic support fault diagnosis method based on laser three-dimensional scanning technology

Abstract

The invention aims to provide a coal face hydraulic support fault diagnosis method based on a laser three-dimensional scanning technology, and belongs to the technical field of coal mining. Three-dimensional point cloud data of the surface of a hydraulic support are obtained through the laser three-dimensional scanning technology, and the average curvature and the normal vector of the point cloud data are calculated; comprehensively extracting characterization information of the hydraulic support; and in combination with the distribution condition of the point cloud data, calculating an abnormal score, locking the position of an abnormal region and an abnormal range, extracting quantitative information and three-dimensional representation information for fitting, judging the severity of a fault, and realizing fault diagnosis. Simulation experiment results show that when fault diagnosis is carried out on the hydraulic support, the severity of different faults can be effectively judged, tiny faults can be effectively recognized, and the proportion of correct diagnosis samples is obviously higher, so that the complexity of characterization information of the hydraulic support is reduced, and the fault diagnosis effect of the hydraulic support on the coal face is improved.

Description

Technical Field

[0001] The invention belongs to the technical field of coal mining, and in particular relates to a fault diagnosis method for hydraulic supports of a coal mining working face based on laser three-dimensional scanning technology. Background Art

[0002] As an important support equipment for coal mining working face, the safe and stable operation of hydraulic support plays a vital role in ensuring mine production safety and improving coal mining efficiency. However, hydraulic support often has various faults in the complex and changeable underground working environment, such as hydraulic leakage, structural deformation, etc. These faults not only affect the normal operation of the support, but also cause certain safety hazards. Therefore, how to accurately and quickly diagnose the fault of hydraulic support has become an important issue facing the current coal industry. In this regard, relevant research has been conducted to explore the fault diagnosis of hydraulic supports. For example: Zhong Jinbao, Qi Jingshuai, Meng Xiangyi, Research on Fault Diagnosis Simulation of Hydraulic Support Push Control Circuit Based on PSO-BP Neural Network [J]. Mining Machinery, 2023, 51(11): 11-15, particle swarm optimization algorithm (PSO) is used to optimize the parameters of BP neural network to realize the fault diagnosis of hydraulic support push control circuit. However, the disadvantage of this method is that the data is mainly obtained by simulating the historical data of the loop, and the real-time monitoring data of the support is not obtained, which affects the accuracy and reliability of subsequent diagnosis. Zhong Jinbao, Qi Jingshuai, Li Lei, et al., Research on Fault Simulation of Hydraulic Support Column Control Circuit Based on Particle Swarm Algorithm Optimization of BP Neural Network [J], Engineering Machinery, 2023, 54 (06): 75-82+10, a fault diagnosis model is constructed by particle swarm algorithm to improve the generalization ability and convergence speed of the network. Due to the high complexity of the characterization information of the hydraulic support in the coal mining face, this method does not diagnose subtle faults such as support top beam bending and support wear, and the diagnostic effect is insufficient. WANG R, SUN W. Fault diagnosis of electrical automatic control system of hydraulic support based on particle swarm optimization algorithm[J]. Journal of Ambient Intelligenceand Humanized Computing, 2023, 14(9): 12091-12097, The parameters of the fault diagnosis model are optimized by particle swarm algorithm to achieve classification diagnosis and identification of different fault types. However, this method mainly diagnoses the internal operation faults of the hydraulic support, and it is difficult to accurately diagnose the surface wear of the support. HUANG K, WU S, LI F, et al. Fault diagnosis of hydraulic systems based on deep learning model with multirate data samples[J]. IEEE Transactions on neural networks and learning systems, 2021, 33(11): 6789-6801, A deep learning model is used to diagnose the faults of the hydraulic system, and the deep learning model is trained with multirate data samples to improve the accuracy and robustness of fault diagnosis. However, this method mainly diagnoses the hydraulic system as a whole, and the diagnostic effect of subtle faults in the three-dimensional characterization of the bracket needs to be improved.

