Graphite calcining furnace external material treatment method and system based on crane

The three-dimensional twin point cloud is generated by lidar, and the suction tube is controlled to absorb powdery filler on the edge of the graphite calciner furnace cover and the calcined coke block, solving the problem of calcined coke block block block and the suction tube colliding with the furnace cover. Improves the removal efficiency and safety of graphite calciner.

CN119027486APending Publication Date: 2024-11-26XINXIANG UNIV +1
View PDF 6 Cites 0 Cited by

Patent Information

Application Number
CN202411496338.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-10-25
Publication Date
2024-11-26

Smart Images

  • Figure CN119027486A_ABST
    Figure CN119027486A_ABST
Patent Text Reader

Abstract

The invention discloses a crane-based graphite calcining furnace external material processing method and system, and the method comprises the following steps: 1, collecting data in a graphite calcining pit, and obtaining original point cloud data; 2, simplifying the original point cloud data to obtain simplified point cloud data; step 3, filtering and noise reduction are carried out on the simplified point cloud data to obtain filtered and noise-reduced point cloud data; 4, performing point cloud segmentation on the filtered and denoised point cloud data, and performing fitting reconstruction according to the segmented point cloud data to obtain a three-dimensional twin point cloud; and 5, controlling a material suction pipe of a material suction crown block according to the three-dimensional twin point cloud to suck the powdery filler near the calcined coke blocks outside the calcining furnace. According to the method, interference in the point cloud data is reduced through simplification and filtering noise reduction of the original point cloud data, the accuracy of laser radar recognition is ensured, the material suction pipe is prevented from colliding with the calcined coke blocks and the furnace cover of the calcining furnace, and powdery filler is convenient to suck.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the technical field of graphite calcining, and in particular to a crane-based external material processing method and system of a graphite calcining furnace. Background Art

[0002] During graphite production, graphite calcination is required. A calcination furnace is set up in the calcination pit, and the outside of the calcination furnace needs to be filled with fillers for heat preservation. After the graphite calcination is completed, the fillers need to be sucked out through the suction pipe on the suction crane under high temperature environment, so that there is no filler around the calcination furnace, thereby allowing the calcination furnace to cool down quickly. However, there are several problems in the existing filler removal process: (1) The filler will form calcined coke lumps under the action of high temperature. When the suction pipe sucks the filler, the calcined coke lumps will be sucked in. The calcined coke lumps are easy to block the suction pipe, so it is generally necessary to pick up the calcined coke lumps manually.

[0003] (2) During the suction process, the suction pipe has poor obstacle avoidance ability and often collides with the calcining furnace cover. Once the calcining furnace cover is damaged, external air will enter the calcining furnace, causing excessive graphite impurities in the calcining furnace to become unusable, resulting in serious losses.

[0004] (3) When the existing recognition technology is used to remove fillers, the amount of calculation is too large and the overall complexity is too large to quickly and accurately control the suction tube to work. Summary of the invention

[0005] In view of the problems that calcined coke agglomerates are prone to block the suction pipe and the suction pipe is prone to hit the cover of the calcining furnace, the present invention provides a crane-based external material handling method and system for a graphite calcining furnace. A laser radar is used to collect data to generate a three-dimensional twin point cloud. A host computer controls the suction pipe according to the three-dimensional twin point cloud to suck powdered filler at the edge of the calcining furnace cover and the calcined coke agglomerates, so that the filler can be quickly sucked in, avoiding the calcining furnace cover from being hit, thereby improving work efficiency and reducing economic losses.

[0006] In order to achieve the above object, the technical solution of the present invention is: The first aspect of the present invention provides a crane-based method for handling materials outside a graphite calcining furnace, comprising the following steps: Step 1: Collect data in the graphite calcination pit to obtain original point cloud data for subsequent processing; Step 2: Simplify the original point cloud data to obtain simplified point cloud data to reduce the amount of calculation and improve data recognition efficiency; Step 3: Filter and denoise the streamlined point cloud data to obtain filtered and denoised point cloud data to avoid various forms of noise influence; Step 4: Segment the point cloud data after filtering and noise reduction, and perform fitting and reconstruction based on the segmented point cloud data to obtain a three-dimensional twin point cloud, which is convenient for identifying the surrounding environment of the calcining pit; Step 5: According to the 3D twin point cloud, the suction pipe of the suction crane is controlled to absorb the powdered filler near the calcined coke agglomerates outside the calcining furnace to facilitate the absorption of the powdered filler. After the powdered filler is absorbed, the calcined coke agglomerates will sink to the bottom of the calcining pit. During the next calcination, the calcined coke agglomerates will turn into powdered filler again.

