Intelligent detection device for wear of consumables in thermal power plant

Abstract

The invention relates to an intelligent detection device for consumable wear of a thermal power plant, which relates to the field of consumable detection and comprises a clamping locking assembly and a multi-degree-of-freedom detection assembly. The clamping locking assembly comprises a limiting supporting frame with the preset operation length and supporting force. The screwing locking rod penetrates through the supporting frame, has a preset operation length and is movably connected with the supporting frame. The multi-degree-of-freedom detection assembly comprises a positioning connection part which is connected with the supporting frame and has a preset operation length and supporting force. The extension adjusting part is movably connected with the positioning connecting part and can carry out movable adjusting operation in a preset range; and the detection searchlighting part is movably connected with the extension adjusting part, has a preset operation length and supporting force, and performs movable adjusting operation in a preset range. The surface of the grinding roller can be shot and imaged through the detection searchlighting part, the abrasion area and the abrasion degree of the grinding roller are captured and analyzed through image analysis, a corresponding analysis report is generated, and manual reinspection is facilitated.

Description

Technical Field

[0001] The present invention relates to the field of consumable material detection, and in particular to an intelligent detection device for consumable material wear in a thermal power plant. Background Art

[0002] A vertical coal mill is a commonly used pulverized coal preparation device in thermal power plants. It is primarily used to grind coal into a fine powder for efficient combustion in boilers. Vertical coal mills are typically compact, occupying a small footprint, making them suitable for space-constrained thermal power plants. Vertical coal mills efficiently grind coal to the desired fineness, typically achieving 85% passing a 200-mesh sieve, thereby improving combustion efficiency. Coal is pressed and ground by rollers on the millstones. The rollers crush the coal into a fine powder, which is then carried away by airflow and passed to a separator for classification. The qualified pulverized coal is then fed into the boiler for combustion. The rollers are a key component of the vertical coal mill, responsible for pulverizing and grinding the coal. Therefore, wear on the rollers is common and unavoidable over time. Testing roller wear is a routine and necessary procedure.

[0003] Traditional roller wear detection systems require the roller to be dismantled before inspection for cracks, wear, and other conditions. Due to the roller's large size and weight, dismantling the roller is a significant undertaking. Furthermore, routine inspections and dismantling waste manpower and resources. While existing vertical mills can partially turn the roller outside for routine inspections, this creates a blind spot near the roller's sidewalls, which are close to the mill body. Rotating the roller for inspection presents safety risks. Summary of the Invention

[0004] Purpose of the invention: To provide an intelligent detection device for wear of consumables in thermal power plants to solve the above-mentioned problems existing in the prior art.

[0005] Technical solution: An intelligent detection device for wear of consumables in thermal power plants, comprising a locking component and a multi-degree-of-freedom detection component.

[0006] The locking assembly includes a position-limiting support frame having a predetermined operating length and supporting force, a predetermined bending arc, and capable of performing corresponding position-limiting and connection operations; a screw-locking rod passing through the support frame, having a predetermined operating length, and being movably connected to the support frame, capable of performing movably adjusted within a predetermined range;

[0007] The multi-degree-of-freedom detection component includes a positioning connection part connected to the support frame, having a predetermined working length and supporting force; an extension adjustment part movably connected to the positioning connection part, having a predetermined working length and supporting force, and capable of performing a predetermined range of movable adjustment operation; and a detection searchlight part movably connected to the extension adjustment part, having a predetermined working length and supporting force, and capable of performing a predetermined range of movable adjustment operation.

[0008] In a further embodiment, the positioning and connecting portion comprises three components: an adapter plate, a rotating platform, and a support column. The adapter plate is connected to the position-limiting support frame, has a predetermined operating size and load capacity, and can perform corresponding adapter and position-limiting operations. The rotating platform is rotatably connected to the adapter plate, has a predetermined operating size and load capacity, and can perform rotational adjustment operations within a predetermined range. The support column is connected to the rotating platform, has a predetermined operating length and support force, and can be adjusted in the corresponding operating direction in accordance with the rotational adjustment of the rotating platform.

[0009] In a further embodiment, the extension adjustment portion comprises an extension arm and an adjustment arm. The extension arm is rotatably connected to the support column, has a predetermined operating length and support force, and can be rotated and adjusted within a predetermined range. The adjustment arm is rotatably connected to the extension arm, has a predetermined operating length and support force, and can be rotated and adjusted within a predetermined range.

