A method and system for testing the safety performance of brake discs

Abstract

This invention relates to the field of image processing technology, and in particular to a method and system for detecting the safety performance of brake discs. When detecting the safety performance of brake discs, this invention performs comprehensive inspections, including appearance, size, and material properties. Simultaneously, it employs an improved adaptive filtering algorithm to denoise the image of the brake disc to be detected. The image is decomposed into multiple scales, resulting in sub-images of various scales. Local features are extracted from each sub-image, and weights are dynamically assigned based on these local features. Adaptive filtering is then applied to each scale of the sub-image based on these weights to improve the filtering effect. The filtering intensity is dynamically adjusted based on local variance, gradient strength, and texture information, which better preserves image edges and details and avoids excessive smoothing. By combining global and local information, adaptive weights are assigned to each pixel to optimize the filtering effect and improve the algorithm's adaptability and robustness.

Description

Technical Field

[0001] This invention relates to the field of image processing technology, and in particular to a method and system for testing the safety performance of brake discs. Background Technology

[0002] Brake discs are one of the core components of a car's braking system. Their main function is to reduce the rotational speed of the wheels through friction with the brake pads, thereby slowing down or stopping the vehicle. The performance of the brake discs directly affects the braking effect and driving safety. After brake discs are manufactured, problems such as scratches, pinholes, and shrinkage can easily occur. If these problems are not detected and addressed in a timely manner, they may lead to brake failure, which could result in serious traffic accidents. Therefore, there is an urgent need for a method and system that can efficiently, accurately, and comprehensively test the safety performance of brake discs.

[0003] Regarding solutions for brake disc testing, Chinese invention patent application CN113819878A provides a brake disc performance testing device and method, which mounts the brake disc on a rotating structure; a synchronous drive structure moves the flatness measuring structure and the braking structure to make contact with the side wall of the brake disc; the rotating structure drives the brake disc to rotate, and after the brake disc rotates, the flatness measuring structure performs flatness measurement on the brake disc; however, the above solution only tests the flatness of the brake disc, which has the problem of incomplete testing.

[0004] Meanwhile, in the existing technology, when using brake disc images to detect brake discs, the preprocessing or image filtering of the brake disc images is a challenge. Since brake disc defects are relatively few in the image, existing adaptive filtering algorithms (such as adaptive Gaussian filtering) can achieve the detection target of brake discs to a certain extent; however, they still have shortcomings in some cases, such as insufficient detail preservation. While removing noise, important details in the image may be blurred, especially tiny pinholes and scratches. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention provides a method and system for testing the safety performance of brake discs, thereby resolving the issues existing in the prior art.

[0006] This invention provides a method for testing the safety performance of brake discs, comprising the following steps:

[0007] S1: Perform visual inspection on the brake disc to be inspected;

[0008] Step S1 includes:

[0009] S1.1: Use a camera to acquire an image of the brake disc to be inspected;

[0010] S1.2: Perform image preprocessing on the brake disc image to be detected; use an improved adaptive filtering algorithm to denoise the brake disc image to be detected; specifically including:

[0011] Sa: Decompose the brake disc image to be detected into sub-images of multiple scales; Sb: Extract local features from each sub-image, the local features including local variance, gradient intensity, and texture information; Sc: Dynamically assign weights based on the local features; Sd: Adaptively filter each sub-image according to the weights; Se: Reconstruct the adaptively filtered sub-images at multiple scales to obtain the final filtered image;

[0012] S1.3: Extract the scratch features, pinhole features, and shrinkage features of the brake disc to be tested from the preprocessed image of the brake disc to be tested;

[0013] S1.4: Input the scratch features, pinhole features and shrinkage features of the brake disc to be tested into the image analysis model to obtain the degree of scratches, pinhole conditions and shrinkage conditions of the brake disc to be tested;

[0014] S1.5: The appearance performance of the brake disc under test is comprehensively evaluated based on the degree of scratches, pinholes, and shrinkage.

[0015] S2: Perform dimensional inspection on the brake disc to be inspected.