[0003] According to the analysis of the above research results, it is difficult to effectively extract the three-dimensional representation information of the surface fault of the hydraulic support by data analysis alone, and it is impossible to accurately diagnose subtle faults. However, laser three-dimensional scanning technology, due to its high precision and non-contact characteristics, has shown its application potential in the effective extraction of three-dimensional representation information of defect detection, and has gradually become a hot topic of research. Regarding the application of laser three-dimensional scanning technology in defect detection, Liu Yongzhi, Zhang Zhouqiang, Guo Zhongchao, et al., Research on three-dimensional surface detection system of parts based on line laser scanning [J]. Foreign Electronic Measurement Technology, 2021, 40(2):67-72, Surface defect detection of parts using three-dimensional line laser scanning technology. Defect detection based on three-dimensional point cloud data of part surface. CHENG W, MA Y, HUANG W. Simultaneous compositional and defect measurementsusing laser opto-ultrasonic dual detection for laser 3D printing and welding [J]. 2021, 11677(3):5-9, Composite detection of welding sample defect information using laser ultrasonic technology. Detection of sample surface defects based on emission spectrum and ultrasonic time domain signal.

[0004] However, the above methods use laser scanning technology to monitor surface defects. While expanding the characterization range of hydraulic supports, it also increases the complexity of information and affects the surface fault detection effect of hydraulic supports. Summary of the invention

[0005] In view of the above problems, the present invention proposes a fault diagnosis method for hydraulic supports in coal mining working faces based on laser three-dimensional scanning technology. On the basis of applying laser scanning technology to collect information, an innovative design of the three-dimensional characterization information extraction process of the hydraulic supports is used to analyze the surface smoothness of the hydraulic supports, reduce the interference of information complexity, and accurately reflect its working status. In addition, the point cloud data is combined to analyze the distribution of abnormal areas, capture more subtle fault information, determine the specific location and range of the fault, and improve the accuracy of fault diagnosis.

[0006] The present invention adopts the following technical solution: A method for diagnosing hydraulic support faults in a coal mining face based on laser three-dimensional scanning technology comprises the following steps: The first step is to obtain the 3D point cloud data of the hydraulic support surface through laser 3D scanning technology; The second step is to calculate the average curvature and normal vector of the point cloud data and fully extract the hydraulic support characterization information; The third step is to calculate the anomaly score based on the distribution of point cloud data, and lock the location and range of the abnormal area; The fourth step is to fit the quantitative information with the three-dimensional representation information extraction to determine the severity of the fault and realize fault diagnosis.

[0007] Furthermore, the method for acquiring the three-dimensional point cloud data of the hydraulic support surface in the first step is as follows: in the coal mining face, a relatively stable area is selected as the scanning environment. Ensure that there is no interference from debris in the scanning area and the hydraulic support is clearly visible, place the hydraulic support in the predetermined scanning area, and ensure that it is in a stable working state; Select markers or positioning devices to help the scanner accurately identify the position and posture of the hydraulic support. The laser scanner emits a laser beam to the surface of the hydraulic support and receives the reflected light signal. Combined with the time difference data between the emission and reception of the laser beam, the three-dimensional point cloud of the surface of the hydraulic support is confirmed. The calculation formula is as follows: , where P(x, y, z) represents the three-dimensional coordinates of a point on the surface of the hydraulic support, O represents the origin coordinates of the three-dimensional laser scanner, Δt represents the time difference between the emission and reception of the laser beam in the three-dimensional laser scanning device, and V represents the direction vector of the laser beam; The surface data points of the hydraulic support are continuously collected until the entire scanning process is completed. The three-dimensional coordinate confirmation results of the hydraulic support surface are presented in the form of a three-dimensional image, generating three-dimensional point cloud data to intuitively display the geometric structure and deformation of the support.