[0007] Furthermore, the specific process of step 2 includes: The original point cloud data is voxelized, and the side length of the voxelized grid is determined according to the number of original point cloud data and the volume of the scanned space to ensure that the overall characteristics of the point cloud before and after simplification remain unchanged; The point cloud in the voxelized grid is replaced by the centroid of the voxelized grid to obtain the processed point cloud, which is convenient for realizing the dimensionality reduction and compression of the point cloud data; The processed point cloud is downsampled to obtain streamlined point cloud data, which is convenient for reducing the amount of calculation.

[0008] Furthermore, the side length of the voxelized grid is calculated according to the following formula: Where N is the total number of point clouds, is the x-axis point cloud span, is the y-axis point cloud span, is the span of the z-axis point cloud, g is the average number of point clouds per unit volume, a is the grid edge length adjustment factor, and s is the scale factor.

[0009] Furthermore, the step three specifically includes: Calculate the distance set between each point cloud and its k-neighbors; The mean and standard deviation of the distance between each point cloud and its k-neighborhood are calculated based on the distance set to obtain the maximum threshold; The maximum threshold is calculated according to the mean and standard deviation of the distance, and the point cloud whose distance to the k-neighborhood is greater than the maximum threshold is eliminated to obtain the point cloud data after filtering and denoising, avoiding various forms of noise influence.

[0010] Furthermore, the maximum threshold is calculated according to the following formula: in, is the mean, n is the total number of point clouds, is the distance between the point cloud and its k-neighborhood, is the standard deviation, is the standard deviation coefficient, is the maximum threshold.

[0011] Furthermore, the step five also includes obtaining a three-dimensional model of the graphite calcining pit according to the three-dimensional twin point cloud, locating the position of the calcining furnace cover, and obtaining the coordinates of the calcining furnace cover, so that the suction pipe can avoid the calcining furnace cover; The suction crane controls the suction pipe to suck the powdered filler near the coordinates of the calcining furnace cover, thereby preventing the suction pipe from colliding with the calcining furnace cover.

[0012] The second aspect of the present invention provides a crane-based external material handling system for a graphite calcining furnace, comprising the following steps: The acquisition module is used to collect data in the graphite calcination pit and obtain the original point cloud data for subsequent processing; The simplification module is used to simplify the original point cloud data to obtain the simplified point cloud data, which is used to reduce the amount of calculation and improve the data recognition efficiency; The filtering and denoising module is used to filter and denoise the streamlined point cloud data to obtain the filtered and denoised point cloud data to avoid various forms of noise influence; The reconstruction module is used to segment the point cloud data after filtering and noise reduction, and to perform fitting and reconstruction based on the segmented point cloud data to obtain a three-dimensional twin point cloud, which is convenient for identifying the surrounding environment of the calcining pit; The control module is used to control the suction pipe of the suction crane to suck the powdered filler near the calcined coke agglomerates outside the calcining furnace according to the three-dimensional twin point cloud, so as to absorb the powdered filler. After the powdered filler is absorbed, the calcined coke agglomerates will sink to the bottom of the calcining pit, and will turn into powdered filler again during the next calcination.

[0013] Furthermore, the simplified module specifically includes: The original point cloud data is voxelized, and the side length of the voxelized grid is determined according to the number of original point cloud data and the volume of the scanned space to ensure that the overall characteristics of the point cloud before and after simplification remain unchanged; The point cloud in the voxelized grid is replaced by the centroid of the voxelized grid to obtain the processed point cloud, which is convenient for realizing the dimensionality reduction and compression of the point cloud data; The processed point cloud is downsampled to obtain streamlined point cloud data, which is convenient for reducing the amount of calculation.