[0010] In a further embodiment, the inspection and searchlight unit comprises three components: a connecting rod, an inspection shaft, and a searchlight head. The connecting rod is connected to the adjustment arm and has a predetermined operating length, enabling corresponding engagement and support limiting operations. The inspection shaft is rotatably connected to the connecting rod and has a predetermined operating length and support force, enabling rotational adjustment within a predetermined range. The searchlight head is connected to the inspection shaft and has a corresponding industrial camera mounted on its working surface, which can adjust the operating direction accordingly with the rotational adjustment of the inspection shaft.

[0011] In a further embodiment, the adapter plate is threadedly connected to the position limiting support frame; the screw locking rod passes through the position limiting support frame and is threadedly connected to the position limiting support frame, and can be screwed and adjusted accordingly.

[0012] In a further embodiment, the industrial camera provided in the search head performs imaging analysis on the grinding roller and analyzes the detected data results. The specific steps are as follows:

[0013] S1, performing surround imaging of the side wall surface of the grinding roller and preprocessing the captured images;

[0014] S2. Extracting edge features and texture features of the grinding roller surface from the pre-processed captured images, and obtaining depth information of the grinding roller surface through stereo vision technology to perform corresponding wear analysis;

[0015] S3. Evaluate the degree of wear on the grinding roller surface based on the wear analysis data results and generate a corresponding data analysis report.

[0016] In a further embodiment, in step S1, the industrial camera provided on the search head is synchronously adjusted to the corresponding operating position by adjusting the operating direction of the extension adjustment part, and the imaging of the side wall of the grinding roller is completed by continuously shooting images.

[0017] In a further embodiment, the specific steps of extracting the edge features and texture features of the grinding roller surface in step S2 are as follows:

[0018] S201, read the original image information and amplify the image using an interpolation algorithm, as follows:

[0019] I(x′,y′)=(1-dx)(1-dy)I(x n ,y n )+dx(1-dy)I(x n +1,y n )+(1-dx)dyI(x n ,y n +1)+dx·dyI(x n +1,y n +1)

[0020] Where I(x′,y′) is the magnified image pixel, (x n ,y n ) is the neighborhood pixel;

[0021] S202, using a filter to reduce image noise, identifying edges in the image using an edge detection operator, and using a tracking algorithm to track the edges to enhance edge recognition;

[0022] S203, segmenting the image after edge feature capture, generating a corresponding co-occurrence matrix, and extracting image features, as follows:

[0023]

[0024] Where i, j are gray levels, θ is the direction, d is the distance, N is the total number of pixel pairs, and δ is a discrete function;

[0025] S204: Compare the grayscale values of the central pixel of the image with those of its neighboring pixels, perform transformation and normalization of the corresponding pattern, and extract the corresponding image texture features, as follows:

[0026]

[0027] Among them, P is the field point, p x ,p y is the neighborhood point coordinate, c is the center pixel value, and s is the sign function.

[0028] In a further embodiment, in step S2, the depth information of the grinding roller surface is obtained by stereo vision technology. Specifically, the original captured image of the grinding roller is used as the original image, and the real-time grinding roller image placed on site is used as the reference image. The disparity d between the original image and the reference image is obtained by a matching algorithm, as follows:

[0029] d=x L -x R

[0030] Among them, x L and x R are the horizontal coordinates of the corresponding points in the original image and the control image respectively;

[0031] Calculate the corresponding depth information Z according to the disparity d as follows:

[0032]

[0033] Where f is the focal length of the camera, B is the baseline distance between the cameras, and d is the parallax;

[0034] The extracted texture features and depth information are fused. Based on the fused features, a machine learning algorithm is used to identify and classify the wear areas on the grinding roller surface. The degree of wear is quantified based on the area and depth information of the identified wear areas.