[0008] Furthermore, the average curvature describes the average degree of curvature at a point on the surface. For each point P (x, y, z) in the point cloud data, its average curvature can be obtained by calculating the curvature of all points in its neighborhood and taking the average value. The specific calculation formula of the average curvature k is as follows: , where n represents the total number of coordinate nodes in the neighborhood, k i Represents the curvature corresponding to each coordinate node, and R1 and R2 represent the two principal curvature radii at point p respectively.

[0009] Furthermore, the normal vector of the point cloud data is calculated as follows: for each point P (x, y, z) in the point cloud data, its normal vector N (x n ,y n , z n ) is obtained by calculating the covariance matrix of the points in its neighborhood and obtaining its eigenvector. The specific expression of the covariance matrix is ​​as follows: , where P i represents a random point in the neighborhood of the 3D point cloud data of the hydraulic support, P avg represents the mean value of the point cloud in the neighborhood, and the vector corresponding to the minimum eigenvalue is the normal vector N(x n ,y n , zn ), after calculating the normal vector of each point, set the angle threshold between the normal vectors of adjacent points to N φ ,If the angle is greater than a certain threshold, it is considered that there may be defects in the area.

[0010] Furthermore, the determination method of the abnormal area and abnormal range is as follows: project a laser beam onto the working surface, the focus of the laser beam corresponds to the center position of the hydraulic support to be measured, and mark the abnormal point cloud data by the density and abnormality of the point cloud data. Assume that the mean and standard deviation of the hydraulic support point cloud data set are μ i and σ i , calculate the anomaly score for each region: , where f represents the characteristic value of the hydraulic support characterization information, and a basic threshold is set. If z i The value exceeds the fluctuation threshold z under normal operating conditions φ , it means that the point cloud data of this part is abnormal, and the area is judged as an abnormal area. At the same time, quantitative information such as wear depth and area is obtained.

[0011] Furthermore, the method for determining the severity of the fault is as follows: while the laser beam is projected onto the work surface, a visual marking tool is used to mark the area near the abnormal area, the location of the abnormal area and the range of the abnormal area are locked, and from the point cloud data that has been marked as abnormal, quantitative information is extracted and fitted with three-dimensional characterization information to extract the coordinate information of the abnormal area. The extracted coordinates of the abnormal area are combined with the feature threshold judgment result to clarify the severity of the fault.

[0012] The beneficial effects of the present invention are as follows: By using the method of the present invention, when diagnosing faults of hydraulic supports, the severity of different faults can be effectively determined, subtle faults can be effectively identified, and the proportion of correctly diagnosed samples is significantly higher, thereby reducing the complexity of the hydraulic support characterization information and improving the fault diagnosis effect of the hydraulic support in the coal mining working face.

[0013] Based on the application of laser scanning technology to collect information, the present invention innovatively designs the process of extracting three-dimensional characterization information of hydraulic supports to analyze the surface smoothness of hydraulic supports, reduce the interference of information complexity, and accurately reflect its working status. In addition, the point cloud data is combined to analyze the distribution of abnormal areas, capture more subtle fault information, determine the specific location and range of the fault, and improve the accuracy of fault diagnosis. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 This is a flow chart of fault diagnosis of hydraulic supports in coal mining working faces based on laser three-dimensional scanning technology of the present invention.

[0015] Figure 2 This is the experimental scene of coal mining face.

[0016] Figure 3 This is a point cloud image scan of the hydraulic support according to an embodiment of the present invention.

[0017] Figure 4 This is the distribution of laser scanning data of the hydraulic support of the present invention.

[0018] Figure 5 This is a comparison chart of fault visualization recognition effects.

[0019] Figure 6 This is the percentage of samples with correct diagnosis. DETAILED DESCRIPTION

[0020] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0021] The fault diagnosis method of hydraulic support fault in coal mining working face is as follows: 1. Acquisition of 3D point cloud data of hydraulic support surface In the coal mining face, select a relatively stable area as the scanning environment. Make sure there is no interference from debris in the scanning area and the hydraulic support is clearly visible. Place the hydraulic support in the predetermined scanning area and ensure that it is in a stable working state.

