A quality detection method for temperature element

By building a multi-dimensional working environment parameter space and introducing impact testing scenarios, combined with multi-source data fusion and correction mechanism, the limitations under a single environmental condition in temperature component quality detection are solved, and more accurate and reliable detection results are achieved. Temperature component performance evaluation in complex working environments.

CN118794568BActive Publication Date: 2024-11-26ZIBO YIYUAN ENG & TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411287689.0
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-09-14
Publication Date
2024-11-26
Estimated Expiration
2044-09-14

AI Technical Summary

Technical Problem

The existing temperature component quality detection technology is limited to performance testing under a single environmental condition, and cannot comprehensively and accurately test the performance of temperature components in actual complex working environments, resulting in insufficient accuracy and reliability of quality detection results.

Method used

By building a multi-dimensional working environment parameter space, the impact testing scenario is introduced, and the multi-source data fusion and correction mechanism is adopted, including randomly generating test scenarios, conducting continuous and impact testing, analyzing test results, performing data correction and weighting calculations, and obtaining comprehensive testing. Accuracy information.

Benefits of technology

It effectively improves the accuracy and reliability of the quality detection results of temperature components, and can more truly reflect the performance of temperature components in complex working environments.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118794568B_ABST
    Figure CN118794568B_ABST
Patent Text Reader

Abstract

The present invention discloses a quality detection method for temperature elements, which relates to the field of data processing technology. The method comprises: obtaining a working environment parameter space within a working environment of the temperature element; randomly generating a plurality of test working environment parameters, screening and obtaining a plurality of impact test scenes; obtaining a plurality of impact temperature test result arrays and a plurality of impact image test result array sequences; obtaining a plurality of continuous detection accuracy information; obtaining a plurality of impact detection accuracy information; performing weighted calculation on a plurality of continuous detection accuracy information and a plurality of impact detection accuracy information to obtain detection accuracy information. The method solves the technical problem that the existing quality detection of temperature elements is limited to performance testing under a single environmental condition, and it is difficult to comprehensively and accurately test the performance of temperature elements in an actual complex working environment, which leads to insufficient accuracy and reliability of quality detection results, and achieves the technical effect of effectively improving the accuracy and reliability of quality detection results.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the technical field related to data processing, and in particular to a quality detection method for a temperature element. Background Art

[0002] In today's industrial manufacturing and automation control fields, temperature components, as core sensing components, are widely used in aerospace, petrochemical, electric power, medical equipment and other fields. Their stability and reliability are directly related to the overall performance and safety of the system. However, in actual working environments, temperature components often face complex and changeable temperature and pressure conditions. These extreme conditions not only test the tolerance of the components, but also put forward strict requirements on their detection accuracy. Traditional temperature component quality inspections often focus on performance testing under static conditions, ignoring the dynamic changes and extreme shocks in the actual working environment. It is impossible to simulate the various working conditions and sudden extreme environmental changes that temperature components may encounter in actual applications. It is difficult to fully and truly reflect the actual performance of temperature components in actual complex working environments, resulting in some potential problems being exposed in subsequent use, increasing maintenance costs and safety hazards.

[0003] Therefore, in the current technology related to the quality inspection of temperature components, there is a technical problem that the performance test is limited to a single environmental condition, and it is difficult to comprehensively and accurately test the performance of the temperature components in an actual complex working environment, which in turn leads to the technical problem of insufficient accuracy and reliability of the quality inspection results. Summary of the invention

[0004] The present application provides a quality inspection method for temperature components. By adopting technical means such as constructing a multi-dimensional working environment parameter space, introducing an impact test scenario, and adopting a multi-source data fusion and correction mechanism, the present application solves the technical problem that the existing quality inspection of temperature components is limited to performance testing under a single environmental condition, and it is difficult to comprehensively and accurately test the performance of the temperature components in an actual complex working environment, which in turn leads to insufficient accuracy and reliability of the quality inspection results. The present application achieves the technical effect of effectively improving the accuracy and reliability of the quality inspection results.

[0005] The present application provides a quality detection method for temperature elements, the method comprising: obtaining a working environment parameter space within a working environment of the temperature element, wherein the working environment parameter space comprises an environment temperature space and an environment pressure space; randomly generating a plurality of test working environment parameters within the working environment parameter space as a plurality of continuous test scenarios, traversing and combining the plurality of test working environment parameters, and screening and obtaining a plurality of impact test scenarios based on the working environment parameter records; performing continuous testing on a multi-point temperature element array under the plurality of continuous test scenarios to obtain a plurality of continuous temperature test result arrays and a plurality of continuous image test result arrays, and performing impact testing on the multi-point temperature element array under the plurality of impact test scenarios to obtain a plurality of impact temperature test result arrays and a plurality of impact image test result array sequences; analyzing and obtaining a plurality of basic continuous detection accuracy information based on the plurality of continuous temperature test result arrays, According to the multiple continuous image test result arrays, multiple continuous damage trend information is obtained by analysis, multiple continuous detection influence coefficients are obtained by classification, and the multiple basic continuous detection accuracy information is corrected to obtain multiple continuous detection accuracy information; according to the multiple impact temperature test result arrays, multiple basic impact detection accuracy information is obtained by analysis, and according to the multiple impact image test result array sequences, impact dynamic damage analysis is performed to obtain multiple impact damage trend information, multiple impact detection influence coefficients are obtained by classification, and the multiple basic impact detection accuracy information is corrected to obtain multiple impact detection accuracy information; multiple extreme degrees and multiple continuous occurrence degrees of the multiple continuous test scenarios are obtained, and multiple impact degrees and multiple impact occurrence degrees of the multiple impact test scenarios are obtained, and weighted calculations are performed on the multiple continuous detection accuracy information and the multiple impact detection accuracy information respectively to obtain detection accuracy information as a quality inspection result.

[0006] In a possible implementation, a plurality of test working environment parameters are randomly generated in the working environment parameter space as a plurality of continuous test scenarios, and the plurality of test working environment parameters are traversed and combined, and a plurality of impact test scenarios are screened and obtained according to the working condition parameter records, and the following processing is performed: a plurality of test working environment parameters are randomly generated in the working environment parameter space as a plurality of continuous test scenarios; the plurality of test working environment parameters are traversed and combined to obtain a plurality of test working environment parameter combinations; the working environment parameter records are obtained in a historical time, and the plurality of test working environment parameter combinations are used to perform impact scenario retrieval in the working environment parameter records, and the plurality of test working environment parameter combinations that have appeared are constructed into a plurality of impact test scenarios.

