System and method for detecting dynamic defects of cooling pipeline of new energy automobile

Abstract

The invention discloses a new energy automobile cooling pipeline dynamic defect detection system and method, and relates to the technical field of new energy automobiles. Comprising dynamic load application, dynamic image acquisition and intelligent image analysis. By applying a dynamic pressure load spectrum simulating a real working condition, closed or hidden tiny defects in a static state can be actively excited, and the problem that early hidden dangers are difficult to find in traditional static detection is solved. In cooperation with in-situ continuous imaging of a high-resolution dynamic X-ray microscope, the dynamic response process of the defect along with load change can be completely captured, the risk of missing detection of a single static image is avoided, and the sensitivity and comprehensiveness of defect identification are remarkably improved. Through the synergistic effect of defect movement fingerprint construction, a classification recognition model and an abnormal early warning mechanism, the defect type can be accurately recognized, and the failure risk of the cooling pipeline can be pre-judged in advance by analyzing the defect evolution trend and quantitatively predicting the defect extension life.

Description

Technical Field

[0001] This invention relates to the field of new energy vehicle technology, and in particular to a dynamic defect detection system and method for cooling pipes in new energy vehicles. Background Technology

[0002] With the global energy crisis and increasing environmental awareness, new energy vehicles, as a core solution to replace traditional fuel vehicles, have become the mainstream direction of the automotive industry. Among these, the power battery and drive motor, as the core power components of new energy vehicles, directly determine the driving safety and lifespan of the entire vehicle. During the operation of new energy vehicles, the charging and discharging process of the power battery and the high-speed operation of the drive motor both generate a large amount of heat. If this heat cannot be dissipated in time, the temperature of the power components will rise sharply, which will not only reduce energy conversion efficiency but may also cause serious malfunctions such as battery thermal runaway and motor insulation aging, and even induce safety accidents such as fires and vehicle loss of control.

[0003] As a core component of the thermal management system in new energy vehicles, the cooling piping system plays a crucial role in transferring heat generated by core components such as the power battery and drive motor to the cooling medium. Its structural integrity and sealing reliability are prerequisites for ensuring the normal operation of the thermal management system. However, the working environment of the cooling piping in new energy vehicles is significantly complex and demanding. On the one hand, the pipeline needs to withstand the pressure of the cooling medium for a long time, and on the other hand, the pipeline will be subjected to periodic vibration loads due to the acceleration, deceleration, turning and other conditions during vehicle operation. On the other hand, the start-stop and operation of the drive motor will trigger periodic thermal cycles, causing the pipeline materials to repeatedly undergo thermal expansion and contraction. At the same time, key parts such as pipeline weld joints are prone to micro-cracks due to problems such as residual welding stress and uneven cooling. In addition, factors such as the impact of sand and gravel during vehicle operation and the friction between pipelines and surrounding components will also exacerbate the aging and damage of pipelines.

[0004] Practice has shown that the failure of cooling pipes in new energy vehicles often stems from the dynamic expansion of early, minute defects. These minute defects (such as weld microcracks, micropores within the material, and localized thinning of the pipe wall) may be difficult to detect at the time of product delivery due to their extremely small size. However, under the combined effects of long-term dynamic pressure loads, vibration loads, and thermal cycling loads, they gradually expand and evolve, eventually leading to pipe rupture and coolant leakage. Coolant leakage not only causes the thermal management system to fail and triggers overheating of core power components, but may also induce secondary faults such as short circuits due to coolant seeping into surrounding electrical components, seriously threatening the safety of the entire vehicle. Therefore, comprehensive and accurate defect detection of cooling pipes in new energy vehicles, especially identifying potential early, minute defects and assessing their dynamic failure risk, is a crucial step in ensuring the safe operation of new energy vehicles and improving product reliability.

[0005] Currently, non-destructive testing (NDT) of cooling pipes in new energy vehicles primarily employs traditional static testing techniques, with X-ray NDT being a widely used method. Traditional X-ray NDT involves applying static X-ray irradiation to the pipes to obtain static images, which are then used to analyze the pipes' geometric dimensional accuracy and macroscopic defects (such as obvious cracks, holes, and inclusions). This technology has certain advantages in detecting static geometric deviations and macroscopic defects in pipes and has been widely used in fields such as mechanical manufacturing and automotive parts inspection. However, for the testing needs of cooling pipes in new energy vehicles, traditional X-ray NDT has significant limitations, failing to meet the requirements for accurate detection of early-stage minute defects and dynamic failure risk assessment. Specific limitations are as follows: First, the testing conditions do not match the actual working conditions, making it impossible to detect potential minor defects.