[0022] Select markers or positioning devices to help the scanner accurately identify the position and posture of the hydraulic support. The laser scanner emits a laser beam to the surface of the hydraulic support and receives the reflected light signal. Combined with the time difference data of the laser beam emission and reception, the three-dimensional point cloud of the surface of the hydraulic support can be confirmed: the calculation formula is as follows: , where P(x, y, z) represents the three-dimensional coordinates of a point on the surface of the hydraulic support, O represents the origin coordinates of the three-dimensional laser scanner, Δt represents the time difference between the emission and reception of the laser beam in the three-dimensional laser scanning device, and V represents the direction vector of the laser beam.

[0023] The surface data points of the hydraulic support are continuously collected until the entire scanning process is completed. The three-dimensional coordinate confirmation results of the hydraulic support surface can be presented in the form of a three-dimensional image, generating three-dimensional point cloud data to intuitively display the geometric structure and deformation of the support.

[0024] 2. 3D characterization information extraction and defect probability determination As a long-term monitoring method, laser scanning technology expands the characterization range of hydraulic supports, but also increases the complexity of information. In this regard, a three-dimensional characterization information extraction process of hydraulic supports is designed to analyze the surface smoothness of hydraulic supports and reflect their working status and health status. In order to reduce the interference of information complexity and accurately reflect the working status of the support, the average curvature and normal vector of the three-dimensional coordinate data of the hydraulic support are calculated to evaluate the smoothness and deformation of the support. The average curvature (K) describes the average degree of curvature at a point on the surface. For each point P (x, y, z) in the point cloud data, its average curvature can be obtained by calculating the curvature of all points in its neighborhood and taking the average value. The specific calculation formula of the average curvature k is as follows: , where n represents the total number of coordinate nodes in the neighborhood, k i Represents the curvature corresponding to each coordinate node, and R1 and R2 represent the two principal curvature radii at point p respectively.

[0025] When an abnormal area appears on the surface of the bracket, it often means that some parts of the bracket are faulty or have potential problems. The abnormal area may also be a leak or failure point of some components in the hydraulic system, such as wear of the seals of the hydraulic cylinder, blockage of the hydraulic valve, etc. In order to detect the wear and surface quality of the bracket, it is also necessary to identify the surface defects of the bracket. Surface defects may appear as abnormal geometric shapes or texture changes. Therefore, the normal vector of the point cloud data is calculated, and its degree of change is analyzed to identify potential defects. For each point P (x, y, z) in the point cloud data, its normal vector N (x n ,y n , z n ) can be obtained by calculating the covariance matrix of the points in its neighborhood and obtaining its eigenvector. The specific expression of the covariance matrix is ​​as follows: , where P i represents a random point in the neighborhood of the 3D point cloud data of the hydraulic support, P avg represents the mean value of the point cloud in the neighborhood, and the vector corresponding to the minimum eigenvalue is the normal vector N(x n ,y n , z n ). After calculating the normal vector of each point, set the angle threshold between the normal vectors of adjacent points to N φ ,If the angle is greater than a certain threshold, it is considered that there may be defects in the area.

[0026] 3. Laser positioning of abnormal areas Based on the three-dimensional characterization information, the fault degree of the hydraulic support is analyzed. The failure of the hydraulic support is a three-dimensional problem. The use of three-dimensional point cloud data can completely capture its surface morphology. Therefore, the point cloud data is combined to analyze the distribution of abnormal areas and capture more subtle wear details, so as to determine the specific location and scope of the fault and realize the fault diagnosis of the hydraulic support.

