High-precision plasma cutting method and system in radioactive environment

Abstract

The invention discloses a high-precision plasma cutting method and system in a radioactive environment, and relates to the technical field of nuclear industry decommissioning, and the method comprises the steps: collecting a two-dimensional image of a target workpiece, carrying out the splicing processing, obtaining a two-dimensional spliced image, carrying out the target recognition of the two-dimensional spliced image, and obtaining the positioning data of the target workpiece; collecting three-dimensional point cloud data of the target workpiece, and splicing the three-dimensional point cloud data to obtain a cutting boundary of the target workpiece; and obtaining an optimal target workpiece cutting track based on the positioning data and the cutting boundary of the target workpiece, and cutting the target workpiece by using a plasma cutting tool head according to the optimal target workpiece cutting track. According to the method, the high-precision image and point cloud data processing technology is adopted, the advanced template matching algorithm is combined, the optimal cutting path is generated for the to-be-cut objects of various complex structures, and therefore the cutting efficiency and quality are improved.

Description

Technical Field

[0001] The invention relates to the technical field of nuclear industry decommissioning, in particular to a high-precision plasma cutting method and system in a radioactive environment. Background Art

[0002] Nuclear facility decommissioning is an important task in the nuclear energy industry. Its core goal is to safely and efficiently dismantle and dismantle various equipment, components and pipelines in nuclear facilities. This process is of vital importance to reducing the generation of radioactive waste and ensuring environmental safety. However, the cutting and dismantling methods of existing technologies in nuclear facility decommissioning have several defects: The structures of the objects to be cut in nuclear facilities are complex and diverse, which is different from the regular structure objects commonly seen in the industrial field. This makes it difficult for traditional plasma cutting technology to adapt, and often leads to unreasonable cutting paths. Since nuclear facility decommissioning sites are usually radioactive, cutting operations need to be performed remotely, which increases the complexity of the operation. In particular, remote control becomes more difficult when the distance between the cutting torch and the cut surface is kept at an appropriate distance (5-8 mm). Conventional plasma jets are mostly transferred arcs, which may lead to reduced cutting efficiency and accuracy during remote control operations. Summary of the invention

[0003] In view of the problems existing in the existing high-precision plasma cutting method in a radioactive environment, the present invention is proposed.

[0004] Therefore, the problem to be solved by the present invention is how to achieve safe, efficient and precise cutting and dismantling of decommissioned nuclear facilities.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions: In a first aspect, an embodiment of the present invention provides a high-precision plasma cutting method in a radioactive environment, which includes collecting a two-dimensional image of a target workpiece and performing stitching processing to obtain a two-dimensional stitched image, performing target recognition on the two-dimensional stitched image, and obtaining positioning data of the target workpiece; collecting three-dimensional point cloud data of the target workpiece, performing stitching processing on the three-dimensional point cloud data, and obtaining a cutting boundary of the target workpiece; obtaining an optimal cutting trajectory of the target workpiece based on the positioning data and cutting boundary of the target workpiece, and using a plasma cutting tool head to cut the target workpiece according to the optimal cutting trajectory of the target workpiece.

[0006] As a preferred solution of the high-precision plasma cutting method in a radioactive environment of the present invention, the two-dimensional image of the target workpiece is collected and preprocessed, and the preprocessed two-dimensional image is stitched by ORB based on the ROI area to obtain a two-dimensional stitched image; the two-dimensional stitched image is converted from the RGB color space to the HSV color space, which is expressed as: , Among them, R represents red, G represents green, B represents blue, H represents hue, S represents saturation, V represents brightness, max represents the maximum value of grayscale in the three color channels of red, green and blue in the RGB image, and min represents the minimum value of grayscale in the three color channels of red, green and blue in the image; further, threshold segmentation based on hue is performed to extract the two-dimensional contour of the target workpiece; the workpiece shape in the DXF file is used as a template, and the extracted two-dimensional contour of the target workpiece is matched with the template to obtain the target workpiece positioning data.

[0007] As a preferred solution of the high-precision plasma cutting method in a radioactive environment described in the present invention, the matching is performed based on template matching based on the halcon image processing library.

