Radiographic nondestructive testing method for armored temperature sensors

Abstract

The present invention relates to a radiographic nondestructive testing method for armored temperature sensors, comprising the following steps: S1, establishing a standard image database for armored temperature sensors; S2, obtaining image data to be tested of the armored temperature sensors to be tested; S3, comparing and analyzing the image data to be tested with the standard image database to obtain a test result. The present invention can be applied to different stage tests of armored temperature sensors, and the different stage tests include pre-manufacturing test, welding or post-assembly test, and finished product test, etc.; if applied to pre-manufacturing test, quality hazards or defective components can be eliminated, and the subsequent product rate can be improved; if applied to welding or post-assembly test, it can be used to evaluate the welding process and structural process effects; if applied to finished product test, the internal components of the armored temperature sensor can be fully tested to find possible quality weaknesses and potential quality hazards, thereby reducing the risks brought by equipment during the operation of nuclear power plants.

Description

Technical Field

[0001] The invention relates to the technical field of ray nondestructive testing, and in particular to a ray nondestructive testing method for an armored temperature sensor. Background Art

[0002] The armored temperature sensor is a temperature-sensitive device, which consists of one or more temperature sensing elements (usually in the form of wire winding) installed in the sheath, inner leads and external terminals for connecting electrical measuring instruments. It mainly includes temperature sensing elements, inner leads, signal leads, insulating materials, inner protective tubes, temperature sensing sleeves, etc. The temperature sensing elements and inner leads and signal leads are connected by welding. With the rapid development of my country's nuclear power projects, armored temperature sensors are widely used in engineering applications. Their quality and stable performance are related to the safe operation of nuclear power plants. Due to the harsh application environment, the presence of radiation, vibration, high temperature and other factors, the armored temperature sensors used in nuclear power are extremely difficult to replace after failure. Their design life is more than 20 years, so it is particularly important to check for hidden dangers and quality inspection before use.

[0003] Usually, a complete and qualified armored temperature sensor for nuclear power needs to go through more than 100 processing and quality inspection processes, and the quality control is relatively strict. However, the current technical solution still has potential quality weaknesses and hidden dangers in the internal components, which mainly include temperature sensing elements, inner leads, signal leads, insulating materials, etc. The detection rate of these weaknesses and hidden dangers is very low, or even cannot be detected. The main manifestations are: 1. Excessive stretching, overlapping or wrinkling of the wire windings between turns during the pressing process of the wire winding coil of the temperature sensing element; 2. False soldering / false connection between the temperature sensing element and the inner lead, and false soldering / false connection between the signal lead; 3. Excessive stretching of the inner lead during the filling and drawing process of the insulating material; 4. Defects in the body material of the temperature sensing element, etc. The above hidden dangers or defects will cause overcurrent disconnection and turn-to-turn short circuit problems in the armored temperature sensor during application. Summary of the invention

[0004] The technical problem to be solved by the present invention is to provide a convenient radiation non-destructive testing method for armored temperature sensors in view of at least one defect of the prior art.

[0005] The technical solution adopted by the present invention to solve the technical problem is: constructing a ray non-destructive testing method for an armored temperature sensor, comprising the following steps:

[0006] S1. Establish a standard image database of armored temperature sensors;

[0007] S2. Acquire the image data to be detected of the armored temperature sensor to be detected;

[0008] S3: Compare and analyze the image data to be detected with the standard image database to obtain a detection result.

[0009] Preferably, the armored temperature sensor includes multiple types of armored temperature sensors; and the step S1 includes:

[0010] S11. Acquire corresponding radiation energy value data according to parameters of the armored temperature sensors of various types;

[0011] S12, acquiring the corresponding standard image data of the armored temperature sensor according to the ray energy value data;

[0012] S13. Establish the standard image database according to the parameters of the various types of armored temperature sensors, the corresponding ray energy value data and the standard image data.

[0013] Preferably, the parameters of the multiple types of armor temperature sensors include the model of the armor temperature sensor, the detection position and the wall thickness parameter of each detection position.

[0014] Preferably, step S2 comprises:

[0015] S21, performing digital image acquisition on the armored temperature sensor to be detected through the ray detection system to obtain an acquired image;

[0016] S22: pre-process the collected image to obtain the image data to be detected.