[0006] Traditional X-ray inspection is performed under static, no-load, or constant-load conditions. However, cooling pipes are subjected to a combination of dynamic pressure loads, vibration loads, and thermal cycling loads during actual vehicle operation. Extensive engineering practice shows that many early-stage micro-defects are closed or stable under static no-load conditions. Their size is extremely small, and their grayscale difference from the surrounding material is insignificant, making them difficult to identify in static X-ray images. However, under dynamic loads and thermal cycling, these micro-defects undergo dynamic evolution, such as opening and expansion, becoming the core hidden danger leading to pipe failure. Traditional static inspection methods, because they do not simulate the dynamic load environment of real-world operating conditions, cannot stimulate the dynamic characteristics of these potential micro-defects, thus easily leading to missed detections.

[0007] Second, the detection targets are limited to static defects, and dynamic failure risks cannot be assessed.

[0008] The core objective of traditional X-ray inspection is to identify existing static geometric and macroscopic defects in a product, enabling only "post-inspection"—detecting defects that have already formed and have reached a certain size. However, the safety requirements for cooling pipes in new energy vehicles go beyond simply identifying current defects; they also require predicting the evolution and failure risks of early-stage, minute defects under long-term dynamic operating conditions, achieving "early warning." Traditional static inspection techniques cannot capture the dynamic evolution of defects, nor can they establish a correlation between defects and dynamic loads. Therefore, they cannot assess the dynamic failure risk of pipes under real operating conditions, making it difficult to fundamentally guarantee the long-term operational reliability of the pipes.

[0009] Third, the defect identification sensitivity is insufficient, making it difficult to detect early, minute defects.

[0010] Early micro-defects in the cooling pipes of new energy vehicles (such as weld microcracks less than 0.1 mm wide and internal micropores less than 0.2 mm in diameter) are extremely small. Traditional X-ray inspection technology has limited image resolution and contrast, and the grayscale difference between micro-defects and surrounding materials is not significant in static images, resulting in low sensitivity for identifying early micro-defects. Even with high-resolution X-ray equipment, it is difficult to effectively distinguish micro-defects from interference signals such as material texture and noise under static, unloaded conditions, thus failing to achieve accurate detection of early micro-defects.

[0011] Besides the aforementioned X-ray inspection technology, other existing cooling pipe inspection technologies also have their own limitations: for example, ultrasonic inspection technology is easily affected by the pipe material and shape, resulting in low accuracy in detecting minute defects inside the pipe; pressure leak detection technology can only detect leaks that have already penetrated, and cannot identify early, minute defects that have not yet penetrated; infrared thermal imaging inspection technology mainly targets abnormal heat conduction and has limited ability to identify non-thermally conductive defects inside the pipe. Therefore, existing inspection technologies cannot effectively solve the problem of accurate detection of early, minute defects in the cooling pipes of new energy vehicles and dynamic failure risk assessment.

[0012] In summary, existing non-destructive testing technologies for cooling pipes in new energy vehicles generally suffer from problems such as mismatch between testing conditions and actual operating conditions, inability to elicit potential minor defects, low defect identification sensitivity, and difficulty in assessing dynamic failure risks. These shortcomings fail to meet the high reliability testing requirements of new energy vehicles for cooling pipes. Therefore, developing a cooling pipe testing technology capable of simulating real dynamic operating conditions, accurately identifying early minor defects, and assessing their dynamic failure risks has become a pressing technical challenge for the new energy vehicle industry, possessing significant engineering application value and practical implications. Summary of the Invention

[0013] To address the above problems, this invention provides a method for testing cooling pipes in new energy vehicles that can simulate real dynamic working conditions, accurately identify early minor defects, and assess their dynamic failure risk.