[0016] S3: Perform material performance testing on the brake disc to be tested.

[0017] Preferably, in step Sa, the brake disc image to be detected is decomposed into sub-images of multiple scales, and each sub-image corresponds to image information of different frequencies;

[0018] Step Sa is expressed by the formula:

[0019] ;

[0020] Where I is the image of the brake disc to be detected, I s Let S represent the sub-image at the s-th scale, where S is the total number of scales.

[0021] Preferably, in step Sb, the specific formula is as follows:

[0022] ;

[0023] ;

[0024] ;

[0025] In the formula, (x, y) represents the pixel coordinates in the image; Window represents a local window region centered at pixel (x, y), used to calculate local statistics; N represents the number of pixels in the local window Window; μ s Represents a sub-image I in a local window (Window). s The average value; Var(I s ) represents the sub-image I in the local window Window. s Local variance; x I s Sub-image I s Gradient in the horizontal direction; y I s Sub-image I s Gradient in the vertical direction; Grad(I) s ) represents sub-image I s gradient strength; Texture(I s ) represents the sub-image I in the local window Window. s Texture information; I s (x, y) represents the sub-image I s The pixels in.

[0026] Preferably, in step Sc, the formula for dynamically allocating weights ω(x, y) is:

[0027] ;

[0028] Where α(x,y), β(x,y) and γ(x,y) are weighting coefficients used to balance the influence of different features; ε is a constant.

[0029] Preferably, the method for determining the weighting coefficients α(x,y), β(x,y), and γ(x,y) specifically includes:

[0030] Calculate the global noise level (Global Var) of the image of the brake disc to be detected;

[0031] Calculate the global texture complexity (Global Texture) of the brake disc image to be detected;

[0032] Calculate the local noise level Local Var(x,y) and the local texture complexity Local Texture(x,y) for each pixel (x,y);

[0033] The weighting coefficients α(x,y), β(x,y) and γ(x,y) are determined based on the global noise level, global texture complexity, local noise level of each pixel, and local texture complexity of each pixel.

[0034] Preferably, the specific formulas for determining the weight coefficients α(x,y), β(x,y), and γ(x,y) based on the global noise level, global texture complexity, local noise level of each pixel, and local texture complexity of each pixel are as follows:

[0035] ;

[0036] ;

[0037] ;

[0038] In the formula, It is a constant.

[0039] Preferably, step S1.1 includes:

[0040] S1.1.1: Install the camera at the monitoring station to ensure that the camera can capture images of the brake disc to be inspected from multiple angles;

[0041] S1.1.2: Install an adjustable light source around the camera;

[0042] S1.1.3: Obtain the image of the brake disc to be detected.

[0043] Preferably, in step S1.1.3, acquiring the image of the brake disc to be detected includes:

[0044] An image of the surface of the brake disc to be tested is taken from the front.

[0045] A side image of the brake disc to be tested is taken from the side of the disc.

[0046] The images were taken from the top, bottom, and oblique angles of the brake disc under test.

[0047] Preferably, in step S1.2, the image preprocessing further includes grayscale processing.

[0048] According to another aspect of the present invention, a brake disc safety performance testing system is provided, wherein the brake disc safety performance testing system employs the above-described brake disc safety performance testing method, and the brake disc safety performance testing system comprises:

[0049] The appearance inspection module is used to perform appearance inspection on the brake disc to be inspected.

[0050] A size detection module is used to detect the size of the brake disc to be tested.

[0051] The material performance testing module is used to perform material performance testing on the brake disc to be tested.