[0017] Preferably, the step S21 includes:

[0018] S211, setting a corresponding ray energy value according to the model, detection position and wall thickness parameters of the armored temperature sensor to be detected;

[0019] S212, driving the armored temperature sensor to be detected to move, and moving each of the detection positions to the irradiation center of the radiation penetration field one by one, and then obtaining the image data to be detected of each of the detection positions.

[0020] Preferably, the detection position includes a first detection point set at the temperature sensing element of the armored temperature sensor, a second detection point set at the internal welding point of the armored temperature sensor, a third detection point set at the inner lead of the armored temperature sensor and / or a fourth detection point set at the signal lead and welding of the armored temperature sensor.

[0021] Preferably, step S3 comprises:

[0022] S31, according to the image data to be detected, retrieve the corresponding standard image data in the standard image database;

[0023] S32, comparing and analyzing the image data to be detected and its corresponding standard image data.

[0024] Preferably, the method further comprises:

[0025] Step S4: construct a computer neural network data set to train and test the standard image database.

[0026] Preferably, the step S4 specifically includes constructing a computer neural network data set based on a convolutional neural network algorithm, and normalizing the standard image data of various types of the armored temperature sensors to train and test the standard image database.

[0027] Preferably, the feature graph expression of the convolutional neural network is:

[0028]

[0029]

[0030] Where b is the deviation, Z l and Z l+1 They represent the convolution input and output of the l+1th layer respectively, K is the number of channels, s0 is the convolution step size, f is the convolution kernel size, and p is the number of padding layers.

[0031] The implementation of the present invention has the following beneficial effects: the present invention can be applied to different stage detections of armored temperature sensors, including pre-manufacturing detection, welding or post-assembly detection, and finished product detection; if applied to pre-manufacturing detection, quality hazards or defective components can be eliminated, thereby improving the subsequent product rate; if applied to welding or post-assembly detection, it can be used to evaluate the welding process and structural process effects; if applied to finished product detection, the internal components of the armored temperature sensor can be fully inspected to discover possible quality weaknesses and potential quality hazards, thereby reducing the risks brought by equipment during the operation of nuclear power plants. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] The present invention will be further described below with reference to the accompanying drawings and embodiments, in which:

[0033] Figure 1 It is a flow chart of the radiation nondestructive testing method of the armored temperature sensor of the present invention;

[0034] Figure 2 It is a flow chart of step S1 of the radiation nondestructive testing method of the armored temperature sensor of the present invention;

[0035] Figure 3 is a flow chart of step S2 of the radiation nondestructive testing method of the armored temperature sensor of the present invention;

[0036] Figure 4 is a flow chart of step S21 of the radiation nondestructive testing method of the armored temperature sensor of the present invention;

[0037] Figure 5 is a flow chart of step S3 of the radiation nondestructive testing method of the armored temperature sensor of the present invention;

[0038] Figure 6 It is a detection schematic diagram of the radiation nondestructive detection method of the armored temperature sensor of the present invention;

[0039] Figure 7 It is a flowchart of an embodiment of a radiation nondestructive testing method for an armored temperature sensor of the present invention;

[0040] Figure 8 This is an image of a qualified armored temperature sensor of the present invention;

[0041] Fig. 9 This is an image of the armored temperature sensor of the present invention that has the problem of indication fluctuation;

[0042] Fig.10 This is an image of the armored temperature sensor having a circuit breaking problem of the present invention;

[0043] Fig.11 This is an image of the armored temperature sensor of the present invention having the problem of overlapping turns of resistance wires;

[0044] Fig.12 This is an image of the armored temperature sensor of the present invention that has the problem of metal poisoning. DETAILED DESCRIPTION

[0045] In order to have a clearer understanding of the technical features, purposes and effects of the present invention, the specific embodiments of the present invention are now described in detail with reference to the accompanying drawings. In the following description, it should be understood that the directions or positional relationships indicated by "front", "back", "up", "down", "left", "right", "longitudinal", "horizontal", "vertical", "horizontal", "top", "bottom", "inside", "outside", "head", "tail", etc. are based on the directions or positional relationships shown in the accompanying drawings, are constructed and operated in a specific direction, and are only for the convenience of describing the present technical solution, rather than indicating that the device or element referred to must have a specific direction, and therefore cannot be understood as a limitation to the present invention.