[0014] The technical solution of this invention is: A detection method for a dynamic defect detection system for cooling pipes in new energy vehicles includes the following steps: Step 1, Applying dynamic load: A designed dynamic pressure load spectrum is applied to the tested cooling pipeline; the dynamic pressure load spectrum simulates pressure fluctuations under real working conditions, employing a curved wave cyclic loading method, including a continuous cyclic process of pressurization, depressurization, and repressurization; the pressure P of the dynamic pressure load changes with time t according to a continuous curve relationship, expressed as: P(t) = P0 + Asin(2π) ft+φ); Where P0 is the average pressure, A is the pressure amplitude, f is the cycle frequency, and φ is the initial phase; the pressure is smoothly oscillated between P0-A and P0+A to simulate the periodic pressure changes in the pipeline during actual operation. S2, Dynamic Imaging Acquisition: While applying the dynamic load in step S1, a dynamic X-ray microscope is used to perform non-destructive, in-situ, and continuous imaging of the cooling pipe under test. The imaging system continuously acquires X-ray images penetrating the cooling pipe under test at a rate of tens to hundreds of frames per second, generating a dynamic image sequence that reflects the internal structure of the pipe as pressure changes. S3, Intelligent Image Analysis: Intelligent image analysis is performed on the dynamic X-ray image sequence obtained in step S2 to extract the defect motion features that are invisible in the static image and sensitive to dynamic stress. The grayscale changes, deformation patterns or motion trajectories of the defect area on the time axis are identified and analyzed by image algorithms, so as to realize the transformation from "static defect detection" to "dynamic failure risk prediction".

[0015] Specifically, the intelligent image analysis in step S3 includes a spatiotemporal feature extraction step: The displacement and strain fields of the defect region are calculated using at least one of the optical flow method and digital image correlation (DIC) method, while the dynamic change region is highlighted by the time difference method.

[0016] Specifically, the intelligent image analysis in step S3 includes the defect motion fingerprint construction step: Extract the temporal grayscale curve, deformation rate, and vibration mode features of the defect region to construct a feature vector that reflects the dynamic behavior of the defect.

[0017] Specifically, the intelligent image analysis in step S3 includes classification and recognition model steps: At least one of support vector machine, random forest, and convolutional neural network is used to automatically classify defect types, and the evolution trend of defects is analyzed by combining a time series model; the time series model is an LSTM model.

[0018] Specifically, the intelligent image analysis in step S3 includes anomaly detection and early warning steps: Abnormal motion patterns are identified using an autoencoder or a single-class SVM, and the defect propagation life is predicted based on a fatigue damage model.

[0019] Specifically, the imaging rate of the high-resolution dynamic X-ray microscope described in step S2 is 50-500 frames per second.

[0020] Specifically, the cyclic frequency f of the dynamic pressure load spectrum is set according to the actual operating conditions of the cooling pipeline of the new energy vehicle, and the value range is 0.1-10Hz.

[0021] Specifically, in step S1, when applying dynamic pressure load, the load application data is collected in real time by a pressure sensor, and the load application parameters are adjusted accordingly to ensure the accuracy of the dynamic pressure load spectrum.

[0022] A dynamic defect detection system for cooling pipes in new energy vehicles includes: Dynamic load application module: used to apply the dynamic pressure load spectrum as described in claim 1 to the tested cooling pipeline; Dynamic imaging acquisition module: Employs a high-resolution dynamic X-ray microscope to perform non-destructive, in-situ, and continuous imaging of the tested cooling pipeline during dynamic load application, generating dynamic image sequences; Intelligent analysis module: used to perform intelligent image analysis on dynamic image sequences as described in claim 1 or any one of claims 2-5, to achieve dynamic feature extraction of defects and prediction of failure risks.

[0023] Specifically, it also includes a data feedback module, which includes a pressure sensor for real-time acquisition of dynamic load application data and feedback to the dynamic load application module to achieve closed-loop adjustment of load application parameters.

[0024] This invention, through its core technical solution of "dynamic load application - dynamic imaging acquisition - intelligent image analysis," achieves multi-dimensional technological breakthroughs compared to traditional static detection techniques. Specific technical effects are as follows: 1. Breaking through the limitations of static detection, it accurately stimulates and identifies early-stage minute defects.

[0025] This invention applies a dynamic pressure load spectrum simulating real-world working conditions (cycle frequency 0.1-10Hz, pressure smooth oscillation) to actively excite minute defects that are closed or hidden in a static state, solving the problem of traditional static inspection's difficulty in detecting early-stage defects. Combined with in-situ continuous imaging using a high-resolution dynamic X-ray microscope (imaging rate 50-500 frames / second), it can completely capture the dynamic response process of defects as the load changes, avoiding the risk of missed detections from a single static image and significantly improving the sensitivity and comprehensiveness of defect identification.