[0014] Furthermore, the side length of the voxelized grid is calculated according to the following formula: Where N is the total number of point clouds, is the x-axis point cloud span, is the y-axis point cloud span, is the span of the z-axis point cloud, g is the average number of point clouds per unit volume, a is the grid edge length adjustment factor, and s is the scale factor.

[0015] Furthermore, the filtering and noise reduction module specifically includes: Calculate the distance set between each point cloud and its k-neighbors; The mean and standard deviation of the distance between each point cloud and its k-neighborhood are calculated based on the distance set to obtain the maximum threshold; The maximum threshold is calculated according to the mean and standard deviation of the distance, and the point cloud whose distance to the k-neighborhood is greater than the maximum threshold is eliminated to obtain the point cloud data after filtering and denoising, avoiding various forms of noise influence.

[0016] Furthermore, the maximum threshold is calculated according to the following formula: in, is the mean, n is the total number of point clouds, is the distance between the point cloud and its k-neighborhood, is the standard deviation, is the standard deviation coefficient, is the maximum threshold.

[0017] Furthermore, the control module also includes obtaining a three-dimensional model of the graphite calcining pit according to the three-dimensional twin point cloud, locating the position of the calcining furnace cover, and obtaining the coordinates of the calcining furnace cover, so that the suction pipe can avoid the calcining furnace cover; The suction crane controls the suction pipe to suck the powdered filler near the coordinates of the calcining furnace cover, thereby preventing the suction pipe from colliding with the calcining furnace cover.

[0018] Beneficial effects of the present invention: (1) The present invention simplifies the amount of point cloud data by performing voxel downsampling processing on the point cloud data, calculates the average number of point clouds in a unit volume through the point cloud spans of the x-axis, y-axis and z-axis, and then calculates the side length of the voxelized grid based on the average number of point clouds in the unit volume, so that the overall characteristics of the point cloud data before and after voxelization remain unchanged, thereby reducing the amount of calculation.

[0019] (2) The present invention filters and reduces noise on point cloud data through a statistical filtering algorithm. The maximum threshold is obtained by calculating the mean and standard deviation of the distance between each point cloud and its k-neighborhood. Points whose distance between the point cloud and the k-neighborhood is greater than the maximum threshold are extracted, thereby achieving filtering and reducing noise on the point cloud data and improving the accuracy of the point cloud data.

[0020] (3) The present invention obtains the coordinates of the calcining furnace cover and the calcined coke agglomerate based on the three-dimensional twin point cloud, so that the suction pipe can absorb the powdered filler at the edge of the calcining furnace cover and the calcined coke agglomerate, thereby ensuring the stability of the automatic suction of the powdered filler and avoiding the collision of the suction pipe with the calcining furnace cover and the calcined coke agglomerate and the clogging of the suction pipe by the calcined coke agglomerate. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 A flow chart of a method for handling materials outside a graphite calcining furnace based on a crane provided in an embodiment of the present invention.

[0022] Figure 2 A schematic diagram of a time-of-flight ranging method for a crane-based graphite calcining furnace external material handling method provided in an embodiment of the present invention.

[0023] Figure 3 A schematic diagram of a triangulation distance measurement method for a crane-based method for handling materials outside a graphite calcining furnace provided in an embodiment of the present invention.

[0024] Figure 4 A schematic diagram of a frequency modulated continuous wave ranging method for a crane-based graphite calcining furnace external material handling method provided in an embodiment of the present invention.

[0025] Figure 5 A schematic diagram of the operation of a laser radar in a crane-based graphite calcining furnace external material handling method provided in an embodiment of the present invention.

[0026] Figure 6 A schematic diagram of radar scanning information displayed by radar configuration software of a crane-based graphite calcining furnace external material handling method provided in an embodiment of the present invention.

[0027] Figure 7 A schematic diagram showing a statistical filtering method for handling materials outside a graphite calcining furnace based on a crane provided in an embodiment of the present invention.