[0035] Beneficial effects: The present invention relates to an intelligent detection device for wear of consumables in thermal power plants, which relates to the field of consumables detection and includes two parts: a locking assembly and a multi-degree-of-freedom detection assembly. The locking assembly includes a limiting support frame with a predetermined operating length and support force, a predetermined bending arc, and can perform corresponding limiting and connection operations; a screw locking rod that passes through the support frame, has a predetermined operating length, and is movably connected to the support frame and can be adjusted within a predetermined range. The multi-degree-of-freedom detection assembly includes a positioning connection part connected to the support frame, with a predetermined operating length and support force; an extension adjustment part that is movably connected to the positioning connection part, has a predetermined operating length and support force, and can be adjusted within a predetermined range; and a detection search part that is movably connected to the extension adjustment part, has a predetermined operating length and support force, and can be adjusted within a predetermined range. The present invention can capture and image the surface of the grinding roller through the detection search part, and capture and analyze the wear area and wear degree of the grinding roller through image analysis, and generate a corresponding analysis report to facilitate manual re-inspection. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 It is a schematic diagram of the present invention as a whole.

[0037] Figure 2 Schematic diagram of the multi-degree-of-freedom detection component of the present invention.

[0038] Figure 3 Schematic diagram of the locking assembly of the present invention.

[0039] Figure 4 It is an enlarged schematic diagram of some components of the present invention.

[0040] The reference numerals in the figure are: position limiting support frame 1, screw locking rod 2, connecting plate 3, rotating platform 4, support column 5, extension arm 6, adjustment arm 7, detection search unit 8, connecting rod 801, detection shaft 802, search head 803. DETAILED DESCRIPTION

[0041] In the following description, numerous specific details are provided to provide a more thorough understanding of the present invention. However, it will be apparent to those skilled in the art that the present invention may be practiced without one or more of these details. In other instances, certain technical features well known in the art are not described to avoid confusion with the present invention.

[0042] The applicant believes that conventional roller wear detection devices require the roller to be disassembled before testing for cracks, wear, and other conditions. Due to the large size and weight of the roller itself, disassembly is a significant effort. Furthermore, routine maintenance and disassembly wastes manpower and resources. While existing vertical mills can partially turn the roller outside for routine inspection, this creates a blind spot near the roller's sidewalls, which are close to the mill body. Rotating the roller for inspection presents a safety hazard.

[0043] To this end, the applicant designed an intelligent detection device for wear of consumables in thermal power plants. The detection and illumination unit 8 can capture and image the surface of the grinding roller, and capture and analyze the wear area and wear degree of the grinding roller through image analysis, and generate a corresponding analysis report to facilitate manual re-inspection.

[0044] The intelligent detection device for wear of consumables in thermal power plants according to the present invention mainly comprises two parts: a locking assembly and a multi-degree-of-freedom detection assembly. The locking assembly comprises a limiting support frame 1 having a predetermined operating length and supporting force, a predetermined bending arc, and capable of performing corresponding limiting and connection operations; a screw locking rod 2 passing through the support frame, having a predetermined operating length, and being movably connected to the support frame, capable of performing movable adjustment within a predetermined range. The multi-degree-of-freedom detection assembly comprises a positioning connection portion connected to the support frame, having a predetermined operating length and supporting force; an extension adjustment portion movably connected to the positioning connection portion, having a predetermined operating length and supporting force, and capable of performing movable adjustment operations within a predetermined range; and a detection search portion 8 movably connected to the extension adjustment portion, having a predetermined operating length and supporting force, and capable of performing movable adjustment operations within a predetermined range. In the later operation process, the positioning support frame 1 cooperates with the screw locking rod 2 to perform corresponding limiting and fixing of the entire device. Then, the imaging of the grinding roller surface is completed by adjusting the multi-degree-of-freedom detection component, and the wear condition of the grinding roller is further determined by image analysis. In a further preferred embodiment, the limit support frame 1 is in the shape of a U-shaped character, and the screw locking rod 2 passes through the limit support frame 1, and can be adapted to grinding roller support plates of different thicknesses by adjusting the length of the screw locking rod 2 passing through the limit support frame 1. The grinding roller support plate described here is a reserved outer flap that limits the bearing of the grinding roller and can be sealed with the machine body. This is an existing component structure, and can be understood by referring to the component structure of the existing vertical coal mill. The intelligent detection device involved in this application is installed at the grinding roller support plate. By clamping the recessed part of the limit support frame 1 to the grinding roller support plate, and then adjusting the screw locking rod 2, the installation and fixation of the intelligent detection device involved in this application can be completed.