[0007] In a possible implementation, based on the multiple continuous temperature test result arrays, multiple basic continuous detection accuracy information is analyzed and obtained, based on the multiple continuous image test result arrays, multiple continuous damage trend information is analyzed and obtained, and multiple continuous detection influence coefficients are classified and obtained, and the multiple basic continuous detection accuracy information is corrected, and the following processing is performed: based on the multiple continuous temperature test result arrays, multiple average test temperatures are calculated and obtained; based on the multiple average test temperatures and the ambient temperatures of the multiple continuous test scenes, multiple basic continuous detection accuracy information is calculated and obtained; temperature component damage is identified on the multiple continuous image test result arrays to obtain multiple continuous damage trend information; the multiple continuous damage trend information is input into a continuous damage detection classifier, and classified to obtain multiple continuous detection influence coefficients, wherein the continuous damage detection classifier is constructed based on the mapping relationship between a sample continuous damage trend information set and a sample continuous detection influence coefficient set; the multiple basic continuous detection accuracy information is corrected and calculated using the multiple continuous detection influence coefficients to obtain multiple continuous detection accuracy information.

[0008] In a possible implementation, temperature element damage identification is performed on the multiple continuous image test result arrays to obtain multiple continuous damage trend information, and the following processing is also performed: based on the damage identification data of the temperature element, a sample continuous image test result set is collected, and a sample continuous damage trend information set is collected; using the sample continuous image test result set and the sample continuous damage trend information set, a continuous damage identifier is trained based on a convolutional neural network; using the continuous damage identifier, the continuous image test results in the multiple continuous image test result arrays are identified to obtain multiple element continuous damage trend information arrays; based on the multiple element continuous damage trend information arrays, multiple continuous damage trend information are calculated.

[0009] In a possible implementation, based on the multiple shock temperature test result arrays, multiple basic shock detection accuracy information is analyzed and obtained; based on the multiple shock image test result array sequences, shock dynamic damage analysis is performed to obtain multiple shock damage trend information; multiple shock detection influence coefficients are classified and obtained; the multiple basic shock detection accuracy information is corrected; and the following processing is performed: based on the multiple shock temperature test result arrays, multiple pre-shock average test temperatures and multiple post-shock average test temperatures are calculated and obtained; based on the multiple pre-shock average test temperatures and multiple post-shock average test temperatures, combined with the temperature combinations of the multiple shock test scenarios, multiple pre-shock average test temperatures and multiple post-shock average test temperatures are calculated and obtained. The invention discloses a method for detecting the impact of a plurality of impact images, and obtains a plurality of impact detection accuracy information, and calculates the average value to obtain a plurality of basic impact detection accuracy information; performs a dynamic damage analysis of the impact of a temperature element according to the array sequence of the plurality of impact image test results, and obtains a plurality of impact damage trend information; inputs the plurality of impact damage trend information into an impact loss detection classifier, and classifies them to obtain a plurality of impact detection influence coefficients, wherein the impact loss detection classifier is constructed based on the mapping relationship between a set of sample impact damage trend information and a set of sample impact detection influence coefficients; uses the plurality of impact detection influence coefficients to perform correction calculations on the plurality of basic impact detection accuracy information, and obtains a plurality of impact detection accuracy information.

[0010] In a possible implementation, a temperature element impact dynamic damage analysis is performed based on the multiple impact image test result array sequences to obtain multiple impact damage trend information, and the following processing is also performed: image extraction and downsampling processing are performed on the multiple impact image test result array sequences according to the first step length to obtain multiple first impact image test result array sets; image extraction is performed on the multiple impact image test result array sequences according to the second step length to obtain multiple second impact image test result array sets, and the second step length is greater than the first step length; based on a slowfast network, an impact dynamic damage analyzer is constructed and trained to obtain, wherein the impact dynamic damage analyzer is trained to convergence based on multiple sample first impact image test result array sets, multiple sample second impact image test result array sets and multiple sample element impact damage trend information array sets; using the impact dynamic damage analyzer, the multiple first impact image test result array sets and the multiple second impact image test result array sets are input recognized to obtain multiple element impact damage trend information arrays; based on the multiple element impact damage trend information arrays, multiple impact damage trend information are calculated to obtain.

[0011] In a possible implementation, multiple extreme degrees and multiple continuous occurrence degrees of the multiple continuous test scenarios are obtained, and multiple impact degrees and multiple impact occurrence degrees of the multiple impact test scenarios are obtained, and weighted calculations are performed on the multiple continuous detection accuracy information and multiple impact detection accuracy information respectively, and the following processing is also performed: the deviation amplitudes between the multiple continuous test scenarios and the conventional environmental parameters are calculated as multiple extreme degrees; the working environment parameter records within the historical time are retrieved to obtain the occurrence ratios of the multiple continuous test scenarios as multiple continuous occurrence degrees; according to the sizes of the multiple extreme degrees and the multiple continuous occurrence degrees, double weight allocation and superposition are performed, and the multiple continuous detection accuracy information is weightedly calculated to obtain first detection accuracy information, wherein the sizes of the extreme degrees and the continuous occurrence degrees are positively correlated with the sizes of the weights; multiple impact degrees and multiple impact occurrence degrees of the multiple impact test scenarios are obtained, and the multiple impact detection accuracy information is weightedly calculated to obtain second detection accuracy information; according to the first detection accuracy information and the second detection accuracy information, comprehensive detection accuracy information is calculated as the quality detection result.

[0012] In a possible implementation, multiple impact degrees and multiple impact occurrence degrees of the multiple impact test scenarios are obtained, and weighted calculation is performed on the multiple impact detection accuracy information. The following processing is also performed: the deviation amplitude of two test working environment parameters in the multiple impact test scenarios is calculated as multiple impact degrees; the working environment parameter records within the historical time are retrieved to obtain the occurrence ratios of the multiple impact test scenarios as multiple impact occurrence degrees; double weight allocation and superposition are performed according to the sizes of the multiple impact degrees and the multiple impact occurrence degrees, and the multiple impact detection accuracy information is weighted calculated to obtain second detection accuracy information.

[0013] It is intended to obtain a working environment parameter space within the working environment of the temperature element through a quality detection method for temperature elements proposed in the present application; randomly generate multiple test working environment parameters within the working environment parameter space as multiple continuous test scenarios, and traverse and combine the multiple test working environment parameters, and screen and obtain multiple impact test scenarios based on the working environment parameter records; in multiple continuous test scenarios, continuously test a multi-point temperature element array to obtain multiple continuous temperature test result arrays and multiple continuous image test result arrays; in multiple impact test scenarios, impact test a multi-point temperature element array to obtain multiple impact temperature test result arrays and multiple impact image test result array sequences; based on the multiple continuous temperature test result arrays, analyze and obtain multiple basic continuous detection accuracy information; based on the multiple continuous image test result arrays, analyze and obtain multiple continuous damage trend information; classify and obtain multiple continuous detection influence coefficients; correct the multiple basic continuous detection accuracy information to obtain multiple continuous detection The invention discloses a method for measuring the accuracy of a plurality of impact detection images and the like, and obtains the accuracy information of a plurality of basic impact detection images. The method comprises the following steps: obtaining the accuracy information of a plurality of basic impact detection images and the like, and obtaining the accuracy information of a plurality of impact detection images. The method comprises the following steps: obtaining the accuracy information of a plurality of basic impact detection images and the like, and obtaining the accuracy information of a plurality of impact detection images. The method comprises the following steps: obtaining the accuracy information of a plurality of basic impact detection images and the like, and obtaining the accuracy information of a plurality of impact detection images. The method comprises the following steps: obtaining the accuracy information of a plurality of basic impact detection images and the like, and obtaining the accuracy information of a plurality of impact detection images. The method comprises the following steps: BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the technical solution of the embodiment of the present disclosure, the accompanying drawings of the embodiment of the present disclosure will be briefly introduced below. A flow chart is used in the present application to illustrate the operations performed by the method according to the embodiment of the present application. It should be understood that the preceding or following operations are not necessarily performed precisely in order. On the contrary, various steps can be processed in reverse order or simultaneously as needed. At the same time, other operations can also be added to these processes, or one or more operations can be removed from these processes.