[0028] Figure 8 An architectural diagram of a crane-based graphite calcining furnace external material handling system provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0029] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution in the embodiment of the present invention will be clearly described below in conjunction with the drawings in the embodiment of the present invention. Obviously, the described embodiment is a part of the embodiment of the present invention, not all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0030] Example 1 like Figure 1 As shown, a crane-based method for handling materials outside a graphite calcining furnace comprises the following steps: S101: Collect data in the graphite calcination pit to obtain original point cloud data.

[0031] S102: Simplify the original point cloud data to obtain simplified point cloud data.

[0032] S103: Filter and reduce noise on the simplified point cloud data to obtain filtered and denoised point cloud data.

[0033] S104: performing point cloud segmentation on the point cloud data after filtering and denoising, and performing fitting and reconstruction based on the segmented point cloud data to obtain a three-dimensional twin point cloud.

[0034] S105: Controlling the suction pipe of the suction crane to suck the powdered filler near the calcined coke agglomerate outside the calcining furnace according to the three-dimensional twin point cloud.

[0035] Example 2 On the basis of the above embodiments, the embodiment of the present invention provides a specific process of a method for handling materials outside a graphite calcining furnace based on a crane, and the specific process is as follows: S201: Collect data in the graphite calcination pit to obtain original point cloud data.

[0036] Specifically, laser radar recognition is a new type of sensing technology that integrates laser measurement technology, radar detection technology and computer science technology. Laser radar recognition can obtain three-dimensional spatial information of the target through the scanning principle. It has high resolution, low noise and high detection accuracy, and is suitable for long-distance detection. The laser radar emits a laser beam. When the laser beam intersects with the surface of the target object, part of the light energy can be reflected back by the target surface, and the laser receiver receives the reflected signal. The three-dimensional coordinate space information, echo intensity and color information of the object to be measured are obtained through the received reflected signal, and the three-dimensional point cloud model of the object to be measured can be quickly reconstructed. Therefore, laser radar recognition technology is used to control the operation of the suction tube.

[0037] As an implementation method, lidar recognition technology collects data through time-of-flight ranging.

[0038] like Figure 2 As shown in the figure, the principle of the time-of-flight ranging method is to emit a laser pulse signal to the target object. By measuring the time difference between the emission and reflection of the pulse signal, that is, the flight time, the time for the signal to be emitted from the sensor to the target and then returned can be determined. The distance of the target object can be calculated based on the flight time and the laser speed. The distance S (unit: m) between the radar and the object being measured can be calculated according to Formula 1. The time-of-flight method can be divided into pulse modulation and continuous wave modulation according to the different modulation signals. The pulse modulation measurement method is simple, has a fast response speed, a wide ranging range, can reduce the interference of background light to a certain extent, and is widely used in disaster prevention, surveying and mapping, etc.

[0039] (1) in, is the round-trip time difference of the pulse, in seconds (s), v is the laser speed, which is approximately 3×108m / s, and S is the distance between the laser radar and the object.

[0040] As an implementation method, lidar recognition technology can also collect data through triangulation.

[0041] The principle of triangulation is as follows Figure 3 As shown, it uses an indirect measurement method, setting a laser emitter and a charge coupled device (CCD) at a known position. The two form a triangle structure with the object to be measured, and the distance D (unit: m) of the object to be measured is calculated according to triangle formula 2 according to the different imaging positions of objects at different distances. When the light is strong, the detector cannot effectively receive the reflected light, the error is large, and it is not suitable for long-distance detection.

[0042] (2) Among them, f is the distance between the receiving lens and the charge coupled device (linear CCD), L is the distance between the laser and the main optical axis, d is the distance between the main optical axis and the target at the projection position of the linear CCD, and D is the distance between the laser radar and the object.

[0043] As an implementation method, lidar recognition technology can also collect data through frequency modulated continuous wave ranging method.

[0044] The principle of frequency modulated continuous wave ranging (AMCW) is as follows Figure 4 As shown, by modulating the intensity of the light wave, a continuous wave signal with a constantly changing frequency is sent, and the frequency change of the echo signal is measured to calculate the distance between the target and the lidar.

