JPH03128447A - Insulation inspecting method for power cable and connection part thereof - Google Patents

Abstract

PURPOSE:To improve reliability of a power cable by irradiating the power cable by X-ray, reading out an image pickup by a reading-out device, expanding a picture density, and automatically discriminating harmful defects in an insulator by a neuro-computer. CONSTITUTION:The image pickup of power cable 2 generated on a picture medium 5 consisting of an imaging plate, etc., by the transmitted amount of X-ray which is irradiated by an X-ray device 1, is read out by a reading-out device 6, and the picture density is converted to an electrical signal and sent to a picture processing device 7. In the device 7, the picture density is strongly expanded further by the internal process to perform a picture improving process such as a frequency process, etc. Next, a density signal of the processed picture is sent to the neuro-computer 8, whereby the harmful defects in the insulator4 of cable 2 such as a void and a foreign matter, etc. are automatically discriminated. The reliability of cable 2 can be thus improved.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a method for inspecting insulators of power cables and their connections.

[Conventional technology]

In recent years, C* cables, CD cables, etc. have been used as typical power cables, and polyethylene, which has an X-ray absorption rate close to that of air, is used as the insulator of these power cables.

If the insulator of a power cable using such a polyethylene insulator contains defective voids (air bubbles) or harmful foreign matter, a serious accident may occur. Therefore, the conventional methods for detecting these defects have been to visually confirm the image projected onto a film or imaging plate using X-ray fluoroscopic imaging, or to use specialized equipment such as a Neumann-type computer or image processing. The images were visually sorted by one person while improving the image quality.

[Problem to be solved by the invention]

Even with the above-mentioned conventional X-ray fluoroscopic imaging method, harmful foreign substances such as minute metals in the capacitor insulator can be easily detected.

However, in the case of the above-mentioned voids (air bubbles) formed in the insulator, there is no large difference in the transmitted X-ray dose between the χ rays that pass through the voided part and the X*I that passes through the healthy part without voids. This is because the image signal of the void portion is buried in the image noise signal due to scattered radiation generated from the insulator, granularity of the film, quantum mottle, structural mottle, etc., and is difficult to identify.

Incidentally, if a void with a diameter of 0.5M occurs in the insulator thickness of 150 holes, the amount of XLS transmitted through the void portion is equal to the amount of XLS transmitted through the healthy portion.
The amount increases by approximately 1/100 of the dose.

Therefore, an edge enhancement method that emphasizes the outer periphery of the image of the void region to be detected is used to highlight the void region and clearly divide the image signal of the void region and the healthy region into two.
There is also a method of automatically identifying voids as defects by converting them into values, but since the voids are spherical at -i, there is no part corresponding to an edge, and this method cannot be applied.

However, since humans have the ability to pattern and identify film images, they can identify clusters of voids from patterns without binarizing them, so humans, not machines, can identify void patterns. After doing 11 memorization exercises, I had no choice but to identify them.

However, testing by humans increases testing costs and
There was a problem in that defect detection accuracy decreased due to fatigue. In addition, with human visual inspection methods, it is impossible to conduct high-speed inspections over long periods of time during cable manufacturing, so even if we wanted to improve reliability through full-length inspections, it was practically impossible. There were also problems.

This invention was made in view of the current state of the conventional electric cable inspection method as described above, and its purpose is to read the image recorded on an image medium by an X-ray device with an image reading device, and to obtain the TI image. To provide a method for inspecting insulators of electrical cables and connections thereof, in which signals are processed by an image processing device and harmful defects are automatically identified by a neurocomputer trained to identify patterns of harmful defects.

[Means to solve the problem]

Therefore, in the present invention, as a means for solving the above problems, an xLA device irradiates the power cable or its connection part with X-rays, the image recorded on the image medium is read by the image reading device, and the image signal is The company adopted a power cable insulator inspection method in which the image is processed by an image processing device to enlarge the shading of the image, and the shading signals of this image are pattern-recognized by a neurocomputer to automatically identify harmful defects in the insulator.

[Effect]

In the above-mentioned method according to the present invention, the electric cable or its connection part is irradiated with X-rays, the image is read by a reading device, and then the shading of the image is expanded by image processing, and the shading signal of this image can be identified as a pattern. A neurocomputer trained to automatically identify harmful defects in insulators.