[0015] Figure 1 A schematic flow chart of a method for detecting the quality of a temperature element provided in an embodiment of the present application;

[0016] Figure 2A schematic diagram of a process for obtaining impact detection accuracy information in a quality detection method for a temperature element provided in an embodiment of the present application. DETAILED DESCRIPTION

[0017] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below.

[0018] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings. The described embodiments should not be regarded as limiting the present application. All other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of this application.

[0019] In the following description, reference is made to "some embodiments", which describe a subset of all possible embodiments, but it is understood that "some embodiments" may be the same subset or different subsets of all possible embodiments, and may be combined with each other without conflict, and the terms "first\second" involved are merely to distinguish similar objects and do not represent a specific ordering of objects. The terms "including" and "having" and any variations are intended to cover non-exclusive inclusions, for example, a process, method, product, or server that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or modules that are not clearly listed or inherent to these processes, methods, products, or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those generally understood by technicians in the technical field of this application. The terms used herein are for the purpose of describing the embodiments of the present application only.

[0020] The present application embodiment provides a method for detecting the quality of a temperature element. Figure 1 As shown, the method includes:

[0021] Step S100, obtaining the working environment parameter space in the working environment of the temperature element, wherein the working environment parameter space includes the ambient temperature space and the ambient pressure space. Temperature measurement is performed under high temperature and high pressure in the reactor of the refining industry, and data collection is performed on the temperature space and the ambient pressure space of the refining production environment to construct the working environment parameter space. The working environment parameter space refers to the set of physical environmental conditions in which the temperature element (high-pressure multi-point flexible thermocouple or thermal resistor) is located in actual application. The working environment parameters have a direct impact on the performance and stability of the temperature element. The working environment parameter space includes the ambient temperature space and the ambient pressure space. Specifically, the ambient temperature space refers to the temperature range of the air or other medium surrounding the temperature element, which can range from extremely low temperatures (such as tens of degrees below zero) to extremely high temperatures (such as more than 100 degrees). The temperature It is one of the important factors affecting the performance of electronic components. Changes in temperature may cause changes in parameters such as resistance and capacitance of components, thereby affecting their measurement accuracy and stability; ambient pressure space refers to the pressure range of air or other media around the temperature element, which can range from low pressure (such as high altitude areas) to high pressure (such as deep sea or inside a high-pressure container). For some types of temperature elements, ambient pressure may not be the main factor affecting their performance, but in some specific application scenarios (such as aerospace, deep-sea exploration, etc.), changes in ambient pressure may affect the sealing and thermal conductivity of the temperature element, thereby affecting its measurement accuracy and stability.

[0022] Step S200, randomly generate multiple test working environment parameters in the working environment parameter space as multiple continuous test scenarios, and traverse and combine the multiple test working environment parameters, and screen and obtain multiple impact test scenarios based on the working environment parameter records. In the determined working environment parameter space (including ambient temperature space and ambient pressure space), use a random number generator to randomly select multiple test working environment parameters, representing various working environment conditions that the temperature element may encounter, and simulate a variety of different working environments to comprehensively evaluate the performance and stability of the temperature element under different conditions, and then set the randomly generated test working environment parameters as continuous test scenarios, that is, the temperature element is continuously tested for a long time under these parameter conditions. Through continuous testing, the performance changes and stability performance of the temperature element under long-term operation can be observed, so as to evaluate its durability and reliability; traverse and combine the randomly generated multiple test working environment parameters to form more test scenarios, that is, different parameters such as temperature and pressure are used. Cross-matching is performed to cover a wider range of working environments and generate more diverse test scenarios. Finally, in the traversed combined test scenarios, multiple impact test scenarios are screened out according to the working environment parameter records (retrieval and screening of test working environment parameter combinations that have appeared in historical production, that is, situations where temperature and pressure have changed). (Including temperature and pressure before the change and temperature and pressure after the change. Sudden changes in temperature and pressure will have an impact on the components and affect the temperature detection accuracy). These scenarios represent extreme or sudden environmental conditions. Impact test scenarios emphasize the suddenness and extremeness of the test, and are designed to simulate the performance of temperature components under extreme or sudden situations. Through impact testing, the tolerance and recovery capabilities of temperature components under extreme conditions are evaluated, thereby further verifying their reliability and stability.

[0023] In a possible implementation, step S200 further includes step S210, randomly generating multiple test working environment parameters in the working environment parameter space as multiple continuous test scenarios. Using a random number generator, randomly selecting multiple spatial points in the working environment parameter space, each point represents a set of specific working environment parameter values, and the randomly generated working environment parameter combination is used as a continuous test scenario, that is, the temperature element will be subjected to a long-term operation test under these continuous test scenarios (simulated environmental conditions) to evaluate its stability and durability. It also includes step S220, traversing and combining the multiple test working environment parameters to obtain multiple test working environment parameter combinations. The multiple test working environment parameters are traversed in pairs, that is, every two test working environment parameters are combined, each combination represents a set of specific working environment parameter values, and finally multiple test working environment parameter combinations are obtained, which are used to simulate the test environment of the temperature element. It also includes step S230, obtaining the working environment parameter record in the historical time, using the multiple test working environment parameter combinations to perform impact scenario retrieval in the working environment parameter record, and constructing the multiple test working environment parameter combinations that have appeared into multiple impact test scenarios. Historical working environment parameter records refer to the recorded data of the actual working environment parameters of the temperature element over a period of time in the past (that is, the situations in which temperature and pressure have changed). Shock scenario retrieval refers to using the previously generated test working environment parameter combination to search in the historical working environment parameter records to find test scenarios that are similar to the extreme or abnormal conditions recorded in the historical data. Scenarios that have actually appeared in the historical data and are similar to the test working environment parameter combination are constructed as shock test scenarios, which represent the extreme or sudden situations that the temperature element may face, and are used to evaluate its shock resistance and recovery capabilities.

