Three-dimensional CT internal defect imaging recognition system

Abstract

A three-dimensional CT internal defect imaging identification system disclosed by the present invention comprises a front-end device and a control platform, the front-end device comprises an X-ray machine and an imaging plate, and the X-ray machine and the imaging plate are both installed on a mechanical flexible electric device. The mechanical flexible device is used for adjusting the position of the front-end equipment so as to ensure that a detected position and the front-end equipment are located on the same horizontal plane, the control platform is used for imaging control and data output of the front-end equipment and movement of the mechanical flexible device, and data obtained by the operation platform is transmitted to the equipment health state image data platform for imaging processing. According to the method, a tool and a tool linkage device suitable for the GIS equipment operation environment are adopted, the types of defects in the GIS pipeline are divided based on experience in actual production, and a neural network model is built based on the experience; and aiming at the actual condition of GIS operation, a circular accelerator is adopted to replace a closed tube X-ray source, so that the recognition capability of the system is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of image recognition, and specifically to a three-dimensional CT internal defect imaging recognition system. Background Art

[0002] Gas Insulated Metal Enclosed Transmission Line (GIS) is an underground power transmission method, which has the advantages of safety, reliability, high efficiency, intensiveness, intelligence, etc. GIS is a power transmission device composed of a grounded alloy aluminum shell and an internal tubular alloy aluminum conductor and using insulating gases such as sulfur hexafluoride as the insulating medium. GIS has the advantages of large transmission capacity, low unit loss, flexible layout, high operation reliability, long service life, and being unaffected by external environmental factors, and is suitable for power transmission occasions with harsh climate environments or restricted corridor selections.

[0003] Since GIS is large in volume and the devices are inside a metal-sealed cavity, direct visual inspection cannot be carried out, and the on-site conditions do not allow equipment disassembly, so on-site "visual" inspection and troubleshooting are required. Currently, on-site "visual" inspection and troubleshooting generally use X-ray non-destructive testing technology. The detection system of this technology consists of several parts such as an X-ray source, a sample, a detector, and an image workstation; the detection methods can be divided into two types: CT and DR. CT is Industrial Computed Tomography (abbreviated as CT), and DR refers to Digital Radiography (abbreviated as DR). The continuous development of these two technologies in recent years has provided a more scientific and objective basis for the defect detection of industrial products. When using industrial CT images and DR images to detect sample defects, it is mainly to locate the sample defects in the images and obtain information such as geometric parameters of the defects from a two-dimensional or three-dimensional perspective, so as to judge whether the sample can meet the requirements according to the type and size of the defects.

[0004] DR obtains a two-dimensional image in a single direction; while CT can obtain the three-dimensional structure of the defect. Relatively speaking, the three-dimensional structure diagram is more objective and comprehensive than the two-dimensional one, and the defect situation understood is more specific. This is the advantage of CT compared with DR.

[0005] The prior art uses an X-ray source and a detector to move around the GIS device. However, the prior art mostly uses a hoisting method for the arrangement of the ray source and the detector, and this method can only rely on manual experience for placement and alignment. The existing methods can only image the cross-section of the GIS device and obtain a two-dimensional image in a single direction, and experienced operators are required to judge whether there are problems inside the GIS pipeline. If the experience of the operator is insufficient, it will cause human errors and affect the operation of the substation.

[0006] The deficiencies of the prior art are as follows: 1) There is a lack of special tooling for three-dimensional CT scanning equipment in the dangerous and narrow environment of substations; 2) There is a lack of methods for evaluating the types and positioning of internal defects in GIS equipment, and there is a lack of an intuitive imaging system; 3) Traditional closed-tube X-ray sources have problems that X-rays cannot penetrate GIS equipment and the irradiation range is insufficient. Summary of the Invention

[0007] The present invention is to solve the above-mentioned deficiencies of the prior art, and proposes a three-dimensional CT internal defect imaging and recognition system for the operation and maintenance of GIS equipment. It is mainly aimed at the operation and maintenance process before and during the operation of GIS equipment, and provides an intuitive, accurate and safe internal defect imaging and recognition system to solve the problems such as the lack of an intuitive imaging system mentioned in the background technology.