[0027] Project the laser beam onto the working surface. The focus of the laser beam should correspond to the center position of the hydraulic support to be tested. The abnormal point cloud data is marked by the density and abnormality of the point cloud data. Assume that the mean and standard deviation of the hydraulic support point cloud data set are μ i and σ i , calculate the anomaly score for each region: , where f represents the characteristic value of the hydraulic support characterization information. Set a basic threshold, if z i The value exceeds the fluctuation threshold z under normal operating conditions φ , it means that there is an abnormality in the point cloud data of this part, and the area is judged as an abnormal area. At the same time, quantitative information such as wear depth and area is obtained.

[0028] While the laser beam is projected onto the work surface, a visual marking tool (such as spray paint, marking pen, etc.) is used to mark the area near the abnormal area, lock the location and range of the abnormal area, and extract and fit the quantitative information with the three-dimensional representation information from the point cloud data that has been marked as abnormal, extract the coordinate information of the abnormal area, and combine the extracted coordinates of the abnormal area with the feature threshold judgment results to clarify the severity of the fault.

[0029] Example Simulation experiment In order to verify the superiority of the hydraulic support fault diagnosis method for coal mining face based on laser 3D scanning technology proposed in this paper in terms of actual diagnostic effect, a coal mining face in a mine was selected as a simulation experiment scene. The face is located deep in the mine, with relatively complex geological conditions. The coal seam thickness is between 2.5 and 3.5 m, and the inclination is about 15°. Figure 2 shown.

[0030] Figure 2 The hydraulic support model of the working face is ZY6800 / 18 / 38, and its main parameters are configured as follows: the support height ranges from 1.8 m to 3.8 m, which can adapt to the mining of coal seams of different thicknesses; the support working resistance reaches 6800 kN, which can effectively support the coal seam and prevent the roof from collapsing; the center distance of the support is 1.5 m, which ensures the advancement speed of the working face and the passability of the coal mining machine.

[0031] The Leica BLK360 3D laser scanner was used to collect the working status data of the hydraulic support on the working surface. The sampling threshold of the 3D laser scanner was set to 0.75, and the sampling interval of the fault data sample was set to 0.5 s to simulate the actual sampling environment, so as to obtain the fault distribution data of the hydraulic support. The time difference Δt between the emission and reception of the laser beam was 0.5, and the normal vector angle threshold N was set to 0.5. φ is 0.71, and the abnormal score fluctuation threshold Z φ is 0.76, and the three-dimensional point cloud data is obtained. The point cloud image and structural surface of the hydraulic support are as follows Figure 3 shown.

[0032] The collected surface data samples of the hydraulic support, including data samples under normal operating conditions and fault data samples, are input into HDFS. Based on the Windows 10 operating system, i5-9400F processor and NVIDIA GeForce GTX1070 graphics card, Python programming language is used, and the test simulation is carried out using the MATLAB 2020a simulation platform. Figure 3 The distribution of laser scanning data of hydraulic support is as follows: Figure 4 shown.

[0033] In order to improve the diversity of the experimental data set, a fault type library was constructed for the hydraulic support fault sample data set. The fault type library covers a variety of fault type data including column failure. The specific characteristic value distribution of the fault type library is shown in Table 1.

[0034] Table 1 Distribution of fault type library characteristics Two groups of conventional fault diagnosis methods for hydraulic supports in coal mining working faces were selected as comparison objects, namely, a fault diagnosis method based on neural network (Zhong Jinbao, Qi Jingshuai, Li Lei, et al. Research on fault simulation of hydraulic support column control loop based on particle swarm algorithm optimization BP neural network [J]. Engineering Machinery, 2023, 54 (06): 75-82+10.), and a deep learning model (HUANG K, WU S, LI F, et al. Fault diagnosis of hydraulic systems based on deep learning model with multirate data samples [J]. IEEE Transactions on neural networks and learning systems, 2021, 33 (11): 6789-6801). The three methods were used to simulate fault diagnosis for the experimental data set. The fault diagnosis results of the hydraulic support by the method of the present invention are shown in Table 2.