[0026] 2. Achieve a leap from "static defect detection" to "dynamic failure risk prediction".

[0027] By leveraging the synergistic effects of defect motion fingerprint construction (extracting features such as temporal grayscale curves and deformation rates), classification and recognition models (combined with LSTM temporal models), and anomaly warning mechanisms (based on fatigue damage models), this invention can not only accurately identify defect types but also quantitatively predict defect propagation lifespan by analyzing defect evolution trends, thus enabling early prediction of cooling pipeline failure risks. Compared to traditional "post-event detection," this "early warning" mode effectively avoids sudden leaks, ruptures, and other failures in cooling pipelines during operation, providing proactive protection for the thermal management safety of core power components in new energy vehicles. Attached Figure Description

[0028] Figure 1 This is a schematic diagram of the process of this invention. Detailed Implementation

[0029] Embodiments of the present invention are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0030] The following is for reference. Figure 1 Described according to the present invention; A detection method for a dynamic defect detection system for cooling pipes in new energy vehicles includes the following steps: Step 1, Applying dynamic load: A designed dynamic pressure load spectrum is applied to the tested cooling pipeline; the dynamic pressure load spectrum simulates pressure fluctuations under real working conditions, employing a curved wave cyclic loading method, including a continuous cyclic process of pressurization, depressurization, and repressurization; the pressure P of the dynamic pressure load changes with time t according to a continuous curve relationship, expressed as: P(t) = P0 + Asin(2π) f t+φ); Where P0 is the average pressure, A is the pressure amplitude, f is the cycle frequency, and φ is the initial phase; the pressure is smoothly oscillated between P0-A and P0+A to simulate the periodic pressure changes in the pipeline during actual operation. The cyclic frequency f of the dynamic pressure load spectrum is set according to the actual operating conditions of the cooling pipes of new energy vehicles, and the value range is 0.1-10Hz.

[0031] When applying dynamic pressure load, the load application data is collected in real time by pressure sensor, and the load application parameters are adjusted accordingly to ensure the accuracy of the dynamic pressure load spectrum.

[0032] To achieve precise closed-loop control of the dynamic pressure load spectrum, the pressure sensor sampling frequency is typically set to 100Hz or higher to satisfy the Nyquist sampling theorem. The response time of the feedback control system is required to be less than 50ms to ensure rapid suppression of load deviations. When the real-time monitored load parameters (such as amplitude and frequency) deviate from the set threshold, the system uses a proportional-integral (PI) controller to dynamically correct the drive command. Its gain parameter is tuned according to the transfer function identified by the system, thereby achieving high-fidelity tracking of the load spectrum amplitude-frequency characteristics.

[0033] S2, Dynamic Imaging Acquisition: While applying the dynamic load in step S1, a high-resolution dynamic X-ray microscope is used to perform non-destructive, in-situ, and continuous imaging of the cooling pipe under test. The imaging system continuously acquires X-ray images penetrating the cooling pipe under test at a rate of tens to hundreds of frames per second, generating a dynamic image sequence that reflects the internal structure of the pipe as pressure changes. The imaging rate of the high-resolution dynamic X-ray microscope is 50-500 frames per second.

[0034] S3, Intelligent Image Analysis: Intelligent image analysis is performed on the dynamic X-ray image sequence obtained in step S2 to extract the defect motion features that are invisible in the static image and sensitive to dynamic stress. The grayscale changes, deformation patterns or motion trajectories of the defect area on the time axis are identified and analyzed by image algorithms, so as to realize the transformation from "static defect detection" to "dynamic failure risk prediction".

[0035] Intelligent image analysis includes spatiotemporal feature extraction steps: The displacement and strain fields of the defect region are calculated using at least one of the optical flow method and digital image correlation (DIC) method, while the dynamic change region is highlighted by the time difference method.

[0036] Intelligent image analysis includes the defect motion fingerprint construction step: Extract the temporal grayscale curve, deformation rate, and vibration mode features of the defect region to construct a feature vector that reflects the dynamic behavior of the defect.

[0037] The extraction of temporal grayscale curves uses a sliding window based on consecutive frames, typically fixing 10 frames as an analysis unit to balance temporal resolution, noise suppression, and signal stability.