[0008] As a preferred solution of the high-precision plasma cutting method in a radioactive environment described in the present invention, the three-dimensional point cloud data of the target workpiece is collected and preprocessed; the preprocessed three-dimensional point cloud data is roughly spliced ​​based on the covariance matrix to obtain three-dimensional point cloud rough splicing data; the point cloud is set and point cloud For a pair of point clouds with overlapping areas, calculate the point clouds separately and point cloud The covariance matrix of is: , in, represents the centroid of point cloud X, Represents the centroid of point cloud Y; x i represents the i-th point in the point cloud X, y j represents the jth point in the point cloud Y, M x , M y Point Cloud and point cloud The covariance matrix of Mx; Assume that the eigenvectors corresponding to the eigenvalues ​​of Mx are p 1 , p 2 , p 3 , My The eigenvalues ​​corresponding to the eigenvectors are q 1 , q 2 , q 3 ; According to the covariance matrix theory, the rigid transformation matrix of point cloud X and point cloud Y is obtained as: , in, R 0 represents the rotation matrix, T 0 represents the translation matrix; get the transformation between two point clouds Based on the iterative nearest algorithm, the three-dimensional point cloud rough stitching data is finely stitched to obtain the three-dimensional point cloud fine stitching data; based on the target workpiece positioning data and the three-dimensional point cloud fine stitching data, the cutting boundary of the target workpiece is obtained.

[0009] As a preferred solution of the high-precision plasma cutting method in a radioactive environment described in the present invention, wherein: a plurality of features to be cut are obtained based on the positioning data of the target workpiece and the cutting boundary of the target workpiece; the starting cutting point and the ending cutting point of each feature to be cut are determined, and each feature to be cut is connected step by step through the cutting points. For a non-final feature to be cut, the distance between its ending cutting point and the starting cutting point of the feature to be cut at the next level is smaller than the distance between its ending cutting point and the starting cutting points of other features to be cut, thereby forming an optimal cutting trajectory for the target workpiece.

[0010] As a preferred solution of the high-precision plasma cutting method in a radioactive environment described in the present invention, the feature to be cut includes one or more of a straight line feature, an arc feature, a polygonal feature and a circular feature.

[0011] As a preferred solution of the high-precision plasma cutting method in a radioactive environment described in the present invention, wherein: the optimal target workpiece cutting trajectory data is DXF file data.

[0012] In the second aspect, an embodiment of the present invention provides a high-precision plasma cutting system in a radioactive environment, which includes: a positioning module, used to perform target recognition on a two-dimensional stitching image of a target workpiece and stitching processing on three-dimensional point cloud data to obtain positioning data and cutting boundaries of the target workpiece; an optimization module, used to obtain an optimal cutting trajectory of the target workpiece based on the positioning data and cutting boundaries of the target workpiece; a cutting module, in which the plasma cutting tool head of the nuclear decommissioning robot cuts the target workpiece according to the optimal cutting trajectory of the target workpiece.

[0013] In a third aspect, an embodiment of the present invention provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, wherein: when the processor executes the computer program, any step of the above-mentioned high-precision plasma cutting method in a radioactive environment is implemented.

[0014] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, any step of the above-mentioned high-precision plasma cutting method in a radioactive environment is implemented.

[0015] The beneficial effect of the present invention is that the present invention generates the optimal cutting path for various complex structures of the objects to be cut by adopting high-precision image and point cloud data processing technology, combined with advanced template matching algorithm, so as to improve the cutting efficiency and quality. The nuclear decommissioning robot equipped with a plasma cutting tool head is used, combined with accurate positioning data, to achieve precise control of the cutting tool head, and even under remote control operation conditions, the appropriate distance between the cutting torch and the cut surface can be maintained. It can adapt to nuclear facility decommissioning objects of different shapes and structures, and has high adaptability and flexibility. It also reduces the potential impact on the environment, which helps to achieve environmental friendliness in the decommissioning of nuclear facilities. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. Among them: Figure 1 is a schematic flow chart of a method for high-precision plasma cutting in a radioactive environment described in an embodiment; Figure 2 Schematic diagram of transformation of each coordinate system in the embodiment; Figure 3 is a schematic diagram of determining the starting point of straight line feature processing in an embodiment; Figure 4 Schematic diagram of determining the starting point of arc feature processing in the embodiment; Figure 5 Schematic diagram of determining the starting point of polygon feature processing in the embodiment; Figure 6 is a schematic diagram of determining the starting point of circular feature processing in an embodiment; Figure 7 1. It is a schematic diagram of the processing optimization algorithm flow in the embodiment; Figure 8 It is the code structure diagram of DXF file; Fig. 9It is a schematic diagram of the process of redrawing a pattern after reading a DXF file in an embodiment. DETAILED DESCRIPTION