[0052] Compared with the prior art, the present invention has the following technical effects:

[0053] When testing the safety performance of brake discs, the brake discs undergo comprehensive inspection, including appearance, size, and material properties. Simultaneously, this invention employs an improved adaptive filtering algorithm to denoise the brake disc image under test. The image is decomposed into sub-images at multiple scales. Local features are extracted from each sub-image, including local variance, gradient intensity, and texture information. Weights are dynamically assigned based on these local features. Adaptive filtering is then applied to each sub-image at each scale based on these weights. By analyzing and processing the image at multiple scales, both high-frequency noise (such as random noise) and low-frequency noise (such as uneven lighting) can be removed simultaneously, improving the filtering effect. Dynamically adjusting the filtering intensity based on local variance, gradient intensity, and texture information better preserves image edges and details, avoiding excessive smoothing. Combining global and local information, adaptive weights are assigned to each pixel to optimize the filtering effect and improve the algorithm's adaptability and robustness.

[0054] Meanwhile, this invention proposes a method for determining weight coefficients in the adaptive filtering process, which dynamically adjusts the weights based on the local and global features of the image, thus better adapting to different types of images and noise distributions. Furthermore, by combining global and local features, it can better balance noise removal, detail preservation, and texture processing. Attached Figure Description

[0055] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0056] Figure 1 This is a flowchart of a brake disc safety performance testing method provided in an embodiment of the present invention;

[0057] Figure 2 This is a flowchart of the appearance inspection of the brake disc to be inspected provided in an embodiment of the present invention;

[0058] Figure 3 This is a flowchart of a process for denoising the brake disc image to be detected using an improved adaptive filtering algorithm, provided in an embodiment of the present invention. Detailed Implementation

[0059] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below. Obviously, the described embodiments are only a part of the embodiments of this invention, and not all of them. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0060] Example 1

[0061] like Figure 1 As shown, a method for testing the safety performance of a brake disc includes the following steps:

[0062] S1: Perform visual inspection on the brake disc to be inspected;

[0063] Among them, such as Figure 2 As shown, step S1 specifically includes:

[0064] S1.1: Use a camera to acquire an image of the brake disc to be inspected;

[0065] Image acquisition is the first step in brake disc safety performance testing. Its purpose is to obtain high-quality brake disc images, providing a reliable data foundation for subsequent image preprocessing, feature extraction, and safety performance assessment. High-quality brake disc images can more clearly reflect surface scratches, pinholes, shrinkage, and other features of the brake disc, thereby improving the accuracy and reliability of the inspection.

[0066] To ensure that the acquired images meet the inspection requirements, a high-resolution industrial camera is used as the image acquisition device in this step. The preferred high-resolution industrial camera is the SVS-Vistek SHR series camera, with a resolution of 25 megapixels (5120×5120). It uses a CMOS sensor, which has advantages such as high sensitivity, low power consumption, and fast readout. Furthermore, the frame rate of this high-resolution industrial camera is 23fps (12 megapixels), which is suitable for dynamic scenes and rapid continuous shooting.

[0067] Furthermore, step S1.1 specifically includes:

[0068] S1.1.1: Install the camera at the monitoring station to ensure that the camera can capture images of the brake disc under inspection from multiple angles;

[0069] The position and angle of the camera should be adjusted according to the size and shape of the brake disc to ensure that the captured image covers the entire surface and sides of the brake disc.

[0070] S1.1.2: Install an adjustable light source around the camera;

[0071] The adjustable light source ensures that light shines evenly onto the surface of the brake disc. The intensity and angle of the adjustable light source should be adjusted according to the material and surface reflectivity of the brake disc to avoid reflected light interfering with image acquisition.

[0072] Before officially acquiring images, the camera and adjustable light source are tested to ensure that the camera's autofocus and lighting adjustment functions are working properly. A small number of test images are taken to check the image sharpness, contrast, and brightness. The camera parameters and light source settings are adjusted until satisfactory images are obtained.

[0073] S1.1.3: Acquire the image of the brake disc to be inspected;

[0074] Specifically, step S1.1.3 includes:

[0075] Take a surface image of the brake disc to be inspected from the front. The camera should be adjusted to the optimal focal length to ensure that the image clearly reflects scratches, pinholes and other defects on the surface of the brake disc. At the same time, in order to improve the comprehensiveness of the inspection, take surface images from multiple angles, such as taking a picture from the center of the brake disc to be inspected outwards, and taking pictures of the edge area from different directions.