[0038] The deformation rate can be calculated with sub-micron precision using digital image correlation and sub-pixel displacement algorithms, with a typical value on the order of 0.01 μm / s.

[0039] The vibration mode characteristic parameters are extracted by time-frequency analysis to obtain the dominant frequency, harmonic components and their corresponding amplitudes, which together constitute a high-dimensional characteristic vector characterizing the dynamic evolution of defects. The sampling theorem guarantees its effectiveness.

[0040] Intelligent image analysis includes classification and recognition model steps: At least one of support vector machine, random forest, and convolutional neural network is used to automatically classify defect types, and the evolution trend of defects is analyzed by combining a time series model; the time series model is an LSTM model.

[0041] In the defect classification and evolution analysis framework, LSTM models often adopt a stacked structure, typically with 2 to 3 hidden layers, each containing 128 or 256 neurons, and are trained iteratively for 50 to 100 epochs to prevent overfitting.

[0042] The choice of kernel function for support vector machines is based on the linear separability of the feature space. Linear kernels are suitable for high-dimensional features, Gaussian kernels can be adapted to nonlinear classification by parameter tuning, while polynomial kernels are suitable for decision boundaries with moderate complexity.

[0043] Random forests set the subtree size (e.g., 100-500 trees) based on the number of features and optimize the depth and feature subset size through out-of-bag error estimation to balance generalization ability and computational efficiency.

[0044] Intelligent image analysis includes anomaly detection and early warning steps: Abnormal motion patterns are identified using an autoencoder or a single-class SVM, and the defect propagation life is predicted based on a fatigue damage model.

[0045] Abnormal motion patterns can be determined by setting a deformation rate threshold (such as an upper limit of 0.1 μm / s based on statistical process control) or by using a confidence interval based on the autoencoder reconstruction error.

[0046] Fatigue damage models typically employ the Miner linear cumulative damage criterion. Key parameters (such as the stress-life curve exponent) are calibrated using material standard test data and corrected online using maximum likelihood estimation or Bayesian update methods in conjunction with the actual load spectrum.

[0047] A dynamic defect detection system for cooling pipes in new energy vehicles includes: A dynamic load application module is used to apply the dynamic pressure load spectrum as described in claim 1 to the tested cooling pipeline; The dynamic imaging acquisition module uses a high-resolution dynamic X-ray microscope to perform non-destructive, in-situ, and continuous imaging of the cooling pipeline under test during the application of dynamic loads, generating a dynamic image sequence. The intelligent analysis module is used to perform intelligent image analysis on dynamic image sequences as described in any one of claims 1 or 2-5, thereby realizing defect dynamic feature extraction and failure risk prediction.

[0048] The data feedback module includes a pressure sensor, which is used to collect dynamic load application data in real time and feed it back to the dynamic load application module to realize closed-loop adjustment of load application parameters.

[0049] This invention first applies a "dynamic pressurization → depressurization → increased curved wave cycle" load spectrum to the tested cooling pipe to actively excite and accelerate the dynamic evolution of its internal potential defects. Second, it uses a high-resolution dynamic X-ray microscope to acquire non-destructive, in-situ, and continuous image data of this stress evolution process. Finally, it performs intelligent analysis on the acquired dynamic image sequence to identify features that are sensitive to dynamic stress but cannot be observed in a single static image. This achieves a shift from detecting static defects to predicting dynamic failure risks, significantly improving the sensitivity of defect identification.

[0050] Regarding the information disclosed in this case, the following points need to be clarified: (1) The accompanying drawings of the embodiments disclosed in this case only involve the structures involved in the embodiments disclosed in this case. Other structures can refer to the general design. (2) Where there is no conflict, the embodiments and features disclosed in this case can be combined with each other to obtain new embodiments; The above are merely specific embodiments disclosed in this case, but the scope of protection of this disclosure is not limited thereto. The scope of protection disclosed in this case shall be determined by the scope of protection of the claims.