[0017] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0018] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0019] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0020] Example 1

[0021] This embodiment provides a high-precision plasma cutting method in a radioactive environment. The method is implemented based on a nuclear decommissioning robot equipped with a plasma cutting tool head. The nuclear decommissioning robot is provided with a six-axis mechanical arm, and the plasma cutting tool head is installed at the end of the six-axis mechanical arm. The nuclear decommissioning robot described in this embodiment also has a two-dimensional camera and a three-dimensional camera, which are used to collect a two-dimensional image of the target workpiece and three-dimensional point cloud data of the target workpiece respectively. The two-dimensional image can obtain better results than the three-dimensional point cloud in terms of workpiece target recognition, and the three-dimensional point cloud processing can obtain the height information of the workpiece, obtain a high-precision cutting trajectory, effectively ensure the distance between the plasma cutting head and the surface of the cutting workpiece, improve the cutting quality, and prevent the plasma cutting head from being welded to the surface of the cutting workpiece due to insufficient cutting distance.

[0022] Reference Figure 1 to Figure 8 , which is the first embodiment of the present invention, provides a high-precision plasma cutting method in a radioactive environment, comprising: S1: Collect the two-dimensional image of the target workpiece and perform stitching processing to obtain a two-dimensional stitching image, perform target recognition on the two-dimensional stitching image, and obtain the positioning data of the target workpiece.

[0023] In this embodiment, camera calibration and hand-eye calibration are performed. The transformation matrix from the camera coordinate system to the calibration plate coordinate system and the transformation matrix between the camera coordinate system and the robot base coordinate system are obtained; The ultimate goal of eye-in-hand calibration is to solve the rigid transformation relationship between the camera coordinate system and the robot tool coordinate system.

[0024] like Figure 2 As shown in the figure, camera is used to represent the camera coordinate system at the end of the robot arm, tool is used to represent the robot tool coordinate system, base is used to represent the robot base coordinate system, and object is used to represent the calibration plate coordinate system. The robot base coordinate system is a rectangular coordinate system fixed on the robot mounting base, which is used to describe the movement of the robot body. When the robot is at position 1, the relationship between the robot tool coordinate system tool1 and the robot base coordinate system is expressed as: , The relationship between the robot tool coordinate system tool1 and the camera coordinate system camera1 is expressed as: , The camera coordinate system camera1 and the calibration plate coordinate system object satisfy the following formula: , Combining the above three equations, we can get: , After moving to position 2, we get: , Combining the above two equations, we can get: , in, Represents the transformation relationship from the camera coordinate system to the robot tool coordinate system, Represents the transformation relationship from the robot base coordinate system to the calibration plate coordinate system, Represents the transformation relationship from the robot base coordinate system to the robot tool coordinate system. Indicates the conversion relationship from the camera coordinate system to the calibration plate coordinate system. Indicates the transformation relationship from the base coordinate system to the robot tool coordinate system when the robot is at position 1. Indicates the transformation relationship from the base coordinate system to the robot tool coordinate system when the robot is at position 2. Indicates the conversion relationship from the camera coordinate system to the calibration plate coordinate system when the robot is in position 1. Indicates the conversion relationship from the camera coordinate system to the calibration plate coordinate system when the robot is in position 2.

[0025] In the above formula, only is an unknown quantity, so the solution of hand-eye calibration can be converted into solving the linear equation system AX=BX, and the hand-eye calibration result can be obtained by solving the linear equation system.

[0026] Before the two-dimensional images of the target workpiece are stitched together, the two-dimensional images of the target workpiece are preprocessed, including correction and filtering, to reduce the interference of illumination and noisy background, in preparation for image stitching, wherein the correction is implemented based on a transformation matrix.