[0045] The positioning connection part includes three components: a connection plate 3, a rotating table 4 and a support column 5. The connection plate 3 is connected to the limiting support frame 1, has a predetermined operating size and a carrying space, and can perform corresponding connection and limiting operations. The rotating table 4 is rotatably connected to the connection plate 3, has a predetermined operating size and a carrying space, and can perform a rotation adjustment operation within a predetermined range. The support column 5 is connected to the rotating table, has a predetermined operating length and supporting force, and can adjust the corresponding operating direction following the rotation adjustment of the rotating table 4. In the later operation process, the rotation adjustment of the rotating table 4 can drive the support column 5 as a whole to be adjusted to the corresponding operating position, thereby realizing the synchronous operation direction adjustment of the extension adjustment part and the detection search part 8. In a further preferred embodiment, the support column 5 is a telescopic cylinder, which can perform a telescopic adjustment operation within a predetermined range, and in the later operation process, the operating length of the support column 5 can be flexibly adjusted according to corresponding needs.

[0046] The extension adjustment portion includes two components: an extension arm 6 and an adjustment arm 7. The extension arm 6 is rotatably connected to the support column 5, has a predetermined operating length and supporting force, and can be rotated and adjusted within a predetermined range. The adjustment arm 7 is rotatably connected to the extension arm 6, has a predetermined operating length and supporting force, and can be rotated and adjusted within a predetermined range. In the later operation process, the adjustment arm 7 can be driven to adjust the corresponding operating position as a whole by rotating the extension arm 6, and the detection search portion 8 can be driven to adjust the corresponding operating angle as a whole by rotating the adjustment arm 7. In a further preferred embodiment, both the extension arm 6 and the adjustment arm 7 have a predetermined elasticity, and can perform an elastic adjustment operation within a predetermined range.

[0047] The detection search unit 8 includes three components: a connecting rod 801, a detection shaft 802, and a search head 803. The connecting rod 801 is rotatably connected to the adjustment arm 7, has a predetermined operating length, and can perform corresponding connection and support limit operations. The detection shaft 802 is rotatably connected to the connecting rod 801, has a predetermined operating length and support force, and can perform rotation adjustment operations within a predetermined range. The search head 803 is connected to the detection shaft 802, and a corresponding industrial camera is provided on the working surface, which can be adjusted to the corresponding working direction following the rotation adjustment of the detection shaft 802. In the later operation process, the rotation adjustment of the connecting rod 801 can drive the detection shaft 802 as a whole to adjust the corresponding working angle, and then the rotation adjustment of the detection shaft 802 can drive the search head 803 as a whole to adjust the corresponding working angle, so as to perform all-round photography and imaging processing of the grinding roller.

[0048] The adapter plate 3 is threadedly connected to the position limiting support frame 1; the screw locking rod 2 passes through the position limiting support frame 1 and is threadedly connected to the position limiting support frame 1, allowing for corresponding screwing adjustment. In the later operation process, the screw locking rod 2 can be screwed to adjust the operating length of the screw locking rod 2 passing through the position limiting support frame 1, thereby cooperating with the position limiting support frame 1 to perform the corresponding position limiting and fixing operation.

[0049] In a further preferred embodiment, the adapter plate 3 is threadedly connected to the position limiting support frame 1, the extension arm 6 is threadedly connected to the support column 5 via a rotating gasket, the adjustment arm 7 is threadedly connected to the extension arm 6 via a rotating gasket, and the connecting rod 801 is threadedly connected to the adjustment arm 7 via a rotating gasket, so that when not in use, the components can be disassembled and separated, which is convenient for storage of the equipment. The intelligent detection device involved in this application is externally connected to a corresponding display terminal, the searchlight 803 and the external motor are both wirelessly connected to the display terminal, and the rotating gasket is wirelessly connected to the external motor. The display terminal is provided with a corresponding control module, which can control the start-up of the external motor and the searchlight 803 through the control module, so that the external motor can drive the rotating gasket to perform corresponding rotation adjustment, and the searchlight 803 can perform corresponding shooting operations.

[0050] The industrial camera provided in the search head 803 performs imaging analysis on the grinding roller and analyzes the detection data results. The specific steps are as follows:

[0051] Firstly, the side wall surface of the grinding roller is imaged in a surround manner and the image is pre-processed;

[0052] Next, the edge features and texture features of the grinding roller surface are extracted from the pre-processed images, and the depth information of the grinding roller surface is obtained through stereo vision technology to perform corresponding wear analysis.