[0024] Step S300, in the multiple continuous test scenarios, the multi-point temperature element array is continuously tested to obtain multiple continuous temperature test result arrays and multiple continuous image test result arrays. In the multiple impact test scenarios, the multi-point temperature element array is impact tested to obtain multiple impact temperature test result arrays and multiple impact image test result array sequences. In the multiple continuous test scenarios and impact test scenarios, the multi-point temperature element array is continuously tested (the continuous test time is set, such as 10 minutes) and impact tested respectively, and the temperature and element images of the multiple element tests after the continuous test are obtained (high temperature and high pressure will cause the surface of the temperature element to crack and affect the test accuracy), and the images before and after the impact test are acquired as the impact image test result array sequence. Specifically, the multi-point temperature element array includes multiple multi-point temperature elements. In each continuous test scenario, the multi-point temperature element array is continuously tested for a long time. During the test, the temperature readings of each temperature element at different time points are recorded to form multiple continuous temperature test result arrays, which include the temperature changes of each temperature element in different environments, in order to more intuitively observe the working status of the temperature element. Or detect potential damage, collect images through cameras or other image acquisition devices, record image information of temperature elements during the test, form multiple continuous image test result arrays, and record abnormal conditions such as deformation, cracks, and color changes of temperature elements; in each impact test scenario, perform a short-term impact test on the multi-point temperature element array. During the test, simulate extreme or sudden temperature changes, pressure changes and other conditions, observe and record the response of the temperature elements, and form multiple impact temperature test result arrays and multiple impact image test result array sequences. The impact temperature test result array records the temperature changes of the temperature elements during the impact process, and the impact image test result array sequence records the image information of the temperature elements during the impact process, including the occurrence of abnormal conditions such as deformation, cracks, and color changes.

[0025] Step S400: Analyze and obtain multiple basic continuous detection accuracy information according to the multiple continuous temperature test result arrays, analyze and obtain multiple continuous damage trend information according to the multiple continuous image test result arrays, classify and obtain multiple continuous detection influence coefficients, and correct the multiple basic continuous detection accuracy information to obtain multiple basic continuous detection accuracy information. According to the multiple continuous temperature test result arrays, analyze the temperature of each temperature element at different time points, compare it with the standard value, calculate the temperature deviation, and then obtain multiple basic continuous detection accuracy information (such as 1-the amplitude of the temperature deviation); according to the multiple continuous image test result arrays, perform image processing and feature extraction on the image test results in each continuous test scenario, identify the potential damage or abnormality of the temperature element, observe the damage state change of the temperature element at different time points through time series analysis and other methods, obtain continuous damage trend information, classify and obtain multiple continuous detection influence coefficients based on the damage trend information, and correct the multiple basic continuous detection accuracy information obtained. For example, when the damage trend is 0, it means that it does not affect the temperature detection, and the influence coefficient is 1; the greater the damage trend, the smaller the continuous detection influence coefficient, the smaller the correction amplitude, and the greater the temperature detection impact. By comprehensively analyzing the continuous temperature test result array and the continuous image test result array, the basic continuous detection accuracy information and continuous damage trend information are obtained, and the continuous detection influence coefficient is obtained by classification to correct the basic accuracy information, thereby improving the accuracy and reliability of temperature component quality detection.

[0026] In one possible implementation, Figure 2 As shown, step S400 further includes step S410, calculating and obtaining multiple average test temperatures according to the multiple continuous temperature test result arrays. The continuous temperature test result array is a data set containing the temperatures of multiple temperature test points at different time points. For each continuous test scene, the average value of all its temperatures is calculated to obtain multiple average test temperatures, which reflects the average working temperature of the temperature element in the test scene. It also includes step S420, calculating and obtaining multiple basic continuous detection accuracy information according to the multiple average test temperatures and the ambient temperature of the multiple continuous test scenes. The ambient temperature of the multiple continuous test scenes refers to the ambient temperature set or actually existing in each test scene. Specifically, for each continuous test scene, the difference between the average test temperature and the ambient temperature, that is, the temperature deviation, is calculated, which quantifies the closeness between the temperature element measurement value and the actual ambient temperature, and multiple basic continuous detection accuracy information is calculated. If the ambient temperature in the test scene exceeds the design measurement range of the temperature element, then even if the temperature deviation is very small, it may be considered that its detection accuracy is not high.

[0027] The method further includes step S430, in which the temperature element damage is identified for the plurality of continuous image test result arrays to obtain a plurality of continuous damage trend information. The image is analyzed using image processing technology (such as image segmentation, feature extraction, pattern recognition, etc.), and feature vectors related to damage are extracted to identify damaged areas on the temperature element, such as cracks, deformations, color changes, etc. The identified damaged areas are marked and tracked on the time series by time series analysis and other methods, and the damage state changes of the temperature element at different time points are observed to evaluate its continuous damage trend (i.e., the law and trend of the temperature element damage changes over time), including information such as the rate of change of the damage degree, the direction and speed of damage expansion, etc.

[0028] The method also includes step S440, inputting the multiple continuous damage trend information into a continuous damage detection classifier, and classifying them to obtain multiple continuous detection influence coefficients, wherein the continuous damage detection classifier is constructed based on the mapping relationship between the sample continuous damage trend information set and the sample continuous detection influence coefficient set. The continuous damage detection classifier is a pre-trained model. Based on sample data (i.e., known, historical or experimentally obtained continuous damage trend information and corresponding continuous detection influence coefficients), the classifier learns and establishes a mapping relationship between continuous damage trend information and continuous detection influence coefficients. When receiving new continuous damage trend information, the classifier can predict or infer the corresponding continuous detection influence coefficient. Specifically, a large amount of sample data is collected, including continuous damage trend information of different temperature elements under different damage states, and the actual impact of these damage states on the detection accuracy of the temperature element (i.e., sample continuous detection influence coefficients). Features related to damage trends and detection impacts are extracted from the sample data, such as morphological features, development speed, location, etc. of damage, as well as model and working environment of the temperature element. The extracted features and the corresponding sample continuous detection influence coefficients are used to train the classifier. During the training process, the classifier learns and optimizes internal parameters, establishes a mapping relationship from continuous damage trend information to continuous detection influence coefficients, and inputs multiple continuous damage trend information into the continuous damage detection classifier. The continuous damage detection classifier classifies and obtains multiple continuous detection influence coefficients according to the learned mapping relationship, which reflects the specific impact of the current damage state on the detection accuracy of the temperature element.

[0029] The method further includes step S450, using the multiple continuous detection influence coefficients to perform correction calculations on the multiple basic continuous detection accuracy information to obtain multiple continuous detection accuracy information. Each basic continuous detection accuracy information is matched with the corresponding continuous detection influence coefficient through a timestamp, and then the continuous detection influence coefficient is used to perform correction calculations on the basic continuous detection accuracy information, and the evaluation result of its detection accuracy is adjusted according to the actual damage state of the temperature element, for example, the basic continuous detection accuracy information is multiplied by the corresponding continuous detection influence coefficient to obtain multiple continuous detection accuracy information.