[0027] Afterwards, ORB image stitching is performed on the preprocessed two-dimensional image based on the ROI region to obtain a two-dimensional stitched image.

[0028] Since a two-dimensional camera cannot generally obtain a panoramic view of the entire work platform in a single shot, it can only be achieved through image stitching. By collecting images of different positions of the target workpiece multiple times and then stitching the images, a panoramic view of the target workpiece can be finally obtained.

[0029] Since the ORB algorithm has certain feature invariance, it can identify the same object under the premise of certain illumination changes, translation changes, rotation changes, and scaling changes in the image. The core idea of ​​the algorithm is to use the improved FAST (oriented FAST) algorithm to detect corners on the image pyramid and generate the main direction, so that the feature points have a certain scale invariance. The feature points are described based on the BRIEF algorithm, so that the feature point descriptors have a certain scale invariance. Finally, the RANSAC (Random Sample Consensus) algorithm is used to purify the feature points, find matching feature points to achieve image registration, and use weighted fusion to eliminate the stitching gaps and achieve image stitching.

[0030] In this embodiment, the two-dimensional stitched image is converted from the RGB color space to the HSV color space, and a threshold segmentation based on the hue is performed to extract the two-dimensional contour of the target workpiece.

[0031] The images obtained from the camera are all RGB images, and contour extraction based on the HSV color space requires converting the image from the RGB color space to the HSV color space. First, calculate the maximum and minimum grayscale values ​​in the red, green, and blue color channels of the original RGB image, and then obtain the corresponding values ​​of the hue channel, saturation channel, and brightness channel according to the following four formulas to convert the image in the RGB color space to the HSV color space: , Among them, R represents red, G represents green, and B represents blue; In the formula: ifR=max, it means: if (if) R is the largest value among the three channels (R represents red, G represents green, B represents blue, and the larger R is, the redder the image is), then H (hue) is equal to (GB) / (max-min).

[0032] H stands for hue, which represents the basic property of color and is measured in degrees, ranging from 0° to 360°, where red corresponds to 0°, green corresponds to 120°, and blue corresponds to 240°. The colors on the hue circle are arranged in spectral order, forming a color wheel.

[0033] S stands for saturation, which indicates the purity or depth of the color, and its value range is from 0% to 100%. The higher the saturation, the brighter the color; the lower the saturation, the closer the color is to gray.

[0034] V stands for brightness, which indicates the brightness of the color, and its value range is also 0% to 100%. The higher the brightness, the brighter the color; the lower the brightness, the darker the color.

[0035] After converting the stitched image into HSV space, the threshold segmentation based on hue can intuitively distinguish the workpiece from the background, that is, the two-dimensional outline of the target workpiece is obtained.

[0036] The workpiece shape in the DXF file is used as a template, and the extracted two-dimensional contour of the target workpiece is matched with the template to obtain the target workpiece positioning data.

[0037] In this embodiment, based on the workpiece contour, template matching of the workpiece and positioning of the target workpiece can achieve a higher matching rate and position accuracy.

[0038] The principle of shape-based template matching is to generate a contour template by extracting the contour features of the known target shape, and then calculate the matching scores of all contours through the corresponding matching function. The one with the highest score is the target workpiece. Each workpiece has a corresponding shape design drawing at the beginning of the design. Based on the design drawing, the DXF file is exported through the CAD software, and then the contour template is generated by reading the DXF file to achieve a better recognition effect. Since information such as contour features is insensitive to lighting changes, template matching is more robust.

[0039] The halcon image processing library is used as a template matching tool. The matching process of the DXF template based on the halcon image processing library is as follows: Import DXF model into DXF file, a saved format for CAD files, which contains workpiece contour information, which can be used as a template for template matching algorithm to prepare for subsequent processing; Since the size of the image outline in the DXF file does not match the size of the acquired image, the outline ratio needs to be adjusted to improve the outline matching rate; Read the workpiece contour image obtained by the above steps as the image to be matched; The imported DXF file is used as a template, and the read work platform workpiece contour is used as the image to be matched for search, so as to obtain the target workpiece template matching coefficient and the workpiece center of mass position, that is, the target workpiece positioning data.

[0040] Output the matching area and matching score for subsequent use.