〔Example〕

Embodiments of the present invention will be described below with reference to the accompanying drawings.

FIG. 1 shows an overall schematic block diagram of an apparatus for carrying out the method for inspecting the insulation of a power cable and its connection part according to the present invention, 1 is an X-ray device, 2 is an insulator extruded onto a core conductor 3. The electric cable 5 consists of a body 4, and 5 is an image medium that receives the transmitted radiation of the electric cable 1 irradiated with X-rays from an X-ray device.

The insulator 4 is made of polyethylene, which is commonly used and has an X-ray absorption rate close to that of air.The image medium 5 is an imaging plate or a high-sensitivity photographic film. Note that it is also possible to use an X-ray sensor as this medium.

The image generated on the image medium 5 by the amount of transmitted X-rays irradiated by the X-ray device 1 is read by the reading device 6, and the shading of the image is converted into an electrical signal and sent to the image processing device 7. The image processing device 7 further enhances and enlarges the shading of the image through its internal processing, and performs image improvement processing such as frequency processing.

The gray level signals of the image processed as described above are sent to the neurocomputer 8. This neurocomputer 8
are designed to behave in a manner similar to the behavior of human brain cells. For this reason, as human brain cells learn the same thing over and over again, the connections between specific brain neurons become stronger, and the neural network itself, which is a collection of connections, changes flexibly as a result of learning. By learning, you can multiply what you have learned and turn it into knowledge, and by learning the neurocomputer, you can turn what you have learned into knowledge.

Therefore, initially it is not possible to judge the gray level signal of the above image, but
Since the neurocomputer 8 is configured as described above, a human being can repeatedly teach the computer 8 patterns of images of voids and foreign objects to the computer 8 (and the computer itself can learn images of voids and foreign objects). memorize the pattern,
It will automatically identify harmful defects in the insulation, such as voids and foreign objects, just as the human brain does.

FIG. 2 shows the state of light transmission when voids 9 exist in the insulator 4 of the power cable 2.

FIG. 3 shows a schematic pattern of imaging the image medium 5 as seen from the line ■-■ in FIG. 9' is a void imaging point where the amount of X-ray transmission increases corresponding to the void 9 and is shaded. In the figure, A and C are regions corresponding to the insulator 4, and as the insulator 4 becomes thicker, it becomes whitish in proportion. Area B corresponding to the core conductor 3 appears white.

FIG. 4 is a partially enlarged view of the vicinity of the void imaging point S' in FIG. FIG. 5 shows the concentration distribution along the line X--X in FIG. 4, in which the same degree of flaking as in the void portion occurs in the Y and Y' regions due to image noise.

Therefore, the void part between a and b and Y, Y/j
It is difficult to distinguish between the 1 region and the 1 region based on the concentration distribution alone.

〔effect〕

As explained in detail above, in this invention, an electric cable or its connection part is irradiated with X-rays, the captured image is read, the shading of the image is enlarged, and this image signal is used by a neurocomputer to learn and perform pattern recognition. Since harmful defects in insulators can be automatically identified, defects such as voids can be automatically detected at high speed, and if used in a detection device for power cables being manufactured, it can be detected continuously at high speed over the entire length of power cables. defects can be detected. Furthermore, using this method, defects can be detected in a short time and with high precision even in connection parts that are assembled on-site, thereby improving the reliability of electrical cables.
The reliability of the power cable system can be improved.

[Brief explanation of the drawing]

Fig. 1 is an overall schematic block diagram of an apparatus for carrying out the method for inspecting the insulator of a power cable and its connection portion according to the present invention; Fig. 2 is a diagram showing the state of X-ray transmission of voids in the power cable insulator; The figure shows the imaging pattern on the image medium as seen from the line FIG. 3 is a diagram showing concentration distribution on X-rays. 1... X-ray device, 3... Core conductor, 5... Image medium, 7... Image processing device, 8... Neuro Computer, S...
Void, 9'...Void imaging point. 2... Electric cable, 4... Insulator, 6... Reading device,

Claims (1)

[Claims] (1) A power cable or its connection part is irradiated with X-rays by an X-ray device, the image captured on the image medium is read by an image reading device, and the image signal is processed by an image processing device to enlarge the shading of the image. A method for inspecting insulators of power cables and their connections, characterized by pattern recognition of the grayscale signals of this image using a neurocomputer to automatically identify harmful defects in the insulators.