[0030] In a possible implementation, step S430 further includes step S431, collecting a set of sample continuous image test results and a set of sample continuous damage trend information according to the damage identification data of the temperature element. According to the damage identification data of the temperature element, continuous image test results of the temperature element in different states are collected, and these images record the appearance, color, texture and other features of the temperature element at different time points and under different working conditions; corresponding to the image test results, information about the damage trend of the temperature element in these images is collected, such as manually annotated damage type, damage degree, damage development speed, etc., and damage feature parameters automatically extracted by image processing technology. This information will be used to train the continuous damage identifier to establish a mapping relationship between image features and damage trends.

[0031] The method further includes step S432, using the sample continuous image test result set and the sample continuous damage trend information set to train a continuous damage identifier based on a convolutional neural network. A recognition model is constructed based on a convolutional neural network, and a suitable CNN architecture is selected, including multiple convolutional layers, pooling layers, activation functions, and fully connected layers. The number of network layers, the number of filters in each layer, the filter size, etc. are designed. The sample continuous image test result set and the sample continuous damage trend information set are divided into a training set and a test set. The training set is used to train the CNN model so that it can learn the mapping relationship between images and damage trends. The test set is used to evaluate the performance of the model to ensure its accuracy in practical applications. In the process of training the CNN model with the training set data, the model will continuously adjust its internal parameters (such as weights and biases) to minimize the prediction error (such as classification error rate or regression error). The continuous damage identifier is finally obtained through multiple iterations of training.

[0032] The method further includes step S433, using the continuous damage identifier to identify the continuous image test results in the plurality of continuous image test result arrays, and obtaining a plurality of component continuous damage trend information arrays. The trained continuous damage identifier is applied to the new, continuous image test result array, and the identifier analyzes each image one by one, so as to identify the current damage state of each temperature component, and further infer its damage trend, and finally, the identification results are organized into a plurality of component continuous damage trend information arrays, each array corresponding to a damage trend information sequence of a temperature component.

[0033] The method further includes step S434, wherein multiple continuous damage trend information is calculated based on the multiple element continuous damage trend information arrays. The continuous damage state changes of each temperature element are quantitatively analyzed and processed to extract key information about its damage trend, which is used to evaluate the damage speed of the element, predict the future damage state, and determine whether the element needs maintenance or replacement. Specifically, for the continuous damage trend information array of each element, trend analysis is performed to observe the change law of the damage state over time, identify whether the damage is gradually aggravated, remains stable, or is alleviated, and the results of the trend analysis are quantified into specific values, such as calculating the damage speed (such as the increase in the degree of damage per unit time), damage acceleration (the rate of change of the damage speed), and remaining life estimation (based on the current damage trend to predict how long the element can work normally), etc. Representative features are extracted from the continuous damage trend information of each element, such as numerical (such as the maximum damage degree, average damage speed), or non-numerical (such as the change of damage type, the emergence of a specific damage mode), and the extracted features are combined to comprehensively evaluate the continuous damage trend of each element to obtain multiple continuous damage trend information.

[0034] Step S500: Analyze and obtain multiple basic impact detection accuracy information based on the multiple impact temperature test result arrays, perform impact dynamic damage analysis based on the multiple impact image test result array sequences to obtain multiple impact damage trend information, classify and obtain multiple impact detection influence coefficients, correct the multiple basic impact detection accuracy information, and obtain multiple impact detection accuracy information. According to multiple shock temperature test result arrays, the shock test temperature value of the temperature element is compared with the expected standard value (such as calculating the error rate, etc.) to obtain the basic shock detection accuracy information under each test condition; based on the shock image test result array sequence, the dynamic damage of the material during the shock process is analyzed through image processing technology and data analysis methods (such as image recognition, feature extraction, time series analysis, etc.), and multiple shock damage trend information is obtained. In the process of analyzing the shock damage trend, it is found that certain factors (such as temperature, material type, impact direction, etc.) have a significant impact on the shock detection results. Through classification and regression analysis and other methods, the degree of influence of these factors can be quantified, that is, multiple shock detection influence coefficients are obtained to describe the specific influence of different factors on the shock detection results. Then, the shock detection influence coefficients obtained above are used to correct the basic shock detection accuracy information, eliminate or reduce the influence of external factors on the test results, and improve the accuracy and reliability of shock detection.

[0035] In a possible implementation, step S500 further includes step S510, calculating and obtaining multiple pre-impact average test temperatures and multiple post-impact average test temperatures according to the multiple shock temperature test result arrays. In the shock temperature test result array, firstly, the time point of the shock event is identified by analyzing the sudden change and abnormal peak of the temperature curve, and the shock temperature test result array is divided into two parts, before the shock and after the shock, based on the time point of the shock event, for each shock temperature test result array, the temperature test results before and after the shock event are considered respectively, and multiple pre-impact average test temperatures and multiple post-impact average test temperatures are calculated.

[0036] It also includes step S520, according to the multiple pre-impact average test temperatures and the multiple post-impact average test temperatures, combined with the temperature combination of the multiple impact test scenes, calculate and obtain multiple pre-impact detection accuracy information and multiple post-impact detection accuracy information, and calculate the average to obtain multiple basic impact detection accuracy information. The average test temperature before and after the impact is used, combined with the temperature combination of the impact test scene (i.e., the temperature setting under different impact conditions), to evaluate the accuracy of the results of the temperature element before and after the impact test, specifically, the actual measured temperature is compared with the standard value, and the error or deviation is calculated to obtain the detection accuracy information before and after the impact, that is, to obtain multiple pre-impact detection accuracy information and multiple post-impact detection accuracy information, and then calculate the average of the pre-impact and post-impact detection accuracy information respectively, to obtain multiple basic impact detection accuracy information, which represents the average accuracy level of the temperature element before and after the impact test in different impact test scenes.

[0037] The method further includes step S530, performing dynamic damage analysis of the impact of the temperature element according to the plurality of impact image test result array sequences, and obtaining a plurality of impact damage trend information. The array sequence of the impact image test results is preprocessed, including denoising, contrast enhancement, image registration, etc., to improve the image quality, and features related to the impact damage of the temperature element (such as cracks, deformation, shedding, temperature abnormality area, etc.) are extracted from the image using image processing technology (such as edge detection, texture analysis, shape recognition, etc.), and the extracted damage features are arranged in chronological order, and their changing trends during the impact process are analyzed, and statistical analysis is performed on the damage features under a plurality of impact test scenarios, and statistical quantities such as average value and standard deviation are calculated to evaluate the prevalence and severity of the damage, and obtain a plurality of impact damage trend information about the temperature element, such as damage type, damage location, damage degree, damage development speed, etc.