[0041] S2: Collect the three-dimensional point cloud data of the target workpiece, perform splicing processing on the three-dimensional point cloud data, and obtain the cutting boundary of the target workpiece therefrom.

[0042] Collect the 3D point cloud data of the target workpiece and perform preprocessing.

[0043] After the hand-eye calibration of the 3D camera, the acquired point cloud needs to be filtered to reduce the amount of subsequent stitching calculations. Due to the acquisition method and the influence of the external environment, the directly acquired point cloud data contains a large number of noise points, outliers, holes and other erroneous information. If these noise points are not processed, it will directly affect the subsequent point cloud registration results and cause the algorithm to fall into a local optimal solution. Point cloud filtering is the first step in point cloud preprocessing. Its function is to remove the above noise points so that subsequent point cloud processing can be carried out better. Since the point cloud collected by the 3D camera carried by the nuclear decommissioning robot is an ordered data, Gaussian filtering is used to process the collected point cloud.

[0044] The preprocessed three-dimensional point cloud data is roughly spliced ​​based on the covariance matrix to obtain three-dimensional point cloud rough splicing data.

[0045] The point cloud rough stitching method based on the covariance matrix can stitch the point clouds to be stitched together. Although the accuracy is not high, it reduces the amount of calculation for the subsequent point cloud fine stitching and can effectively prevent the point cloud fine stitching from falling into the local optimal solution.

[0046] The rough stitching of point clouds is a method of transforming two point clouds with the same local features to the same global coordinate system under the premise of low precision, in order to obtain better initial iteration values ​​for subsequent fine stitching. According to the theoretical description of the covariance matrix, a point cloud can be described by a covariance matrix, and the eigenvalues ​​and eigenvectors of the covariance matrix can be used to characterize the characteristics of the point cloud. The point cloud is processed using the matrix properties to obtain the transformation matrix between the two point clouds. The basic principle of the rough stitching of point clouds based on the covariance matrix is ​​to obtain the covariance matrix representing the two point clouds according to the centroid positions of the two point clouds, and then calculate the rigid transformation matrix between the two point clouds according to the eigenvalues ​​and eigenvectors of the covariance matrix to achieve rough stitching of point clouds.

[0047] The rough stitching of point clouds based on the covariance matrix can be calculated as follows: Set point cloud and point cloud For a pair of point clouds with overlapping areas, calculate the point clouds separately and point cloud The covariance matrix of is: , in, represents the centroid of point cloud X, Represents the centroid of the point cloud Y. , The x and y in are the coordinates used to describe the position of a point in the point cloud, and ( =1~m) refers to the number of points.

[0048] In order to obtain the transformation matrix between two point clouds, it is necessary to find M x and M y The eigenvalue of M x The eigenvalues ​​corresponding to the eigenvectors are p 1 , p 2 , p 3 , M y The eigenvalues ​​corresponding to the eigenvectors are q 1 , q 2 , q 3 , According to the covariance matrix theory, the rigid transformation matrix of point cloud X and point cloud Y is obtained as: , Based on the iterative nearest algorithm, the three-dimensional point cloud rough stitching data is finely stitched to obtain the three-dimensional point cloud fine stitching data.

[0049] The accuracy of the obtained three-dimensional point cloud rough stitching data generally cannot meet the cutting requirements, and fine stitching is required to meet the accuracy requirements. The point cloud fine stitching method adopted in this embodiment is the iterative closest algorithm (ICP). The main idea of ​​the ICP algorithm is to calculate the optimal rigid transformation matrix between the original point cloud and the point cloud to be stitched through the objective function to realize point cloud stitching. Based on the idea of ​​least squares method, the iteration is finally realized by calculating the error function value. Generally speaking, the objective function can be realized by the distance between points, the distance from points to lines, and the distance from points to tangent planes.

[0050] The process of stitching point clouds using the ICP algorithm consists of three steps: Search for corresponding points in the original point cloud and the point cloud to be spliced; Remove incorrect matching point pairs; Solve the rigid transformation matrix of two point clouds and implement iteration.