[0053] Finally, the wear degree of the grinding roller surface is evaluated based on the wear analysis data results, and a corresponding data analysis report is generated.

[0054] The step is to adjust the operating direction of the extension adjustment part to drive the industrial camera provided on the search head 803 to adjust the corresponding operating position synchronously, and complete the imaging of the side wall of the grinding roller by continuously shooting images.

[0055] The specific steps of extracting the edge features and texture features of the grinding roller surface in step 2 are as follows:

[0056] First, the original image is read and the image is enlarged using an interpolation algorithm, as follows:

[0057] I(x′,y′)=(1-dx)(1-dy)I(x n ,y n )+dx(1-dy)I(x n +1,y n )+(1-dx)dyI(x n ,y n +1)+dx·dyI(x n +1,y n +1)

[0058] Where I(x′,y′) is the magnified image pixel, (x n ,y n ) is the neighborhood pixel;

[0059] Next, a filter is used to reduce image noise, the Canny edge detection operator is used to identify edges in the image, and a tracking algorithm is used to track edges and enhance edge recognition.

[0060] Subsequently, the image after edge feature capture is segmented, and the corresponding co-occurrence matrix is generated to extract image features, as follows:

[0061]

[0062] Where i, j are gray levels, θ is the direction, d is the distance, N is the total number of pixel pairs, and δ is the Kronecker function;

[0063] Finally, the grayscale values of the central pixel of the image are compared with those of its neighboring pixels, and the corresponding pattern transformation and normalization are performed to extract the corresponding image texture features, as follows:

[0064]

[0065] Among them, P is the field point, p x ,p y is the neighborhood point coordinate, c is the center pixel value, and s is the sign function.

[0066] In the second step, the depth information of the grinding roller surface is obtained by stereo vision technology. Specifically, the original shot image of the grinding roller is used as the original image, and the real-time grinding roller image placed on site is used as the reference image. The disparity d between the original image and the reference image is obtained by a matching algorithm, which is as follows:

[0067] d=x L -x R

[0068] Among them, x L and x R are the horizontal coordinates of the corresponding points in the original image and the control image respectively;

[0069] Calculate the corresponding depth information Z according to the disparity d as follows:

[0070]

[0071] Where f is the focal length of the camera, B is the baseline distance between the cameras, and d is the parallax;

[0072] The extracted texture features and depth information are fused. Based on the fused features, a machine learning algorithm is used to identify and classify the wear areas on the grinding roller surface. The degree of wear is quantified based on the area and depth information of the identified wear areas.

[0073] As described above, although the present invention has been shown and described with reference to specific preferred embodiments, it should not be construed as limiting the present invention itself. Various changes may be made to it in form and detail without departing from the spirit and scope of the present invention as defined in the appended claims.

Claims

1. An intelligent detection device for wear of consumables in thermal power plants, characterized by include: The locking assembly includes a position-limiting support frame having a predetermined operating length and supporting force, a predetermined bending arc, and capable of performing corresponding position-limiting and connection operations; a screw-locking rod passing through the support frame, having a predetermined operating length, and being movably connected to the support frame, capable of performing movably adjusted within a predetermined range; A multi-degree-of-freedom detection assembly includes a positioning connection portion connected to the support frame and having a predetermined operating length and supporting force; An extension adjustment portion movably connected to the positioning and connecting portion, having a predetermined operating length supporting force and capable of performing a predetermined range of movable adjustment operation; The detection and illumination part is movably connected to the extension adjustment part, has a predetermined operating length and supporting force, and performs a predetermined range of movable adjustment operation.

2. The intelligent detection device for wear of consumables in thermal power plants according to claim 1, characterized in that: The positioning and connecting portion includes: The connecting plate is connected to the position-limiting support frame and has a predetermined operating size and carrying space, and can perform corresponding connecting and position-limiting operations; A rotating platform is rotatably connected to the adapter plate, has a predetermined operating size and carrying space, and can perform a rotation adjustment operation within a predetermined range; The support column is connected to the rotating platform, has a predetermined working length and supporting force, and can be adjusted to the corresponding working direction following the rotation adjustment of the rotating platform.