[0038] It also includes step S540, inputting the multiple impact damage trend information into the impact loss detection classifier, and classifying to obtain multiple impact detection influence coefficients, wherein the impact loss detection classifier is constructed based on the mapping relationship between the sample impact damage trend information set and the sample impact detection influence coefficient set. A classifier model is trained based on known sample data (i.e., the sample impact damage trend information set and the corresponding sample impact detection influence coefficient set), and then the model is used to classify and predict new impact damage trend information to obtain the corresponding impact detection influence coefficient. Specifically, the sample impact damage trend information set is a set of impact test data that has been collected and organized, which contains damage trend information of temperature components under different impact conditions. The sample impact detection influence coefficient set corresponds to the sample impact damage trend information, and is a set containing the performance influence coefficients of temperature components under corresponding impact conditions. Machine learning algorithms (such as decision trees, random forests, and support vector machines) are used. , neural networks, etc.) to build a classifier model, taking the sample impact damage trend information set as input features and the sample impact detection influence coefficient set as the target output (or label). Through the training process, the classifier learns the mapping relationship from impact damage trend information to impact detection influence coefficient. During the training stage, the classifier will iteratively adjust its internal parameters (such as weights, biases, etc.) to minimize the difference between the predicted value and the actual value (i.e., optimize the loss function). After the training is completed, an impact loss detection classifier is obtained. Multiple impact damage trend information is input into the impact loss detection classifier, and classification prediction is performed using the mapping relationship to output multiple impact detection influence coefficients.

[0039] The method further includes step S550, using the multiple impact detection influence coefficients to perform correction calculations on the multiple basic impact detection accuracy information to obtain multiple impact detection accuracy information. For each basic impact detection accuracy information, correction calculations are performed according to its corresponding impact detection influence coefficient to obtain multiple impact detection accuracy information, which can more accurately reflect the actual performance of the temperature element under impact conditions.

[0040] In a possible implementation, step S530 further includes step S531, performing image extraction and downsampling processing on the multiple impact image test result array sequences according to the first step length, respectively, to obtain multiple first impact image test result array sets. According to the preset first step length (usually a fixed time interval or frame number interval), extract images from each impact image test result array sequence, ignore other unselected images in the sequence, thereby reducing the amount of data, and perform downsampling processing on each extracted image, specifically, using average pooling, maximum pooling or bilinear interpolation, etc., to reduce the image resolution by reducing the number of pixels in the image, while retaining the main features of the image, further reducing the amount of data, and organizing all images after image extraction and downsampling processing to form multiple first impact image test result array sets, each array set corresponding to an original impact image test result array sequence.

[0041] The method further includes step S532, wherein the plurality of impact image test result array sequences are respectively subjected to image extraction according to a second step length to obtain a plurality of second impact image test result array sets, wherein the second step length is greater than the first step length. Images are extracted from each impact image test result array sequence using a larger time interval or frame number interval (i.e., the second step length). Since the second step length is greater than the first step length, fewer images are extracted, which means that in the sequence, the interval between adjacent extracted images is larger, thereby further reducing the amount of data. All images extracted according to the second step length are organized to form a plurality of second impact image test result array sets.

[0042] The method further includes step S533, constructing and training an impact dynamic damage analyzer based on a slowfast network, wherein the impact dynamic damage analyzer is trained to convergence based on a plurality of sample first impact image test result array sets, a plurality of sample second impact image test result array sets, and a plurality of sample element impact damage trend information array sets. The slowfast network is composed of two parallel convolutional neural networks (CNNs), a slow channel and a fast channel, wherein the slow channel processes low frame rate or high resolution images to extract static or slowly changing features (such as the initial form of the element, background, etc.), and the fast channel processes high frame rate or low resolution images to capture dynamically changing information (such as deformation and speed change during the impact process, etc.); respectively from a plurality of first impact image test result array sets and a plurality of second impact image test result array sets. A part is extracted as a plurality of sample first impact image test result array sets, a plurality of sample second impact image test result array sets, and a plurality of sample component impact damage trend information array sets, which are data about the degree of damage of the component after impact. The image is standardized, such as resizing and normalization, and then converted into a format suitable for SlowFast network input. Specifically, a suitable loss function is selected to measure the difference between the damage trend predicted by the network and the actual damage trend. Hyperparameters such as learning rate, batch size, and number of training rounds are set. Iterative training is performed using the prepared sample data set. During the training process, the network parameters are optimized through the back-propagation algorithm to minimize the loss function until the model converges to obtain an impact dynamic damage analyzer.

[0043] The method further includes step S534, wherein the plurality of first impact image test result array sets and the plurality of second impact image test result array sets are input and identified by the impact dynamic damage analyzer to obtain a plurality of component impact damage trend information arrays. The plurality of first impact image test result array sets and the plurality of second impact image test result array sets are used as input data and input into the impact dynamic damage analyzer for analysis and identification, that is, the impact dynamic damage analyzer extracts features from the input images, matches the input images with known impact damage patterns based on the extracted features, obtains a plurality of component impact damage trend information arrays, and determines the current impact damage status of the components.

[0044] The method further includes step S535, wherein multiple impact damage trend information is calculated based on the multiple component impact damage trend information arrays. The array data containing the component impact damage trend is processed and analyzed to extract key information and trends about the impact damage, for example, the impact damage trend information array is decomposed into trend items using a time series decomposition method (such as seasonal decomposition, trend item extraction, etc.), and based on the trend items, a suitable prediction model (such as linear regression, time series model, etc.) is used to predict the future trend of the impact damage.

[0045] Step S600, obtain multiple extremes and multiple continuous occurrences of the multiple continuous test scenarios, and obtain multiple impacts and multiple impact occurrences of the multiple impact test scenarios, and perform weighted calculations on the multiple continuous detection accuracy information and multiple impact detection accuracy information, respectively, to obtain detection accuracy information as a quality detection result. Calculate the extremes according to the deviation of the temperature and pressure in the continuous test scenario from the normal temperature and pressure (such as the temperature and pressure of normal operation), calculate the proportion of occurrence according to the frequency of occurrence of the multiple continuous test scenarios in the entire test cycle, and then obtain the continuous occurrence; calculate the impact degree according to the change amplitude of the temperature and pressure before and after the change in the impact test scenario (to evaluate the intensity, rate and other parameters of the impact force in each impact test scenario), calculate the frequency or number of occurrences of each impact test scenario during the test process, that is, the proportion of occurrence in work, and calculate the impact occurrence. The extreme degree and impact degree reflect the severity of the test scenario and should be assigned a higher weight. The continuous occurrence degree and impact occurrence degree reflect the stability and frequency of the test scenario and may contribute relatively less to the weight. Weighted calculation is performed on multiple continuous detection accuracy information and multiple impact detection accuracy information respectively. Specifically, weights are assigned to corresponding continuous detection accuracy information or impact detection accuracy information according to the extreme degree, continuous occurrence degree (for continuous test scenarios) or impact degree, impact occurrence degree (for impact test scenarios) of each test scenario. The detection accuracy information under different test scenarios is calculated using the weighted average method to obtain comprehensive detection accuracy information. For example, assuming that there are n continuous test scenarios and m impact test scenarios, a weight coefficient (based on factors such as extreme degree, continuous occurrence degree, impact degree, impact occurrence degree, etc.) can be set for each scenario. For each scenario, its detection accuracy information is multiplied by the corresponding weight coefficient. Finally, all weighted detection accuracy information is added together to obtain the detection accuracy information as the final detection result of product quality detection, which is used to evaluate the overall performance and reliability of the product.