[0051] The cutting boundary of the target workpiece is obtained based on the target workpiece positioning data and the three-dimensional point cloud precision stitching data, including: obtaining the point cloud outer contour of the workpiece upper surface point cloud after stitching to obtain the surface outer contour, extracting all the workpiece upper surface point cloud contour points, using the target workpiece positioning data to mark the single workpiece upper surface point cloud, and then segmenting the single workpiece upper surface outer contour, scaling the single workpiece upper surface point cloud outer contour based on the point cloud centroid to obtain the cutting boundary of the target workpiece.

[0052] S3: Obtain an optimal cutting trajectory of the target workpiece according to the target workpiece positioning data and the cutting boundary of the target workpiece, and transmit the optimal cutting trajectory data of the target workpiece to the nuclear decommissioning robot through TCP / IP communication.

[0053] According to the positioning data of the target workpiece and the cutting boundary of the target workpiece, a number of features to be cut are obtained, and the starting cutting point and the ending cutting point of each feature to be cut are determined. Then, each feature to be cut is connected step by step through the cutting points. For a non-final feature to be cut, the distance between its ending cutting point and the starting cutting point of the feature to be cut at the next level is smaller than the distance between its ending cutting point and the starting cutting points of other features to be cut, thereby forming the optimal cutting trajectory of the target workpiece.

[0054] The cutting features on the plane can be generally divided into closed features and open features. Closed features are connected from beginning to end, and the cutting start point and cutting end point of the feature are at the same point; open features have different beginnings and ends, and the cutting start point and cutting end point are not the same point.

[0055] The features to be cut can be straight line features, arc features, polygonal features and circular features.

[0056] For the cutting of straight line features, Figure 3 As shown, both ends of the straight line can be used as the starting point for cutting the feature, and the distance from the point P where the cutting tool head is located to the two ends of the straight line is d 0 , d 1 , determine the closest point Q 1 As a starting point, another point Q 0 For the end point.

[0057] For the cutting of arc segment features, Figure 4 As shown, similar to the straight line feature, compare the distance between point P and the two end points of the arc, and select the point with the shorter distance as the cutting starting point.

[0058] The cutting path of the polygon is as follows Figure 5 As shown in the figure, a polygon is a graphic feature composed of multiple line segments connected end to end. For cutting this feature, the intersection of two line segments is selected as the starting point and end point of the cutting trajectory. The polygon shown in the figure is a regular pentagon. This feature has 5 corner points. By comparing the distance from the last cutting end point to these 5 corner points, the corner point with the shortest distance is selected. Q 4 The starting point of the cut is the feature, and the end point coincides with this point.

[0059] like Figure 6 The circular feature shown is determined by the center and radius. For this feature, the starting position P of the cutting tool head is connected to the center O of the circle, and the intersection of this line segment and the circle is Q 0 The starting and ending points of the cut.

[0060] Due to the limitations of the cutting environment in a radioactive environment, manually controlled robot cutting is usually used. However, due to other factors such as the robot's arm span, a complete cutting process usually includes multiple cutting steps.

[0061] The overall process of the processing optimization algorithm in this embodiment is as follows: Figure 7 As shown in the figure, the current position of the cutting tool head is used as the starting position of the overall cutting path, the first feature end point is set as the origin O (0, 0), and then O is assigned to P, and the distance between the previous feature end point P and the remaining single feature endpoints is compared. For open features, the point with a shorter distance is set as the starting point and the other point as the end point. For closed features, the determined feature point is used as both the starting point and the end point. Compare the distance between point P and the starting points of all remaining features, select the route with the shortest distance as the next processing route, and update the cutting end point of the connected feature to the new point P, and iterate to obtain the final path.

[0062] In this embodiment, the optimal target workpiece cutting trajectory output result is a DXF file after algorithm optimization.

[0063] The motion path of the robot arm holding the cutting tool needs to be planned by the user, and the end of the robot arm can move along the specified path when cutting by generating spatial discrete target points that the robot arm can recognize. Therefore, the planning of the cutting trajectory needs to generate DXF reading data through modeling. DXF (AutoCAD Drawing Exchange Format) file is a vector drawing file format. It is an intermediate format for data exchange between various CAD software for plane drawing. It uses ASCII code as a storage method. Therefore, secondary development of the original drawing can be achieved by parsing the information in the DXF file. The encoding of the DXF file is a combination of group code and group value in pairs. By observation, it can be found that DXF files are generally composed of 7 coding segments, namely, header segment (HEADER), class segment (CLASSES), table segment (TABLES), block segment (BLOCKS), entity segment (ENTITIES), object segment (OBJECTS), and graphic preview segment (THUMBNAILIMAGE).