Claims

1. A detection method for a dynamic defect detection system for cooling pipes in new energy vehicles, characterized in that, Includes the following steps: Step 1, Applying dynamic load: A designed dynamic pressure load spectrum is applied to the tested cooling pipeline; the dynamic pressure load spectrum simulates pressure fluctuations under real working conditions, employing a curved wave cyclic loading method, including a continuous cyclic process of pressurization, depressurization, and repressurization; the pressure P of the dynamic pressure load changes with time t according to a continuous curve relationship, expressed as: P(t)=P0+Asin(2π f t+φ); Where P0 is the average pressure, A is the pressure amplitude, f is the cycle frequency, and φ is the initial phase; To achieve smooth pressure oscillation between P0 - A and P0 + A, simulating the periodic pressure changes in the pipeline during actual operation; S2, Dynamic Imaging Acquisition: While applying the dynamic load in step S1, a dynamic X-ray microscope is used to perform non-destructive, in-situ, and continuous imaging of the cooling pipe under test. The imaging system continuously acquires X-ray images penetrating the cooling pipe under test at a rate of tens to hundreds of frames per second, generating a dynamic image sequence that reflects the internal structure of the pipe as pressure changes. S3, Intelligent Image Analysis: Intelligent image analysis is performed on the dynamic X-ray image sequence obtained in step S2 to extract the defect motion features that are invisible in the static image and sensitive to dynamic stress. The grayscale changes, deformation patterns or motion trajectories of the defect area on the time axis are identified and analyzed by image algorithms, so as to realize the transformation from "static defect detection" to "dynamic failure risk prediction".

2. The detection method of the dynamic defect detection system for cooling pipes in new energy vehicles according to claim 1, characterized in that, Step S3, the intelligent image analysis, includes a spatiotemporal feature extraction step: The displacement and strain fields of the defect region are calculated using at least one of the optical flow method and digital image correlation method, while the dynamic change region is highlighted by the time-series difference method.

3. The detection method of the dynamic defect detection system for cooling pipes in new energy vehicles according to claim 1, characterized in that, Step S3, the intelligent image analysis, includes the defect motion fingerprint construction step: Extract the temporal grayscale curve, deformation rate, and vibration mode features of the defect region to construct a feature vector that reflects the dynamic behavior of the defect.

4. The detection method of the dynamic defect detection system for cooling pipes in new energy vehicles according to claim 1, characterized in that, Step S3, the intelligent image analysis, includes classification and recognition model steps: At least one of support vector machine, random forest, and convolutional neural network is used to automatically classify defect types, and the evolution trend of defects is analyzed by combining a time series model; the time series model is an LSTM model.

5. The detection method of the dynamic defect detection system for cooling pipes in new energy vehicles according to claim 1, characterized in that, Step S3, the intelligent image analysis, includes anomaly detection and early warning steps: Abnormal motion patterns are identified using an autoencoder or a single-class SVM, and the defect propagation life is predicted based on a fatigue damage model.

6. The detection method of the dynamic defect detection system for cooling pipes in new energy vehicles according to claim 1, characterized in that, The imaging rate of the high-resolution dynamic X-ray microscope described in step S2 is 50-500 frames per second.

7. The detection method of the dynamic defect detection system for cooling pipes in new energy vehicles according to claim 1, characterized in that, The cyclic frequency f of the dynamic pressure load spectrum is set according to the actual operating conditions of the cooling pipeline of new energy vehicles, and the value range is 0.1-10Hz.

8. The detection method of the dynamic defect detection system for cooling pipes in new energy vehicles according to claim 1, characterized in that, When applying dynamic pressure load in step S1, the load application data is collected in real time by pressure sensor, and the load application parameters are adjusted accordingly to ensure the accuracy of dynamic pressure load spectrum.

9. A dynamic defect detection system for cooling pipes in new energy vehicles, characterized in that, include: Dynamic load application module: used to apply the dynamic pressure load spectrum as described in claim 1 to the tested cooling pipeline; Dynamic imaging acquisition module: Employs a high-resolution dynamic X-ray microscope to perform non-destructive, in-situ, and continuous imaging of the tested cooling pipeline during dynamic load application, generating dynamic image sequences; Intelligent analysis module: used to perform intelligent image analysis on dynamic image sequences as described in claim 1 or any one of claims 2-5, to achieve dynamic feature extraction of defects and prediction of failure risks.

10. The dynamic defect detection system for cooling pipes in new energy vehicles according to claim 9, characterized in that, It also includes a data feedback module, which includes a pressure sensor for real-time acquisition of dynamic load application data and feedback to the dynamic load application module to achieve closed-loop adjustment of load application parameters.

Images