[0045] The present invention adopts C16 laser radar, such as Figure 5As shown. The radar mainly uses a built-in laser probe to scan the surrounding environment 360 degrees at a speed of 10Hz. Through the time-of-flight method, the distance between the radar and the target is calculated, a high-precision map of the radar's surroundings can be drawn, and a 3D profile of the target is generated. It can see more accurately and farther than the relatively more popular visible light camera, and is not affected by light. The radar is widely used in the fields of autonomous driving, robot positioning and navigation, surveying and mapping, security, etc.

[0046] Specifically, the data in the graphite calcining pit include the surrounding environment of the calcining pit, the filling material data in the calcining pit, the graphite calcining furnace data, the calcining furnace cover data, and the distance between the suction pipe and the edge of the calcining pit, the calcining furnace and the calcining furnace cover.

[0047] Use a network cable to connect the radar to the host computer and power the laser radar. The laser radar scans the calcination pit to obtain the original point cloud data. After the host computer is connected to the radar, the radar configuration software displays the radar scanning information, such as Figure 6 As shown. The vertical angle, horizontal angle and distance parameters of the radar are obtained through the original point cloud data. The dispatch and distance information in polar coordinates are converted into the X, Y, and Z coordinates of the right-handed Cartesian coordinate system according to the following formula: (3) in, For distance, is the vertical angle, is the horizontal rotation angle correction angle, X, Y and Z are the projection coordinates of the polar coordinates on the x, y and z axes respectively, cos is the cosine function, and sin is the sine function.

[0048] S202: Simplify the original point cloud data to obtain simplified point cloud data.

[0049] Specifically, the raw point cloud data collected by the LiDAR may be very large and may contain various types of noise such as radar errors, obstructing objects or other interference factors, which consumes a lot of computing resources to process. In order to reduce the impact of uncertainty on target recognition performance, the raw point cloud data is preprocessed.

[0050] Since the amount of raw point cloud data collected by the radar is huge, directly filtering, denoising or segmenting the raw point cloud data will seriously affect the speed and efficiency of the recognition processing of the entire system. Therefore, the raw point cloud data is simplified, that is, the downsampling method is used.

[0051] Methods for simplifying point cloud data include voxelization, nearest neighbor sampling, random sampling, grid filtering, and geometric feature-based simplification methods. Voxelization is to divide the point cloud into voxel grids. The grid can be freely defined and covers the entire point cloud area. For each voxel, the center of gravity of all its points is calculated. Each voxel body is represented by a center of gravity. A new point cloud is reconstructed based on the obtained center of gravity, thereby achieving dimensionality reduction and compression of point cloud data. Nearest neighbor sampling searches for the nearest neighbors of points in the point cloud, finds the nearest points around it, and selects one or more of them as sampling results, which can reduce the amount of data while retaining the main features. Random sampling randomly selects a certain proportion of points in the point cloud as the sampling result, or collects point clouds at random intervals, but it may lead to an unbalanced sampling result, and the sampling ratio or interval needs to be adjusted continuously. Grid filtering processes the points in each grid unit by projecting the point cloud data onto a grid to reduce redundant points. The geometric feature-based simplification method selects representative points according to the geometric features of the point cloud, such as curvature and normal vector, but it requires complex algorithms to identify and process geometric features, which has a high computational cost.

[0052] After analyzing the actual scanning point cloud situation and work needs, the present invention chooses to use the voxelized downsampling processing method. First, the size of the voxelized grid is determined according to the number of scanned point clouds and the scanned space volume to ensure that the overall characteristics of the point cloud before and after processing remain unchanged. The side length L of the voxelized grid is calculated by the following formula: (4) Among them, a is the grid side length adjustment factor, s is the scale factor, g is the average number of point clouds per unit volume, and L is the side length of the voxelized grid.

[0053] The average number of point clouds per unit volume is calculated using the following formula: (5) Where N is the total number of point clouds, is the x-axis point cloud span, is the y-axis point cloud span, The z-axis point cloud span.

[0054] Finally, the side length of the voxelized grid is calculated according to the following formula: (6) By calculating the side length of the grid, all the original point cloud data is voxelized, and the centroid of the voxelized grid replaces all the point clouds in the voxelized grid. The processed point cloud can be downsampled to obtain streamlined point cloud data, and the original shape of the point cloud can be completely retained.