3. The intelligent detection device for wear of consumables in thermal power plants according to claim 2, characterized in that: The extension adjustment portion includes: An extension arm is rotatably connected to the support column, has a predetermined operating length and supporting force, and can be rotated and adjusted within a predetermined range; The adjusting arm is rotatably connected to the extension arm, has a predetermined operating length and supporting force, and can perform a rotation adjustment operation within a predetermined range.

4. The intelligent detection device for wear of consumables in thermal power plants according to claim 3, characterized in that: The detection and illumination unit includes: A connecting rod, connected to the adjusting arm, has a predetermined operating length and can perform corresponding connection and support limiting operations; A detection shaft is rotatably connected to the connecting rod, has a predetermined operating length and supporting force, and can be rotated and adjusted within a predetermined range; The search head is connected to the detection axis, and a corresponding industrial camera is provided on the surface of the working surface, which can be adjusted to the corresponding working direction following the rotation adjustment of the detection axis.

5. The intelligent detection device for wear of consumables in thermal power plants according to claim 2, characterized in that: The connecting plate is threadedly connected to the limiting support frame; the screw locking rod passes through the limiting support frame and is threadedly connected to the limiting support frame, and can be screwed and adjusted accordingly.

6. The intelligent detection device for wear of consumables in thermal power plants according to claim 4, characterized in that: The industrial camera provided with the search head performs imaging analysis on the grinding roller and analyzes the detection data results. The specific steps are as follows: S1, performing surround imaging of the side wall surface of the grinding roller and preprocessing the captured images; S2. Extracting edge features and texture features of the grinding roller surface from the pre-processed captured images, and obtaining depth information of the grinding roller surface through stereo vision technology to perform corresponding wear analysis; S3. Evaluate the degree of wear on the grinding roller surface based on the wear analysis data results and generate a corresponding data analysis report.

7. The intelligent detection device for wear of consumables in thermal power plants according to claim 6, characterized in that: In the step S1, the industrial camera provided on the search head is driven to adjust the corresponding working position synchronously by adjusting the working direction of the extension adjustment part, and the imaging of the side wall of the grinding roller is completed by continuously shooting images.

8. The intelligent detection device for wear of consumables in thermal power plants according to claim 6, characterized in that: The specific steps of extracting the edge features and texture features of the grinding roller surface in step S2 are as follows: S201, read the original image information and amplify the image using an interpolation algorithm, as follows: I(x′,y′)=(1-dx)(1-dy)I(xn,yn)+dx(1-dy)I(xn+1,yn)+(1-dx)dyI(xn,yn+1)+dx·dyI(xn+1,yn+1)where I(x′,y′) is the magnified image pixel and (xn,yn) is the neighborhood pixel; S202, using a filter to reduce image noise, identifying edges in the image using an edge detection operator, and using a tracking algorithm to track the edges to enhance edge recognition; S203, segmenting the image after edge feature capture, generating a corresponding co-occurrence matrix, and extracting image features, as follows: Where i, j are gray levels, θ is the direction, d is the distance, N is the total number of pixel pairs, and δ is a discrete function; S204: Compare the grayscale values of the central pixel of the image with those of its neighboring pixels, perform transformation and normalization of the corresponding pattern, and extract the corresponding image texture features, as follows: Among them, P is the number of points in the neighborhood, px,py are the coordinates of the neighborhood points, c is the center pixel value, and s is the sign function.

9. The intelligent detection device for wear of consumables in thermal power plants according to claim 6, characterized in that: In step S2, the depth information of the grinding roller surface is obtained by stereo vision technology. Specifically, the original shot image of the grinding roller is used as the original image, and the real-time grinding roller image placed on site is used as the reference image. The disparity d between the original image and the reference image is obtained by a matching algorithm, which is as follows: d=x L -x R Among them, x L and x R are the horizontal coordinates of the corresponding points in the original image and the control image respectively; Calculate the corresponding depth information Z according to the disparity d as follows: Where f is the focal length of the camera, B is the baseline distance between the cameras, and d is the parallax; The extracted texture features and depth information are fused. Based on the fused features, a machine learning algorithm is used to identify and classify the wear areas on the grinding roller surface. The degree of wear is quantified based on the area and depth information of the identified wear areas.

Images