[0076] Take a side image from the side of the brake disc to be tested. This side image is mainly used to detect the shrinkage of the brake disc to be tested, such as uneven thickness, bending, etc. The camera should be adjusted to a suitable height and angle to ensure that the side profile of the brake disc to be tested can be clearly captured.

[0077] To more comprehensively evaluate the safety performance of the brake disc, step S1.1.3 also includes taking pictures of the brake disc from multiple angles. For example, pictures can be taken from the top, bottom, sides, and oblique angles of the brake disc under test to obtain image information from more angles. Multi-angle shooting can effectively avoid the omission of features caused by shooting from a single angle and improve the accuracy of detection.

[0078] S1.2: Perform image preprocessing on the image of the brake disc to be inspected;

[0079] The purpose of this image preprocessing is to perform preliminary processing on the acquired images to improve image quality, enhance key features, and provide a better data foundation for subsequent feature extraction and analysis. In this step, image preprocessing mainly includes grayscale conversion and noise reduction.

[0080] Grayscale conversion is the process of converting a color image into a grayscale image. A color image typically contains three color channels (red, green, and blue), each with its own pixel value. Grayscale conversion combines the information from the three color channels into a single grayscale channel, where the grayscale value of each pixel represents its brightness.

[0081] In this step, a weighted average method is used to convert the image of the brake disc to grayscale.

[0082] Specifically, this weighted average method assigns different weights to the three color channels (red, green, and blue) based on the human eye's sensitivity to different colors, and calculates the grayscale value (Gray); the calculation formula is as follows:

[0083] ;

[0084] In the formula, R, G, and B are the pixel values ​​of the red, green, and blue channels, respectively.

[0085] This step employs an improved adaptive filtering algorithm to denoise the brake disc image to be detected. In brake disc detection, the purpose of image preprocessing is to remove noise while preserving key features (such as pinholes and scratches). While existing adaptive filtering algorithms (such as adaptive Gaussian filtering) can achieve this goal to some extent, they still have shortcomings in certain situations; for example, insufficient detail preservation may blur important details in the image (especially tiny pinholes and scratches) while removing noise. Therefore, this embodiment proposes an improved adaptive filtering algorithm that combines multiple filtering strategies and optimization mechanisms to improve the filtering effect; specifically, such as... Figure 3 As shown, the specific steps for denoising the brake disc image to be detected using the improved adaptive filtering algorithm include:

[0086] Sa: Decompose the brake disc image to be detected into sub-images of multiple scales.

[0087] In this process, the brake disc image I to be detected is decomposed into sub-images of multiple scales, and each sub-image corresponds to image information of different frequencies.

[0088] In this step, Gaussian pyramid decomposition is used to decompose the brake disc image to be detected into sub-images of multiple scales;

[0089] Specifically, step Sa is expressed by the formula:

[0090] ;

[0091] Where I is the image of the brake disc to be detected, I s Let S represent the sub-image at the s-th scale, where S is the total number of scales.

[0092] Sb: Local feature extraction is performed on each sub-image, which includes local variance, gradient intensity, and texture information;

[0093] Among them, local variance is used to measure noise level, gradient intensity is used to detect edges, and texture information is used to identify complex regions.

[0094] The specific formula is as follows:

[0095]

[0096]

[0097] ;

[0098] In the formula, (x, y) represents the pixel coordinates in the image; Window represents a local window region centered at pixel (x, y), used to calculate local statistics; N represents the number of pixels in the local window Window; μ s Represents a sub-image I in a local window (Window). s The average value; Var(I s ) represents the sub-image I in the local window Window. s Local variance; x I s Sub-image I s Gradient in the horizontal direction; y I s Sub-image I s Gradient in the vertical direction; Grad(I) s ) represents sub-image I s gradient strength; Texture(I s ) represents the sub-image I in the local window Window. s Texture information; I s (x, y) represents the sub-image I s The pixels in.