[0055] S203: Filter and reduce noise on the simplified point cloud data to obtain filtered and denoised point cloud data.

[0056] Specifically, point cloud data may be affected by various forms of noise during the acquisition process, and there may be some unnecessary fluctuations or spikes. These fluctuations may be caused by insufficient equipment power, insufficient measurement accuracy and frequency, lack of vertical angle, etc., or due to the influence of external unstable factors, resulting in strange and complex shapes of the scanned results. The invalid point clouds caused by these abnormal factors will cause problems in the subsequent work of segmentation and reconstruction. Therefore, it is necessary to use filtering methods to remove noise points and outliers. Commonly used filtering methods mainly include median filtering, statistical filtering, Gaussian filtering, and Sobel filtering.

[0057] The statistical filtering algorithm counts and analyzes all point clouds, selects query points and determines the standard point cloud number of neighboring points, calculates the average distance and standard deviation of neighboring points in the query point cloud, and defines the range. Points outside the range can be regarded as noise points or outliers and removed. Figure 7 As shown in the figure, the green circle is the valid point cloud, and the red circle is the discarded point cloud.

[0058] Perform neighborhood calculation on the streamlined point cloud. First, calculate the total number of point clouds n and the distance set between each point cloud and its k-neighborhood. Then calculate the mean and standard deviation of the distance between each point cloud and its k-neighborhood. Use formulas 7 and 8 to calculate the mean and standard deviation. Calculate the maximum threshold based on the mean and standard deviation. Finally, remove the points whose distance between the point cloud and the k-neighborhood is greater than the maximum threshold.

[0059] It is specifically expressed as follows: (7) (8) in, is the mean, n is the total number of point clouds, is the distance between the point cloud and its k-neighborhood, is the standard deviation.

[0060] The maximum threshold between the point cloud and its k-neighborhood is calculated based on the mean and standard deviation, as shown in Equation 9: (9) in, is the maximum threshold, is the standard deviation coefficient, The size of determines the threshold value range. For different point cloud distributions and different degrees of noise, The value of also needs to be changed. If the coefficient is too large, the denoising effect is not obvious. If the coefficient is too small, some useful point clouds will be removed as outliers. Therefore, it is necessary to select appropriate parameters to obtain a better filtering effect. Finally, the mean of each point cloud is compared with the maximum threshold, and the points greater than the maximum threshold are defined as outlier noise points and removed.

[0061] S204: performing point cloud segmentation on the point cloud data after filtering and denoising, and performing fitting and reconstruction based on the segmented point cloud data to obtain a three-dimensional twin point cloud.

[0062] Specifically, point cloud segmentation is the process of segmenting large point cloud data into multiple parts or categories, which is usually used to identify and separate different objects and scenes. The present invention uses region of interest (ROI) extraction to segment point cloud data. ROI usually analyzes and extracts specific attributes, targets or regions of interest.

[0063] The laser radar scans a wide area, scanning the entire workspace. The target location is in a relatively small fixed area. By setting a threshold, the point cloud data in this range is filtered and the point cloud data is regionally abstracted.

[0064] S205: Controlling the suction pipe of the suction crane to suck the powdered filler near the calcined coke agglomerate outside the calcining furnace according to the three-dimensional twin point cloud.

[0065] Specifically, the segmented point cloud data is converted into grid data in the MIRROR software, and the grid data is fitted with points, and the filler, calciner cover and calcined coke agglomerate are fitted. Then, the grid data is fitted with a straight line, and the fitting result is saved as a straight line. The angle difference between the straight line and the X-axis is calculated, and the coordinates and angles of the fitted grid are assigned. The grid coordinates are converted with the crane coordinates, and the scanned object is matched and overlapped with the grid to achieve positioning. The final point cloud data is transmitted to the twin system to generate a three-dimensional twin point cloud.