[0046] It should also be noted that, unless otherwise clearly specified and limited, the terms such as "installed", "connected", "connected", "fixed", "set" and the like should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral one; it can be a mechanical connection or an electrical connection; it can be directly connected or indirectly connected through an intermediate medium, it can be the internal connection of two elements or the interaction relationship between two elements. When an element is referred to as being "on" or "under" another element, the element can be "directly" or "indirectly" located on the other element, or there may be one or more intermediate elements. The terms "first", "second", "third", etc. are only for the convenience of describing the present technical solution, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined as "first", "second", "third", etc. can explicitly or implicitly include one or more of the features. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to the specific circumstances.

[0047] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present invention. However, it should be clear to those skilled in the art that the present invention may be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present invention.

[0048] See also Figures 1 to 7 , is a radiation nondestructive testing method for an armored temperature sensor of the present invention. In this embodiment, X-rays are used as an example for nondestructive testing. It is understandable that other types of rays can also be used for testing. The radiation nondestructive testing method for an armored temperature sensor includes the following steps:

[0049] Step S1, establishing a standard image database of armored temperature sensors according to standard image data; specifically, establishing a standard image database of armored temperature sensors according to parameters of various armored temperature sensors to be detected; the armored temperature sensors include various types of armored temperature sensors; such as Figure 2 As shown, further, step S1 includes:

[0050] Step S11, according to the parameters of the various types of armor temperature sensors, the corresponding radiation energy value data are obtained; wherein the parameters of the various types of armor temperature sensors include the model of the armor temperature sensor, the detection position and the wall thickness parameter of each detection position, and the radiation energy value data of the various types of armor temperature sensors are obtained according to the above parameters. wherein the detection position of each type of armor temperature sensor can be a preset detection position.

[0051] Step S12: acquiring standard image data of the corresponding armored temperature sensor according to the ray energy value data;

[0052] Step S13: establishing a standard image database according to the parameters of various types of armored temperature sensors, their corresponding ray energy value data and standard image data.

[0053] Since armored platinum thermal resistor temperature sensors are made of composite processing of different metal structures, the energy of X-ray photons needs to be adjusted according to the thickness and density of different metal materials. The thickness and density of each metal material corresponds to a ray energy level, otherwise the transmission effect cannot be achieved; that is, each armored temperature sensor corresponds to a ray energy value data. Therefore, the standard image database includes the parameters of all armored temperature sensors that need to be tested. The parameters of each armored temperature sensor include the model of the armored platinum thermal resistor temperature sensor, the wall thickness parameters of each part, the ideal standard image, etc., and then the ray energy value data of the armored temperature sensor is obtained. Finally, the ray energy value data of various types of armored temperature sensors are summarized to establish a standard image database.

[0054] Step S2, obtaining the image data to be detected of the armored temperature sensor to be detected; Figure 3 As shown, further, step S2 includes:

[0055] Step S21, the armored temperature sensor to be detected is subjected to digital image acquisition through the ray detection system to obtain an acquired image; specifically, the armored temperature sensor to be detected can be subjected to image acquisition through the X-ray detection system to obtain an acquired image; the X-ray detection system is also called the X-ray constant frequency emission and imaging system; before the detection, the system is firstly set with relevant parameters, including pre-setting the detection position of the temperature sensor, and adjusting the ray energy value and relevant parameters corresponding to the X-ray detection system according to the pre-set detection position of the temperature sensor;

[0056] like Figure 4 As shown, specifically, step S21 includes:

[0057] Step S211, according to the model, detection position and wall thickness parameters of the armor temperature sensor to be detected, the corresponding ray energy value is set; more specifically, according to the model of the armor temperature sensor to be detected and the detection position corresponding to the armor temperature sensor of this model, the image is collected, that is, the X-ray detection system will adjust the corresponding parameters of the system according to the model of the armor temperature sensor to be detected and the detection position corresponding to the armor temperature sensor of this model, and the X-ray detection system can quickly and accurately collect the detection position of the armor temperature sensor to be detected and obtain the corresponding collection image;

[0058] Step S212, drive the armored temperature sensor to be detected to move, and move each detection position one by one to the irradiation center of the radiation transmission field, and then obtain the image data to be detected at each detection position; further, according to the recognition requirements of the radiation detection system, a sliding inspection device can be set in the radiation detection system, and a suitable sliding track is set according to the preset detection position of the armored temperature sensor. The armored temperature sensor to be detected is transported to the identifiable range of the radiation detection system through the sliding inspection device, and the detection range is within the square detection window of the radiation detection system operating platform; preferably, each preset detection position of the armored temperature sensor is located at the center below the radiation transmission field.