[0035] Table 2 Fault diagnosis results of hydraulic support by the method of the present invention It can be seen from Tables 1 and 2 that the method of the present invention can identify different fault types and determine the severity of the fault. This is because the method of the present invention effectively extracts the three-dimensional characterization information of the hydraulic support, and analyzes the smoothness of the support surface by calculating the average curvature and normal vector of the point cloud data, so as to obtain the wear depth, leakage rate, etc. in a targeted manner.

[0036] Using the method of the present invention, comparative method 1 (Liu Yongzhi, Zhang Zhouqiang, Guo Zhongchao, et al. Research on three-dimensional surface detection system of parts based on line laser scanning [J]. Foreign Electronic Measurement Technology, 2021, 40(2):67-72), and comparative method 2 ([6] CHENG W, MA Y, HUANG W. Simultaneous compositional and defect measurementsusing laser opto-ultrasonic dual detection for laser 3D printing and welding [J]. 2021, 11677(3):5-9), two minor faults that are difficult to diagnose, rod column misalignment and support plate wear, are identified. The identified visualization images are as follows: Figure 5 shown.

[0037] Depend on Figure 5The visualization results show that the method of the present invention accurately identifies the locations of the pole misalignment and the side plate wear faults, while the comparison method 1 incorrectly identifies the pole misalignment location, and the comparison method 2 only identifies the side plate wear fault. The comparison results prove that the method of the present invention is more accurate in identifying minor faults. This is because the method of the present invention innovatively designs the three-dimensional representation information extraction process of the hydraulic support, analyzes the surface smoothness of the hydraulic support, reduces the complexity of the three-dimensional representation information, reduces interference, and improves recognition accuracy.

[0038] The percentage of correctly diagnosed samples after applying different methods (the method of the present invention, comparative method 1, and comparative method 2) is used as a comparative indicator to measure the actual diagnostic accuracy of different methods. The specific experimental results are as follows: Figure 6 shown.

[0039] Depend on Figure 6 It can be seen that when simulating fault diagnosis for the same data set, the diagnostic accuracy of different methods is also different, and the method of the present invention is significantly better than other methods in diagnostic accuracy, and the proportion of correctly diagnosed samples is significantly higher. This is because the method of the present invention combines the average curvature and normal vector of the point cloud data, analyzes the distribution of abnormal areas, captures more subtle fault information, and ensures diagnostic accuracy.

[0040] The present invention designs a fault diagnosis method for hydraulic supports of coal mining working faces based on laser three-dimensional scanning technology, obtains three-dimensional point cloud data of the hydraulic supports of coal mining working faces, deeply explores the three-dimensional characterization information extraction process of the hydraulic supports, and uses the three-dimensional point cloud data to analyze the distribution of abnormal areas, completely capture the surface morphology of the hydraulic supports, and determine the abnormal areas, thereby determining the severity of the fault and ensuring a high proportion of correctly diagnosed samples, thereby achieving effective and high-precision diagnosis of hydraulic support faults and improving the safety production level of coal mining working faces.

Claims

1. A method for diagnosing hydraulic support faults in a coal mining face based on laser three-dimensional scanning technology, characterized in that: The steps include: The first step is to obtain the 3D point cloud data of the hydraulic support surface through laser 3D scanning technology; The second step is to calculate the average curvature and normal vector of the point cloud data and fully extract the hydraulic support characterization information; The third step is to calculate the anomaly score based on the distribution of point cloud data, and lock the location and range of the abnormal area; The fourth step is to fit the quantitative information with the three-dimensional representation information extraction to determine the severity of the fault and realize fault diagnosis.