[0008] To achieve the above object, the present invention provides the following technical solutions: A three-dimensional CT internal defect imaging and recognition system includes a front-end device and a control platform. The front-end device includes an X-ray machine and an imaging plate. Both the X-ray machine and the imaging plate are installed on a mechanical flexible electric device. The mechanical flexible device is used to adjust the position of the front-end device to ensure that the measured position is on the same horizontal plane as the front-end device. The control platform is used for imaging control and data output of the front-end device and the movement of the mechanical flexible device. The data obtained by the control platform is transmitted to the equipment health status image data platform for imaging processing.

[0009] As a further technical solution of the present invention: The imaging processing includes the following steps: Step 1, overlapping imaging: Obtain multiple X-ray images with image intersections, number them in the order of adjacent relationships, and identify whether the main image parameters meet the splicing conditions; Step 2, feature extraction: Extract the main image into feature data formed by single pixels through gray value differences, and store it as a feature mapping image; Step 3, feature matching and image fusion: By reading the corresponding feature mapping images of the main image, perform feature data matching on the feature data within the image intersection range and the corresponding feature mapping images of any adjacent splicing images until all images are spliced to obtain a complete X-ray image after splicing.

[0010] As a further technical solution of the present invention: the result of the imaging processing is used to accurately identify the defect location and type and perform three-dimensional reconstruction.

[0011] As a further technical solution of the present invention: the method for dealing with defects includes defect size correction and defect processing method.

[0012] As a further technical solution of the present invention: the defect size correction is specifically: when measuring the defect size, combined with the actual detection requirements, according to the known device component size, the defect size is calibrated and calibrated using the vertically transmitted X-ray digital image and in the same direction as the object under the same depth of field condition. The calculation method for measuring the component size in the image is the formula: Where: Ni is the number of pixels obtained by measuring the size of the calibration structure by computer; Ns is the number of pixels obtained by measuring the defect size by computer; L is the actual size of the calibration structure; S is the size of the structure to be measured.

[0013] As a further technical solution of the present invention: the defect types are divided into material defects, assembly defects and foreign body defects.

[0014] As a further technical solution of the present invention: the defect handling method is power-off disassembly and maintenance.

[0015] Compared with the prior art, the present invention has the following beneficial effects: 1) The tooling and tooling linkage device (sensor integration device) suitable for the operating environment of GIS equipment are used; 2) Based on the experience in actual production, the defect types inside the GIS pipeline were divided, and a neural network model was built based on this experience; 3) In view of the actual situation of GIS operation, a cyclotron accelerator is used to replace the closed tube X-ray source to improve the recognition ability of the system; 4) A defect processing method based on three-dimensional CT internal defect imaging and recognition system is proposed. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 is a system schematic diagram of the present invention; Figure 2 is a physical diagram of the imaging system of the present invention; Figure 3 It is the imaging reconstruction flow chart of the present invention.

[0017] Figure 4 It is a schematic diagram of the image stitching process.

[0018] Figure 5 is a schematic diagram of a mechanical flexible electric device.

[0019] Figure 6It is the flowchart of the defect recognition part. Detailed implementation manners

[0020] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0021] As Figure 1-6 shown, the present invention discloses a three-dimensional CT internal defect imaging recognition system, which includes a front-end device and a control platform. The front-end device includes an X-ray machine and an imaging plate. Both the X-ray machine and the imaging plate are installed on a mechanical flexible electric device. The mechanical flexible device is used to adjust the position of the front-end device to ensure that the measured position and the front-end device are on the same horizontal plane. The control platform is used for imaging control and data output of the front-end device and the movement of the mechanical flexible device. The data obtained by the control platform is transmitted to the device health status image data platform for imaging processing.