[0059] Step S22, pre-processing the collected image to obtain the image data to be detected. Specifically, step S22 mainly processes the collected image obtained in step S1 through image processing methods such as image filtering and image noise reduction to obtain the image data to be detected.

[0060] Furthermore, in this embodiment, there are four preset detection positions, the diameter of which is 8.4mm, the initial X-ray tube voltage is 130-150KV, the voltage varies with the diameter of the standard part, based on 130KV, the voltage increases by 10KV for every 1mm increase in diameter; the exposure time is set to 333ms, the resolution is 2000*1500, the radiation angle is vertically 90°, and the focal length is automatically adjusted. Figure 6 As shown, specifically, the X-ray detection system includes an X-ray machine 100, and an X-ray beam is emitted by a ray source such as the X-ray machine 100. The ray is attenuated after passing through a relay, and the transmitted ray is converted into an analog signal and a digital signal by a ray receiving and conversion device, and then the detection result image is visually presented on the display screen with the help of the transmission technology of semiconductor materials and the processing and information technology of digital images. Under the premise of not affecting or damaging the performance of the temperature sensor, the images of the internal structures of the various parts of the object on the film 101 can be distinguished due to the different photosensitivity, and the internal defects can be intuitively displayed. In the detection process of this embodiment, four detection points are selected as preset detection positions for detection. The preset detection positions of the armored temperature sensor include a first detection point 1 set at the temperature sensing element of the armored temperature sensor, a second detection point 2 set at the internal welding point of the armored temperature sensor, a third detection point 3 set at the inner lead of the armored temperature sensor, and a fourth detection point 4 set at the signal lead welding of the armored temperature sensor.

[0061] The first test point 1 mainly tests the temperature sensing components, which can identify the following quality and hidden dangers:

[0062] (1) Overlapping shadows and wrinkles of platinum wire coils, and uneven distance between turns of platinum wire coils, such as Fig.11 This problem causes the temperature sensor to have poor measurement accuracy and fluctuate in the measured value;

[0063] (2) The lead of the platinum wire ring is broken and open. This problem causes the temperature sensor to be unable to measure normally.

[0064] The second inspection point 2, the third inspection point 3 and the fourth inspection point 4 mainly inspect the internal leads and their welding points, and can identify the following quality and hidden dangers:

[0065] (1) The lead wire and its welding point are poorly soldered, which will cause the temperature sensor measurement value to fluctuate and the insulation to be low;

[0066] (2) Whiskers and impurities in the leads and their welding points will cause metal poisoning of the temperature sensor, such as Fig.12 As shown, this can lead to fluctuations in measured values, low insulation, open circuits, etc.

[0067] Step S3: Compare and analyze the image data to be detected with the standard image database to obtain the detection result. Figure 5 As shown, specifically, step S3 includes:

[0068] Step S31, according to the image data to be detected, retrieve the corresponding standard image data in the standard image database; for example, according to the image data to be detected, obtain the model of the temperature sensor to be detected, and retrieve the standard image data corresponding to the model of the temperature sensor; if the specific model of the temperature sensor to be detected cannot be accurately obtained, retrieve the standard image data of the temperature sensor of a similar model;

[0069] Step S32, compare and analyze the image data to be detected with the standard image data corresponding thereto; specifically, when the captured image of the armored temperature sensor to be detected is obtained, or after the image data to be detected is obtained through image processing, the standard image data corresponding to the armored temperature sensor to be detected is retrieved from the standard image database, and the current image data to be detected is compared and analyzed with the standard image data to obtain the detection result. Further, the standard image data corresponding to the armored temperature sensor to be detected can be the standard image data corresponding to the model of the armored temperature sensor to be detected, or the standard image data corresponding to the shape of the armored temperature sensor to be detected, or the standard image data corresponding to the relevant parameters of the armored temperature sensor to be detected; the standard image data can include one or more groups, preferably a group of standard image data corresponding to the model of the armored temperature sensor to be detected is retrieved for comparison and analysis; but when the model of the armored temperature sensor to be detected cannot be obtained, it can be considered to retrieve one or more groups of standard image data similar to the shape or relevant parameters of the armored temperature sensor to be detected for comparison and analysis.