[0066] After calcination is stopped, the powdered filler on the top of the calcination pit will condense into calcined coke agglomerates, which will affect the suction pipe from sucking the powdered filler. Therefore, the host computer obtains the three-dimensional model of the graphite calcination pit based on the three-dimensional twin point cloud, locates the positions of the calcination furnace cover and the calcined coke agglomerates, and obtains the coordinates of the calcination furnace cover and the calcined coke agglomerates. The suction crane controls the suction pipe to suck the powdered filler on its edge at a distance of 50mm-90mm from the coordinates of the calcination furnace cover or the calcined coke agglomerates. This avoids the suction pipe from colliding with the calcination furnace cover and also avoids the suction pipe from sucking in the calcined coke agglomerates and causing the suction pipe to be blocked. And when the powdered filler around the calcined coke agglomerates is sucked, the calcined coke agglomerates will sink to the bottom of the calcination pit. When graphite is calcined next time, the high temperature will turn the calcined coke agglomerates at the bottom into powdered filler again, which will not affect the subsequent graphite calcination.

[0067] Example 3 Corresponding to the above method, such as Figure 8 As shown, an embodiment of the present invention provides a crane-based external material handling system for a graphite calcining furnace, comprising: The acquisition module is used to collect data in the graphite calcination pit and obtain the original point cloud data; A simplification module is used to simplify the original point cloud data to obtain simplified point cloud data; A filtering and denoising module is used to filter and denoise the simplified point cloud data to obtain the filtered and denoised point cloud data; The reconstruction module is used to segment the point cloud data after filtering and denoising, and to perform fitting and reconstruction based on the segmented point cloud data to obtain a three-dimensional twin point cloud; The control module is used to control the suction pipe of the suction crane to suck the powdered filling material near the calcined coke agglomerate outside the calcining furnace according to the three-dimensional twin point cloud.

[0068] It should be noted that the crane-based graphite calcining furnace external material handling system provided in the embodiment of the present invention is to realize the above-mentioned crane-based graphite calcining furnace external material handling method. Its specific functions can be referred to the above-mentioned method embodiments, which will not be repeated here.

[0069] In summary, the present invention simplifies the number of point cloud data by voxelizing and downsampling the point cloud data, calculates the average number of point clouds in a unit volume through the point cloud span of the x-axis, y-axis and z-axis, and then calculates the side length of the voxelized grid according to the average number of point clouds in the unit volume, so that the overall characteristics of the point cloud data before and after voxelization remain unchanged, reducing the amount of calculation. The present invention filters and de-noises the point cloud data through a statistical filtering algorithm, obtains the maximum threshold by calculating the mean and standard deviation of the distance between each point cloud and its k-neighborhood, and extracts the points whose distance between the point cloud and the k-neighborhood is greater than the maximum threshold, thereby achieving filtering and de-noising of the point cloud data and improving the accuracy of the point cloud data. The present invention obtains the coordinates of the calcining furnace cover and the calcined coke agglomerate based on the three-dimensional twin point cloud, so that the suction pipe absorbs the powdered filler at the edge of the calcining furnace cover and the calcined coke agglomerate, ensuring the stability of the automatic suction of the powdered filler, and avoiding the collision of the suction pipe with the calcining furnace cover and the calcined coke agglomerate and the clogging of the suction pipe by the calcined coke agglomerate.

[0070] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A crane-based method for handling materials outside a graphite calcining furnace, characterized in that: The following steps are involved: Step 1: Collect data from the graphite calcination pit to obtain original point cloud data; Step 2: Simplify the original point cloud data to obtain simplified point cloud data; Step 3: Filter and denoise the simplified point cloud data to obtain filtered and denoised point cloud data; Step 4: Segment the point cloud data after filtering and noise reduction, and perform fitting and reconstruction based on the segmented point cloud data to obtain a 3D twin point cloud; Step 5: Based on the 3D twin point cloud, the suction pipe of the suction crane is controlled to absorb the powdered filler near the calcined coke agglomerates outside the calciner.

2. The method for handling materials outside a graphite calcining furnace based on a crane according to claim 1, characterized in that: The specific process of step 2 includes: voxelize the original point cloud data, and determine the side length of the voxelized grid according to the amount of original point cloud data and the scanned space volume; The point cloud within the voxelized grid is replaced by the centroid of the voxelized grid to obtain the processed point cloud; The processed point cloud is downsampled to obtain streamlined point cloud data.