[0022] The mechanical flexible electric device mainly consists of 5 parts: (1) Trolley base: responsible for controlling the movement of the tooling; (2) Support feet: used for strengthening stability when raising the device; (3) Electric lifting rod: the lifting height controls the detection height of the X-ray machine; (4) Top rotating platform: the horizontal rotation angle controls the detection direction of the X-ray machine; (5) X-ray machine platform: manually rotate the pitch angle of the X-ray machine.

[0023] Tooling dimensions: 1 m (length) x 0.6 m (width) x 1.62 m (height); The maximum detection height is approximately 4 m, the load capacity is 50 kg, the X-ray machine can be electrically lifted and rotated around the horizontal axis, the lifting accuracy is 10 mm, and the rotation accuracy is 1°.

[0024] The specific operation process is as Figure 1 shown: First, operator A needs to select the tooling of the radiation source according to the actual situation. After determining the tooling, the radiation source, GIS pipeline, and detector are positioned through the control program. The positioning process is realized by the sensors integrated on the tooling. After obtaining a certain number of detection results, operator B processes the collected data, judges whether there are defects inside through the recognition system, and reports according to the prompts built in the program if there are defects, reports the problem to the operation and maintenance department, and the operation and maintenance department repairs the GIS pipeline according to the actual situation.

[0025] The process by which the recognition system determines whether there are internal defects is as follows: By selecting X-ray images under typical defect conditions to construct an image library (typical defects mainly include bubbles / casting porosity, foreign objects / metal particles, etc., about 2,500 images in the atlas), an image recognition algorithm based on the Faster RCNN model is established to achieve automatic recognition, annotation, and positioning of defects in X-ray pictures of GIS equipment, and assist manual judgment. A picture repeatability recognition model is established to realize automatic repeatability analysis of the uploaded pictures with the pictures in the system database to determine whether there is detection fraud behavior and improve the supervision ability of the detection work.

[0026] The imaging result of the system can accurately identify the location and type of defects and perform three-dimensional reconstruction to make the result more intuitive.

[0027] Multiple X-ray images with intersections obtained by discrete shooting are automatically stitched by overlapping the intersection content to obtain a complete X-ray image, which is achieved through the following stitching steps: a) Overlapping imaging: Obtain multiple X-ray images with image intersections, number them in the order of adjacent relationships, and identify whether the main image parameters meet the stitching conditions; b) Feature extraction: Extract the main image into feature data composed of single pixels through gray value differences to form multiple feature points, and store it as a feature mapping image; c) Feature matching + image fusion: By reading the corresponding feature mapping images of the main image, perform feature data matching on the feature data within the image intersection range with the corresponding feature mapping images of any adjacent stitching images until all images are stitched to obtain the stitched complete X-ray image.

[0028] For the defect situations that appear in the pictures, the main countermeasures include defect size correction and defect handling methods: Defect size correction: When measuring the defect size, it should be combined with the actual detection requirements. According to the known dimensions of the equipment components, use the X-ray digital images of vertical penetration and calibrate and calibrate the defect size in the same direction as the object to be inspected under the same depth of field condition. The calculation method for measuring the component size in the image is shown in the formula:

[0029] In the formula: N i —— The number of pixels obtained by computer measurement and calibration of the structural size; N s —— The number of pixels obtained by computer measurement of the defect size; L —— The actual size (mm) of the calibration structure; S —— Dimensions of the structure to be measured (mm).

[0030] ② Defect handling method: Defects in high-voltage switchgear are mainly divided into material defects, assembly defects, and foreign object defects. Typical defect handling measures are shown in Table 1, and other defects not included can be referred to for implementation.

[0031] Table 1: Reference table for typical defect handling measures;

[0032] The present invention adopts tooling and tooling linkage devices (sensor integration devices) suitable for the operating environment of GIS equipment; divides the defect types inside the GIS pipeline based on experience in actual production, and builds a neural network model based on this experience; in view of the actual situation of GIS operation, replaces the closed-tube X-ray source with a cyclotron to improve the system's recognition ability; and proposes a defect handling method based on a three-dimensional CT internal defect imaging recognition system.