[0064] Each DXF file consists of a group code 0 and SECTION combination to introduce each segment, followed by a group code 2 and the name of each segment. Each segment represents a different feature type in the drawing. The group code 0 and ENDSEC combination indicates the end of a segment. When all the segment codes representing the features are completed, the group code 0 and EOF are used to indicate the end of the entire file. The code structure of the DXF file can be used Figure 8 The whole file can be read using high-level languages ​​such as C++. By locating the position of each segment and reading the content in the parsed segment, the required information can be obtained for subsequent reconstruction and secondary development of drawings.

[0065] After reading the DXF file using a high-level programming language, the coded data behind the graphics is obtained, and then the pattern is redrawn according to the read entity segment code. The specific process is as follows Fig. 9 As shown, LINE represents a straight line, CIRCLE represents a circle, ARC represents an arc, and ELLIPSE represents a compound line.

[0066] The six-axis robot arm of the nuclear decommissioning robot is used to move according to the optimal cutting trajectory of the target workpiece, so that the plasma cutting tool head cuts the target workpiece according to the optimal cutting trajectory of the target workpiece.

[0067] In summary, the present invention adopts high-precision image and point cloud data processing technology, combined with advanced template matching algorithm, and can generate the optimal cutting path for various complex structures of objects to be cut, thereby improving cutting efficiency and quality. Using a nuclear decommissioning robot equipped with a plasma cutting tool head, combined with precise positioning data, precise control of the cutting tool head is achieved, and the appropriate distance between the cutting torch and the cut surface can be maintained even under remote control operation conditions. The design and control strategy of the plasma cutting tool head are optimized for remote control operation, ensuring the stability and consistency of the plasma jet and improving the accuracy and efficiency of cutting. It can adapt to nuclear facility decommissioning objects of different shapes and structures, and has high adaptability and flexibility. By reducing the time that personnel are directly exposed to the radioactive environment, the present invention significantly improves the safety of operation. Through precise cutting path planning, it helps to reduce unnecessary material waste, which is in line with the principle of waste minimization, and the precise cutting technology reduces the potential impact on the environment, which helps to achieve the environmental friendliness of nuclear facility decommissioning.

[0068] Example 2 Reference Figure 1 Based on the first embodiment, this embodiment further provides a high-precision plasma cutting system in a radioactive environment, comprising: A positioning module is used to perform target recognition on the two-dimensional stitching image of the target workpiece and stitching processing on the three-dimensional point cloud data to obtain the positioning data and cutting boundary of the target workpiece; An optimization module, used for obtaining an optimal cutting trajectory of the target workpiece according to the positioning data and cutting boundary of the target workpiece; Cutting module: the plasma cutting tool head of the nuclear decommissioning robot cuts the target workpiece according to the optimal cutting trajectory of the target workpiece.

[0069] This embodiment also provides a computer device suitable for the case of a high-precision plasma cutting method in a radioactive environment, comprising a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute computer-executable instructions to implement the high-precision plasma cutting method in a radioactive environment proposed in the above embodiment.

[0070] The computer device may be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen and an input device connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. The display screen of the computer device may be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device may be a touch layer covering the display screen, or a key, trackball or touchpad provided on the housing of the computer device, or an external keyboard, touchpad or mouse, etc.

[0071] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, the method for high-precision plasma cutting in a radioactive environment as proposed in the above embodiment is implemented.

[0072] The storage medium proposed in this embodiment and the data storage method proposed in the above embodiment belong to the same inventive concept. The technical details not fully described in this embodiment can be found in the above embodiment, and this embodiment has the same beneficial effects as the above embodiment.

[0073] 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 preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A high-precision plasma cutting method in a radioactive environment, characterized in that: include, Collecting two-dimensional images of the target workpiece and performing stitching processing to obtain a two-dimensional stitched image, performing target recognition on the two-dimensional stitched image, and obtaining positioning data of the target workpiece; Collect the three-dimensional point cloud data of the target workpiece, perform splicing processing on the three-dimensional point cloud data, and obtain the cutting boundary of the target workpiece; An optimal cutting trajectory of the target workpiece is obtained based on the positioning data and the cutting boundary of the target workpiece, and the target workpiece is cut according to the optimal cutting trajectory of the target workpiece using a plasma cutting tool head.