[0046] In a possible implementation, step S600 further includes step S610, calculating the deviation amplitudes between the multiple continuous test scenarios and conventional environmental parameters as multiple extreme degrees. Calculate the deviation amplitudes between the multiple continuous test scenarios and conventional environmental parameters (including temperature and pressure) as multiple extreme degrees, wherein the larger the deviation, the more extreme the test scenario, and the higher the extreme degree. Also included is step S620, retrieving the occurrence ratios of the multiple continuous test scenarios from the working environment parameter records in the historical time as multiple continuous occurrence degrees. Retrieve data (environmental parameter combinations) matching the multiple continuous test scenarios from the working environment parameter records, calculate the occurrence ratios of the continuous test scenarios in the historical time, and divide the number of occurrences of each scenario by the total number of records in the entire historical time, which reflects the frequency of the test scenario in the actual working environment, that is, the continuous occurrence degree.

[0047] The method further includes step S630, wherein a double weight distribution and superposition are performed according to the magnitudes of the multiple extreme degrees and the multiple continuous occurrence degrees, and the multiple continuous detection accuracy information is weightedly calculated to obtain the first detection accuracy information, wherein the magnitudes of the extreme degree and the continuous occurrence degree are positively correlated with the magnitudes of the weights. The magnitudes of the extreme degree and the continuous occurrence degree are positively correlated with the magnitudes of the weights. That is, the test scenarios with higher extreme degrees or higher continuous occurrence degrees should be given greater weights in the weighted calculation. These scenarios either represent extreme test conditions or frequently appear in practical applications. A corresponding weight is distributed according to the extreme degree of each test scenario, and at the same time, a corresponding weight is also distributed according to the continuous occurrence degree of each test scenario. The extreme degree weight and the continuous occurrence degree weight are superimposed to form a comprehensive weight. The weights obtained by the above distribution and superposition are used to perform weighted calculation on the multiple continuous detection accuracy information to obtain the first detection accuracy information, which reflects the continuous detection accuracy of the temperature element under the test scenarios with different extreme degrees and continuous occurrence degrees.

[0048] The method further includes step S640, obtaining multiple shock degrees and multiple shock occurrence degrees of the multiple shock test scenarios, performing weighted calculation on the multiple shock detection accuracy information, and obtaining second detection accuracy information. For multiple shock test scenarios, firstly, the shock degree data of each scenario is collected, and at the same time, the occurrence degree information of each shock scenario in the historical data is obtained, and then the multiple shock detection accuracy information is weighted calculated, for example, the shock detection accuracy information of each shock scenario is multiplied by its corresponding weight (the weight is jointly determined by the shock degree and the shock occurrence degree, or the weights can be set separately and then combined), and the size of the weight is positively correlated with the size of the shock degree and the shock occurrence degree, that is, the shock test scenario with higher shock degree or higher shock occurrence degree is given a greater weight in the weighted calculation, and finally the second detection accuracy information is obtained, which reflects the detection accuracy and reliability of the temperature element under different shock scenarios.

[0049] The method further includes step S650, wherein comprehensive detection accuracy information is calculated based on the first detection accuracy information and the second detection accuracy information as a quality detection result. The first detection accuracy information and the second detection accuracy information are integrated to obtain comprehensive detection accuracy information that can fully reflect the performance of the temperature element. For example, different weights are assigned to the first detection accuracy information and the second detection accuracy information, and the weights are used to perform weighted averaging on the first detection accuracy information and the second detection accuracy information to obtain comprehensive detection accuracy information as a quality detection result, which can accurately and fully reflect the accuracy and reliability of the temperature element quality detection.

[0050] In a possible implementation, step S640 further includes step S641, calculating the deviation amplitude of the two test working environment parameters in the multiple impact test scenarios as multiple impact degrees. For each impact test scenario, the deviation amplitude of the two test working environment parameters is calculated, and the size of the deviation amplitude reflects the degree of deviation between the test scenario and the normal working environment, that is, the intensity or severity of the impact, and the obtained deviation amplitude is used as the impact degree of the impact test scenario. It also includes step S642, retrieving the occurrence ratio of the multiple impact test scenarios in the working environment parameter records within the historical time, as multiple impact occurrence degrees. Within the historical time range, retrieve and collect the recorded data of the working environment parameters, for each impact test scenario, count the number of times it appears in the historical data, and calculate its proportion of the total records, that is, the occurrence degree of the impact test scenario, which reflects its commonality or persistence in the actual working environment.

[0051] The method further includes step S643, wherein a double weight distribution and superposition are performed according to the magnitudes of the multiple impact degrees and the multiple impact occurrence degrees, and a weighted calculation is performed on the multiple impact detection accuracy information to obtain second detection accuracy information. According to the magnitudes of the impact degree and the impact occurrence degree, corresponding weights are respectively assigned to them, and scenes with higher impact degrees or higher impact occurrence degrees are assigned greater weights, and the impact degree weights and the impact occurrence weights are superimposed to form a comprehensive weight, and the comprehensive weights obtained by the distribution and superposition are used to perform weighted calculation on multiple impact detection accuracy information, specifically, the impact detection accuracy information of each impact scene is multiplied by its corresponding comprehensive weight, and then all weighted accuracy information is summed and averaged to obtain the second detection accuracy information, which reflects the detection accuracy and reliability of the temperature element under different impact scenes.

[0052] The above specific implementations do not constitute a limitation on the protection scope of this application. It should be understood by those skilled in the art that various modifications, combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of this application should be included in the protection scope of this application.