[0033] The X-ray energy for internal imaging in GIS is dynamically adjusted according to the material combination (steel / aluminum / copper), penetration thickness (usually 35 - 100 mm), and detection standards. The general X-ray energy range is 150 - 450 keV, and specific parameters need to be determined in combination with the equipment model and on-site tests. The peak energy of the cyclotron's rays can reach 7.5 MeV, the step size for ray energy adjustment is 0.5 MeV each time, the adjustment range is 2 MeV - 7.5 MeV, the maximum focal spot size is 0.3 × 3 mm, the overall device weight is 115 kg, and the ray dose rate at 1 m is 10 R / min.

[0034] For those skilled in the art, it is obvious that the present invention is not limited to the details of the above exemplary embodiments, and without departing from the spirit or basic characteristics of the present invention, the present invention can be implemented in other specific forms. Therefore, from any point of view, the embodiments should be regarded as exemplary and non-limiting. The scope of the present invention is defined by the appended claims rather than the above description. Therefore, all changes falling within the meaning and scope of the equivalent elements of the claims are intended to be included in the present invention. Any reference signs in the claims should not be regarded as limiting the claimed rights.

[0035] In addition, it should be understood that although this specification is described according to embodiments, not every embodiment only includes an independent technical solution. This narrative way of the specification is only for clarity. Those skilled in the art should regard the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A 3D CT internal defect imaging recognition system, comprising a front-end device and a control platform, characterized in that: The front-end device includes an X-ray machine and an imaging plate, both of which are installed on a mechanical flexible electric device. The mechanical flexible device is used to adjust the position of the front-end device to ensure that the measured position and the front-end device are on the same horizontal plane. The control platform is used for imaging control and data output of the front-end device and movement of the mechanical flexible device. The data obtained by the control platform is transmitted to the equipment health status image data platform for imaging processing.

2. A three-dimensional CT internal defect imaging and recognition system according to claim 1, characterized in that: The imaging process comprises the following steps: Step 1, overlapping imaging: obtain multiple X-ray images with image intersections, number them in order of adjacent relationship, and identify whether the main image parameters meet the stitching conditions; Step 2, feature extraction: extract the feature data containing multiple feature points formed by a single pixel from the main image through gray value difference, and store it as a feature mapping image; Step 3: Feature matching and image fusion: By reading the corresponding feature mapping image of the main image, the feature data within the image intersection range is matched with the corresponding feature mapping image of any adjacent stitched image until all images are stitched together to obtain a complete stitched X-ray image.

3. A 3D CT internal defect imaging and recognition system according to claim 2, characterized in that: The result of the imaging process is used to accurately identify the defect location and type and perform three-dimensional reconstruction.

4. A 3D CT internal defect imaging and recognition system according to claim 3, characterized in that: The methods for dealing with defects include defect size correction and defect treatment methods.

5. A three-dimensional CT internal defect imaging recognition system according to claim 4, characterized in that: The defect size correction is specifically: when measuring the defect size, combined with the actual inspection requirements, according to the known equipment component size, the defect size is calibrated and calibrated using the vertically transmitted X-ray digital image and in the same direction as the object under the same depth of field condition. The calculation method for measuring the component size in the image is the formula: Where: N i It is the number of pixels obtained by measuring the structural dimensions using a computer; N s It is the number of pixels obtained by measuring the defect size by computer; L It is the actual size of the structure used for calibration; S is the size of the structure to be measured.

6. A 3D CT internal defect imaging and recognition system according to claim 4, characterized in that: The types of defects are divided into material defects, assembly defects and foreign body defects.

7. The three-dimensional CT internal defect imaging recognition system according to claim 5, characterized in that: The defect handling method is power outage, disassembly and maintenance.

Images