[0070] Among them, the standard image database includes image data of different measurement positions of various types of armored temperature sensors. Therefore, when performing comparative analysis, the X-ray detection system will compare the image data of different measurement positions of the armored temperature sensor to be detected with the corresponding standard image data. For example, assuming that the model of the armored temperature sensor to be detected is A, and the X-ray detection system obtains the image data of the first detection point of the armored temperature sensor to be detected, then the system will quickly filter out the standard image data corresponding to the temperature sensor of model A according to the model of the armored temperature sensor for comparison. More specifically, the X-ray detection system will directly compare and analyze the image data of the first detection point of the temperature sensor of model A stored in the standard image database in advance, so as to achieve efficient and rapid detection and reduce the workload of system comparative analysis. Figure 7 As shown, in one embodiment, the X-ray detection system includes an X-ray generator, and the detection position of the armored platinum thermal resistor temperature sensor to be detected is imaged by the X-ray generator, and the imaging is collected and processed, and the sensor quality is screened through a visual comparison system.

[0071] In some embodiments, the radiation non-destructive testing method of the armored temperature sensor also includes step S4, constructing a computer neural network data set to train and test a standard image database.

[0072] Specifically, step S4 includes building based on the convolutional neural network algorithm, and the convolution kernel parameter sharing and the sparsity of the inter-layer connection in the hidden layer of the convolutional neural network can automatically learn the deep features of the input data. First, the normalization processing of the standard images of various armored platinum thermal resistor temperature sensors is completed, two channels (black and white) are set, and the standardized image is set to obtain the standard image data. Among them, the feature map expression of the convolutional neural network is:

[0073]

[0074]

[0075] Where b is the deviation, Z l and Z l+1 They represent the convolution input and output of the l+1th layer respectively, K is the number of channels, s0, f and p are the convolution layer parameters, s0 is the convolution step, f is the convolution kernel size, and p is the number of padding layers. Specifically, the number of K channels is set to 3, the number of p padding layers is 2, the f convolution kernel size is set to 5*5, and the s0 convolution step is set to 4.

[0076] The pooling layer uses the Lp pooling model:

[0077]

[0078] s0 and pixel (i, j) are set the same as the above convolutional layer parameters, and p is pre-specified as 1.

[0079] According to the above-mentioned X-ray nondestructive testing method of armored temperature sensor, through comparative analysis, the following situations may be identified:

[0080] 1) Overlapping shadows and wrinkles of platinum wire loops;

[0081] 2) The distance between the turns of the platinum wire is uneven, such as Fig.11 As shown;

[0082] 3) Armored platinum thermal resistor with open circuit at platinum wire coil, signal lead or solder joint.

[0083] After testing the platinum wire resistance coils of different armored temperature sensors, the test results are as follows:

[0084] (1) Figure 8 As shown, the internal platinum wire loop of a qualified temperature sensor is clear and intact without overlap or wrinkles.

[0085] (2) Fig. 9 As shown in the figure, the platinum wire coil inside the temperature sensor probe with fluctuating indication has overlapping shadows and wrinkles. Under slight on-site vibration, the platinum wire coil may have a short circuit between turns, which will cause a short-term drop in the probe resistance and thus cause the temperature display to fluctuate for a short time. The reason is the manufacturing quality defects of the sensor components or excessive stretching during the manufacturing process of the sensor body.

[0086] (3) Fig.10 As shown, the platinum wire loop of the broken temperature sensor also has overlapping shadows and wrinkles, and is obviously stretched, which can easily cause excessive current density (electric stress concentration) at the damaged part, causing the resistance wire to burn and fuse due to overcurrent at that part.