Claims

1. A method for detecting the quality of a temperature element, characterized in that: The method comprises: Acquire a working environment parameter space in a working environment of a temperature element, wherein the working environment parameter space includes an environment temperature space and an environment pressure space; Randomly generate multiple test working environment parameters in the working environment parameter space as multiple continuous test scenarios, traverse and combine the multiple test working environment parameters, and select and obtain multiple impact test scenarios according to the working environment parameter records; In the multiple continuous test scenarios, the multi-point temperature element array is continuously tested to obtain multiple continuous temperature test result arrays and multiple continuous image test result arrays; in the multiple impact test scenarios, the multi-point temperature element array is impact tested to obtain multiple impact temperature test result arrays and multiple impact image test result array sequences; Analyze and obtain multiple basic continuous detection accuracy information according to the multiple continuous temperature test result arrays, analyze and obtain multiple continuous damage trend information according to the multiple continuous image test result arrays, classify and obtain multiple continuous detection influence coefficients, and correct the multiple basic continuous detection accuracy information to obtain multiple continuous detection accuracy information; Analyze and obtain multiple basic shock detection accuracy information according to the multiple shock temperature test result arrays, perform shock dynamic damage analysis according to the multiple shock image test result array sequences to obtain multiple shock damage trend information, classify and obtain multiple shock detection influence coefficients, and correct the multiple basic shock detection accuracy information to obtain multiple shock detection accuracy information; Acquire multiple extreme degrees and multiple continuous occurrence degrees of the multiple continuous test scenarios, and acquire multiple impact degrees and multiple impact occurrence degrees of the multiple impact test scenarios, and respectively perform weighted calculations on the multiple continuous detection accuracy information and the multiple impact detection accuracy information to obtain detection accuracy information as a quality detection result; Analyzing and obtaining a plurality of basic continuous detection accuracy information according to the plurality of continuous temperature test result arrays, analyzing and obtaining a plurality of continuous damage trend information according to the plurality of continuous image test result arrays, classifying and obtaining a plurality of continuous detection influence coefficients, and correcting the plurality of basic continuous detection accuracy information, including: Calculating and obtaining a plurality of average test temperatures according to the plurality of continuous temperature test result arrays; Calculate and obtain multiple basic continuous detection accuracy information according to the multiple average test temperatures and the ambient temperatures of the multiple continuous test scenarios; Performing temperature component damage identification on the plurality of continuous image test result arrays to obtain a plurality of continuous damage trend information; Input the plurality of continuous damage trend information into a continuous damage detection classifier, and classify and obtain a plurality of continuous detection influence coefficients, wherein the continuous damage detection classifier is constructed based on a mapping relationship between a set of sample continuous damage trend information and a set of sample continuous detection influence coefficients; Using the multiple continuous detection influence coefficients, performing correction calculations on the multiple basic continuous detection accuracy information to obtain multiple continuous detection accuracy information; Performing temperature component damage identification on the plurality of continuous image test result arrays to obtain a plurality of continuous damage trend information, including: According to the damage identification data of the temperature element, a set of continuous image test results of the sample is collected, and a set of continuous damage trend information of the sample is collected; Using the sample continuous image test result set and the sample continuous damage trend information set, based on a convolutional neural network, a continuous damage identifier is trained; Using the continuous damage identifier, the continuous image test results in the plurality of continuous image test result arrays are identified to obtain a plurality of component continuous damage trend information arrays; Calculating and obtaining a plurality of continuous damage trend information according to the plurality of element continuous damage trend information arrays; According to the plurality of shock temperature test result arrays, a plurality of basic shock detection accuracy information is obtained by analysis; according to the plurality of shock image test result array sequences, a shock dynamic damage analysis is performed to obtain a plurality of shock damage trend information; a plurality of shock detection influence coefficients are obtained by classification; and the plurality of basic shock detection accuracy information is corrected, including: Calculating and obtaining a plurality of pre-impact average test temperatures and a plurality of post-impact average test temperatures according to the plurality of impact temperature test result arrays; According to the multiple pre-impact average test temperatures and the multiple post-impact average test temperatures, in combination with the temperature combinations of the multiple impact test scenarios, multiple pre-impact detection accuracy information and multiple post-impact detection accuracy information are calculated, and the average is calculated to obtain multiple basic impact detection accuracy information; According to the plurality of impact image test result array sequences, a dynamic impact damage analysis of the temperature element is performed to obtain a plurality of impact damage trend information; Input the multiple impact damage trend information into an impact loss detection classifier, and classify to obtain multiple impact detection influence coefficients, wherein the impact loss detection classifier is constructed based on a mapping relationship between a sample impact damage trend information set and a sample impact detection influence coefficient set; Using the multiple impact detection influence coefficients, performing correction calculations on the multiple basic impact detection accuracy information to obtain multiple impact detection accuracy information; According to the plurality of impact image test result array sequences, a temperature element impact dynamic damage analysis is performed to obtain a plurality of impact damage trend information, including: Performing image extraction and down-sampling processing on the plurality of shock image test result array sequences respectively according to the first step length to obtain a plurality of first shock image test result array sets; performing image extraction on the plurality of shock image test result array sequences respectively according to a second step length to obtain a plurality of second shock image test result array sets, wherein the second step length is greater than the first step length; Based on the slowfast network, an impact dynamic damage analyzer is constructed and trained, wherein the impact dynamic damage analyzer is trained to convergence based on a plurality of sample first impact image test result array sets, a plurality of sample second impact image test result array sets, and a plurality of sample element impact damage trend information array sets; Using the impact dynamic damage analyzer, input recognition is performed on the plurality of first impact image test result array sets and the plurality of second impact image test result array sets to obtain a plurality of component impact damage trend information arrays; A plurality of impact damage trend information are calculated and obtained based on the plurality of element impact damage trend information arrays.

2. A method for detecting the quality of a temperature element according to claim 1, characterized in that: A plurality of test working environment parameters are randomly generated in the working environment parameter space as a plurality of continuous test scenarios, and the plurality of test working environment parameters are traversed and combined, and a plurality of impact test scenarios are screened and obtained according to the working condition parameter records, including: Randomly generating multiple test working environment parameters in the working environment parameter space as multiple continuous test scenarios; Traversing and combining the multiple test working environment parameters to obtain multiple test working environment parameter combinations; The working environment parameter records within the historical time are obtained, the multiple test working environment parameter combinations are used to perform impact scenario retrieval in the working environment parameter records, and the multiple test working environment parameter combinations that have appeared are constructed into multiple impact test scenarios.

3. The quality detection method of a temperature element according to claim 1, characterized in that: Acquiring multiple extreme degrees and multiple continuous occurrence degrees of the multiple continuous test scenarios, and acquiring multiple impact degrees and multiple impact occurrence degrees of the multiple impact test scenarios, and performing weighted calculations on the multiple continuous detection accuracy information and the multiple impact detection accuracy information, respectively, including: Calculating deviations between the plurality of continuous test scenarios and conventional environmental parameters as a plurality of extreme degrees; Recording working environment parameters in historical time, retrieving and obtaining occurrence ratios of the plurality of continuous test scenarios as a plurality of continuous occurrence degrees; According to the magnitudes of the multiple extreme degrees and the multiple continuous occurrence degrees, double weight allocation and superposition are performed, and weighted calculation is performed on the multiple continuous detection accuracy information to obtain first detection accuracy information, wherein the magnitudes of the extreme degrees and the continuous occurrence degrees are positively correlated with the magnitudes of the weights; Acquire multiple impact degrees and multiple impact occurrence degrees of the multiple impact test scenes, perform weighted calculation on the multiple impact detection accuracy information, and obtain second detection accuracy information; Comprehensive detection accuracy information is calculated based on the first detection accuracy information and the second detection accuracy information as a quality detection result.

4. A method for detecting the quality of a temperature element according to claim 3, characterized in that: Acquiring a plurality of shock degrees and a plurality of shock occurrence degrees of the plurality of shock test scenarios, and performing weighted calculation on the plurality of shock detection accuracy information, including: Calculating the deviation amplitudes of two test working environment parameters in the multiple impact test scenarios as multiple impact degrees; Recording working environment parameters in historical time, retrieving and obtaining occurrence ratios of the plurality of impact test scenarios as a plurality of impact occurrence degrees; According to the sizes of the multiple impact degrees and the multiple impact occurrence degrees, double weight distribution and superposition are performed, and weighted calculation is performed on the multiple impact detection accuracy information to obtain second detection accuracy information.