3. The method for handling materials outside a graphite calcining furnace based on a crane according to claim 2, characterized in that: The edge length of the voxelized grid is calculated as follows: Among them, N is the total number of point clouds, is the x-axis point cloud span, is the y-axis point cloud span, is the span of the z-axis point cloud, g is the average number of point clouds per unit volume, a is the grid side length adjustment factor, and s is the scale factor.

4. The method for handling materials outside a graphite calcining furnace based on a crane according to claim 1, characterized in that: The step three specifically includes: Calculate the distance set between each point cloud and its k-neighbors; Calculate the mean and standard deviation of the distance between each point cloud and its k-neighborhood based on the distance set; The maximum threshold is calculated based on the mean and standard deviation of the distance, and the point cloud whose distance to the k-neighborhood is greater than the maximum threshold is eliminated to obtain the point cloud data after filtering and denoising.

5. The method for handling materials outside a graphite calcining furnace based on a crane according to claim 4, characterized in that: The maximum threshold is calculated according to the following formula: in, is the mean, n is the total number of point clouds, is the distance between the point cloud and its k-neighborhood, is the standard deviation, is the standard deviation coefficient, is the maximum threshold.

6. The method for handling materials outside a graphite calcining furnace based on a crane according to claim 1, characterized in that: The step five also includes obtaining a three-dimensional model of the graphite calcining pit based on the three-dimensional twin point cloud, locating the position of the calcining furnace cover, and obtaining the coordinates of the calcining furnace cover; The suction crane controls the suction pipe to suck the powdered filler near the coordinates of the calcining furnace cover.

7. A crane-based external material handling system for a graphite calcining furnace, characterized in that: The following steps are involved: The acquisition module is used to collect data in the graphite calcination pit and obtain original point cloud data; The simplification module is used to simplify the original point cloud data to obtain simplified point cloud data; A filtering and denoising module is used to filter and denoise the simplified point cloud data to obtain filtered and denoised point cloud data; The reconstruction module is used to segment the point cloud data after filtering and noise reduction, and perform fitting and reconstruction based on the segmented point cloud data to obtain a 3D twin point cloud; The control module is used to control the suction pipe of the suction crane to absorb the powdered filler near the calcined coke agglomerates outside the calcining furnace according to the three-dimensional twin point cloud.

8. The crane-based external material handling system for a graphite calcining furnace according to claim 7, characterized in that: The streamlined module specifically includes: voxelize the original point cloud data, and determine the side length of the voxelized grid according to the amount of original point cloud data and the scanned space volume; The point cloud within the voxelized grid is replaced by the centroid of the voxelized grid to obtain the processed point cloud; The processed point cloud is downsampled to obtain streamlined point cloud data.

9. The crane-based external material handling system for a graphite calcining furnace according to claim 8, characterized in that: The edge length of the voxelized grid is calculated as follows: Among them, N is the total number of point clouds, is the x-axis point cloud span, is the y-axis point cloud span, is the span of the z-axis point cloud, g is the average number of point clouds per unit volume, a is the grid side length adjustment factor, and s is the scale factor.

10. The crane-based external material handling system for a graphite calcining furnace according to claim 7, characterized in that: The filtering and noise reduction module specifically includes: Calculate the distance set between each point cloud and its k-neighborhood; Calculate the mean and standard deviation of the distance between each point cloud and its k-neighborhood based on the distance set; The maximum threshold is calculated based on the mean and standard deviation of the distance, and the point cloud whose distance to the k-neighborhood is greater than the maximum threshold is eliminated to obtain the point cloud data after filtering and denoising.

Citation Information

Patent Citations

  • A low-rigidity workpiece assembly deformation prediction method based on point cloud data

    CN109918755A

  • Laser radar three-dimensional point cloud segmentation method

    CN110969624A

  • Distribution network operation robot control method and device based on digital twin system

    CN113954066A

  • Mine three-dimensional model point cloud data combined filtering method and medium

    CN117011175A

  • Method and device for achieving rapid cooling of high-temperature kiln based on movable material suction

    CN117570728A