2. The method for diagnosing faults of hydraulic supports in coal mining working faces based on laser three-dimensional scanning technology according to claim 1 is characterized in that: The method for acquiring the three-dimensional point cloud data of the surface of the hydraulic support in the first step is as follows: in the coal mining working face, a relatively stable area is selected as the scanning environment, ensuring that there is no interference from debris in the scanning area and the hydraulic support is clearly visible, and the hydraulic support is placed in the predetermined scanning area to ensure that it is in a stable working state; Select markers or positioning devices to help the scanner accurately identify the position and posture of the hydraulic support. The laser scanner emits a laser beam to the surface of the hydraulic support and receives the reflected light signal. Combined with the time difference data between the emission and reception of the laser beam, the three-dimensional point cloud of the surface of the hydraulic support is confirmed. The calculation formula is as follows: , where P(x, y, z) represents the three-dimensional coordinates of a point on the surface of the hydraulic support, O represents the origin coordinates of the three-dimensional laser scanner, Δt represents the time difference between the emission and reception of the laser beam in the three-dimensional laser scanning device, and V represents the direction vector of the laser beam; The surface data points of the hydraulic support are continuously collected until the entire scanning process is completed. The three-dimensional coordinate confirmation results of the hydraulic support surface are presented in the form of a three-dimensional image, generating three-dimensional point cloud data to intuitively display the geometric structure and deformation of the support.

3. The method for diagnosing faults of hydraulic supports in coal mining working faces based on laser three-dimensional scanning technology according to claim 1 is characterized in that: The mean curvature describes the average curvature at a point on the surface. For each point P (x, y, z) in the point cloud data, its mean curvature can be obtained by calculating the curvature of all points in its neighborhood and taking the average value. The specific calculation formula of the mean curvature k is as follows: , where n represents the total number of coordinate nodes in the neighborhood, k i Represents the curvature corresponding to each coordinate node, and R1 and R2 represent the two principal curvature radii at point p respectively.

4. The method for diagnosing faults of hydraulic supports in coal mining working faces based on laser three-dimensional scanning technology according to claim 1 is characterized in that: The calculation of the normal vector of the point cloud data is as follows: for each point P (x, y, z) in the point cloud data, its normal vector N (x n ,y n , z n ) is obtained by calculating the covariance matrix of the points in its neighborhood and obtaining its eigenvector. The specific expression of the covariance matrix is ​​as follows: , where P i represents a random point in the neighborhood of the 3D point cloud data of the hydraulic support, P avg represents the mean value of the point cloud in the neighborhood, and the vector corresponding to the minimum eigenvalue is the normal vector N(x n ,y n , z n ), after calculating the normal vector of each point, set the angle threshold between the normal vectors of adjacent points to N φ ,If the angle is greater than a certain threshold, it is considered that there may be defects in the area.

5. The method for diagnosing faults of hydraulic supports in coal mining working faces based on laser three-dimensional scanning technology according to claim 1 is characterized in that: The determination method of the abnormal area and abnormal range is as follows: project a laser beam onto the working surface, the focus of the laser beam corresponds to the center position of the hydraulic support to be measured, and mark the abnormal point cloud data by the density and abnormality of the point cloud data. Assume that the mean and standard deviation of the hydraulic support point cloud data set are μ i and σ i , calculate the anomaly score for each region: , where f represents the characteristic value of the hydraulic support characterization information, and a basic threshold is set. If z i The value exceeds the fluctuation threshold z under normal operating conditions φ , it means that the point cloud data of this part is abnormal, and the area is judged as an abnormal area. At the same time, quantitative information such as wear depth and area is obtained.

6. The method for diagnosing faults of hydraulic supports in coal mining working faces based on laser three-dimensional scanning technology according to claim 1, characterized in that: The method for determining the severity of the fault is as follows: while the laser beam is projected onto the work surface, a visual marking tool is used to mark near the abnormal area, the location of the abnormal area and the abnormal range are locked, and from the point cloud data that has been marked as abnormal, quantitative information and three-dimensional characterization information are extracted and fitted to extract the coordinate information of the abnormal area. The extracted coordinates of the abnormal area are combined with the feature threshold judgment result to clarify the severity of the fault.

Images