[0087] The present invention solves the technical problem that the potential quality weaknesses and quality risks of some parts of armored temperature sensors are not easy to detect in the existing process and quality control processing and manufacturing. Nuclear-grade temperature sensors have harsh working environment conditions and high quality requirements. During the operation of nuclear power plants, due to the influence of harsh operating conditions such as high temperature / irradiation / vibration, temperature sensors with quality weaknesses and potential quality risks will develop into quality defects, and the temperature sensor signal output fluctuations / circuit breakage / open circuit / insulation resistance value lower than the acceptance value will occur, resulting in internal operation events (IOE) of nuclear power plants. The present invention intends to add X-ray non-destructive testing and visual identification processes to the key process quality control or finished parts of armored temperature sensors, detect the internal structure and quality risks of the armored body without destroying the metal structure, and control the product quality risks before the temperature sensor leaves the factory. The present invention can be applied to different stage detections of armored temperature sensors, including pre-manufacturing detection, welding or post-assembly detection, and finished product detection. If applied to pre-manufacturing detection, quality hazards or defective components can be eliminated, thereby improving the subsequent product rate. If applied to welding or post-assembly detection, it can be used to evaluate welding processes and structural process effects. If applied to finished product detection, the internal components of the armored temperature sensor can be completely detected to discover possible quality weaknesses and potential quality hazards, thereby reducing the risks brought by equipment during the operation of nuclear power plants.

[0088] It can be understood that the above embodiments only express the preferred implementation modes of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the patent scope of the present invention. It should be pointed out that, for ordinary technicians in this field, the above technical features can be freely combined without departing from the concept of the present invention, and several deformations and improvements can be made, which all belong to the protection scope of the present invention. Therefore, all equivalent changes and modifications made to the scope of the claims of the present invention should belong to the coverage of the claims of the present invention.

Claims

1. A radiation nondestructive testing method for an armored temperature sensor, characterized in that: The following steps are involved: S1. Establish a standard image database of armored temperature sensors; S2. Acquire the image data to be detected of the armored temperature sensor to be detected; S3, comparing and analyzing the image data to be detected with the standard image database to obtain a detection result; The step S2 comprises: S21, performing digital image acquisition on the armored temperature sensor to be detected through a ray detection system to obtain an acquired image; S22, preprocessing the collected image to obtain the image data to be detected; The armor temperature sensor includes multiple types of armor temperature sensors, and the parameters of the multiple types of armor temperature sensors include the model of the armor temperature sensor, the detection position and the wall thickness parameter of each detection position; the step S1 includes: S11. Acquire a type of ray energy value data corresponding to each type of armored temperature sensor according to the parameters thereof; S12, acquiring the corresponding standard image data of the armored temperature sensor according to the ray energy value data; S13, establishing the standard image database according to the parameters of the various types of armored temperature sensors, the corresponding ray energy value data and the standard image data; Wherein, the step S3 comprises: S31, according to the image data to be detected, retrieve the corresponding standard image data in the standard image database; S32, comparing and analyzing the image data to be detected and its corresponding standard image data.

2. The radiation nondestructive testing method of the armored temperature sensor according to claim 1 is characterized in that: The step S21 comprises: S211, setting a corresponding ray energy value according to the model, detection position and wall thickness parameters of the armored temperature sensor to be detected; S212, driving the armored temperature sensor to be detected to move, and moving each of the detection positions to the irradiation center of the radiation penetration field one by one, and then obtaining the image data to be detected of each of the detection positions.

3. The radiation nondestructive testing method of the armored temperature sensor according to claim 2 is characterized in that: The detection positions include a first detection point set at the temperature sensing element of the armored temperature sensor, a second detection point set at the internal welding point of the armored temperature sensor, a third detection point set at the inner lead of the armored temperature sensor and / or a fourth detection point set at the signal lead and welding of the armored temperature sensor.

4. The radiation nondestructive testing method of the armored temperature sensor according to claim 1 is characterized in that: The method further comprises: Step S4: construct a computer neural network data set to train and test the standard image database.

5. The radiation nondestructive testing method of the armored temperature sensor according to claim 4 is characterized in that: The step S4 specifically includes constructing a computer neural network data set based on a convolutional neural network algorithm, normalizing the standard image data of various types of armored temperature sensors to train and test the standard image database.

6. The radiation nondestructive testing method of the armored temperature sensor according to claim 5 is characterized in that: The feature graph expression of the convolutional neural network is: = , ; Where b is the deviation, and They represent the convolution input and output of the l+1th layer respectively, K is the number of channels, is the convolution step size, f is the convolution kernel size, and p is the number of padding layers.

Images