2. The high-precision plasma cutting method in a radioactive environment as claimed in claim 1, characterized in that: Acquire a two-dimensional image of the target workpiece and perform preprocessing, and perform ORB image stitching on the preprocessed two-dimensional image based on the ROI area to obtain a two-dimensional stitched image; Convert the two-dimensional stitched image from RGB color space to HSV color space, expressed as: ， Where R represents red, G represents green, B represents blue, H represents hue, S represents saturation, V represents brightness, max represents the maximum grayscale value of the red, green and blue color channels in the RGB image, and min represents the minimum grayscale value of the red, green and blue color channels in the image; Further, the threshold segmentation based on the hue is performed to extract the two-dimensional contour of the target workpiece; The workpiece shape in the DXF file is used as a template, and the extracted two-dimensional contour of the target workpiece is matched with the template to obtain the target workpiece positioning data.

3. The high-precision plasma cutting method in a radioactive environment as claimed in claim 2, characterized in that: The matching is based on template matching by the halcon image processing library.

4. The high-precision plasma cutting method in a radioactive environment as claimed in claim 3, characterized in that: Collecting three-dimensional point cloud data of the target workpiece and performing preprocessing; Based on the covariance matrix, the pre-processed three-dimensional point cloud data is roughly spliced ​​to obtain three-dimensional point cloud rough splicing data; Set point cloud and point cloud For a pair of point clouds with overlapping areas, calculate the point clouds separately and point cloud The covariance matrix of is: ， in, represents the centroid of point cloud X, represents the centroid of point cloud Y, x i represents the i-th point in the point cloud X, y j represents the jth point in the point cloud Y, M x , M y Point Cloud and point cloud The covariance matrix of Assumptions M x The eigenvalues ​​corresponding to the eigenvectors are p 1 , p 2 , p 3 , M y The eigenvalues ​​corresponding to the eigenvectors are q 1 , q 2 , q 3 ; According to the covariance matrix theory, the rigid transformation matrix of point cloud X and point cloud Y is obtained as: ， in, R 0 represents the rotation matrix, T 0 means translation matrix; Based on the iterative nearest algorithm, the three-dimensional point cloud rough stitching data is finely stitched to obtain the three-dimensional point cloud fine stitching data; The cutting boundary of the target workpiece is obtained based on the target workpiece positioning data and the three-dimensional point cloud precise stitching data.

5. The high-precision plasma cutting method in a radioactive environment as claimed in claim 4, characterized in that: Acquire a plurality of features to be cut based on the positioning data of the target workpiece and the cutting boundary of the target workpiece; Determine the starting cutting point and the ending cutting point of each feature to be cut, and connect each feature to be cut step by step through the cutting points. For a non-final feature to be cut, the distance between its ending cutting point and the starting cutting point of the feature to be cut at the next level is smaller than the distance between its ending cutting point and the starting cutting points of other features to be cut, thus forming the optimal cutting trajectory of the target workpiece.

6. The high-precision plasma cutting method in a radioactive environment as claimed in claim 5, characterized in that: The features to be cut include one or more of a straight line feature, an arc feature, a polygonal feature and a circular feature.

7. The high-precision plasma cutting method in a radioactive environment as claimed in claim 6, characterized in that: The optimal target workpiece cutting trajectory data is DXF file data.

8. A high-precision plasma cutting system in a radioactive environment, based on the high-precision plasma cutting method in a radioactive environment according to any one of claims 1 to 7, characterized in that: include, A positioning module is used to perform target recognition on the two-dimensional stitching image of the target workpiece and stitching processing on the three-dimensional point cloud data to obtain the positioning data and cutting boundary of the target workpiece; An optimization module, used for obtaining an optimal cutting trajectory of the target workpiece according to the positioning data and cutting boundary of the target workpiece; Cutting module: the plasma cutting tool head of the nuclear decommissioning robot cuts the target workpiece according to the optimal cutting trajectory of the target workpiece.

Images