[0099] Sc: Dynamically assign weights based on the local features;

[0100] Weights are dynamically allocated based on local features to optimize the filtering effect;

[0101] Specifically, the weight allocation strategy is as follows:

[0102] 1. Noise suppression weight: A larger weight is assigned to the high variance region (where there is more noise) to enhance noise suppression;

[0103] 2. Detail-preserving weights: Assign smaller weights to high-gradient regions (edges and details) to avoid over-smoothing;

[0104] 3. Texture weight: Assign appropriate weights to areas with complex textures to balance noise removal and detail preservation.

[0105] Specifically, the formula for dynamically allocating weights ω(x, y) is:

[0106] ;

[0107] Where α(x,y), β(x,y) and γ(x,y) are weighting coefficients used to balance the influence of different features; ε is a constant.

[0108] Furthermore, this embodiment proposes a method for determining the weighting coefficients α(x,y), β(x,y), and γ(x,y), specifically including:

[0109] Calculate the global noise level (Global Var) of the image of the brake disc to be detected;

[0110] The formula for calculating the global noise level is as follows:

[0111] ;

[0112] Where M and N are the width and height of the brake disc image to be detected, μ is the global average value of the brake disc image to be detected, and I(x,y) is the brake disc image to be detected.

[0113] Calculate the global texture complexity (Global Texture) of the brake disc image to be detected;

[0114] The formula for calculating global texture complexity is:

[0115] ;

[0116] Calculate the local noise level Local Var(x,y) and the local texture complexity Local Texture(x,y) for each pixel (x,y); the specific formulas are as follows:

[0117] ;

[0118] ;

[0119] In the formula, I(x+u,y+v) is the pixel value with an offset of (u,v) in a local window centered at (x,y); μlocal is the average value in the local window;

[0120] The weighting coefficients α(x, y), β(x, y), and γ(x, y) are determined based on the global noise level, global texture complexity, local noise level of each pixel, and local texture complexity of each pixel; the specific formula is as follows:

[0121] ;

[0122] ;

[0123] .

[0124] Based on the dynamic adjustment of local and global features of the image, the weight coefficients α, β and γ can better adapt to different types of images and noise distributions; at the same time, by combining global and local features, noise removal, detail preservation and texture processing can be better balanced.

[0125] Sd: Adaptive filtering of sub-images at each scale based on weights;

[0126] Gaussian filtering is used to smooth the sub-images at each scale, while the filter strength is adjusted according to the weights:

[0127] ;

[0128] Among them, I' s (x, y) represents the smoothed sub-image; G(u, v; σ(x, y)·w(x, y)) represents the Gaussian kernel calculated based on the standard deviation adjusted according to adaptive weights. The core of Gaussian filtering is the Gaussian kernel, a discretized version of a two-dimensional Gaussian function. The shape of the Gaussian kernel is determined by the standard deviation. The larger the standard deviation, the more dispersed the distribution of the Gaussian kernel, and vice versa. In image processing, the Gaussian kernel is mainly used for smoothing, blurring, and edge detection. By applying the Gaussian kernel to an image, filtering can be achieved, thereby reducing noise and improving image quality. Simultaneously, the Gaussian kernel can also be used for edge detection, extracting edge information from the image by performing Gaussian filtering on the second-order difference of the image.

[0129] Se: Multi-scale reconstruction is performed on the sub-images at each scale of the adaptive filter to obtain the final filtered image I. filtered ;

[0130] ;

[0131] In this embodiment, by analyzing and processing images at multiple scales, high-frequency noise (such as random noise) and low-frequency noise (such as uneven illumination) can be removed simultaneously, improving the filtering effect. The filtering intensity is dynamically adjusted according to local variance, gradient strength and texture information, which can better preserve the edges and details of the image and avoid over-smoothing. By combining global and local information, adaptive weights are assigned to each pixel to optimize the filtering effect and improve the adaptability and robustness of the algorithm.

[0132] S1.3: Extract the scratch features, pinhole features, and shrinkage features of the brake disc to be tested from the preprocessed image of the brake disc to be tested;

[0133] The scratch feature is used to characterize the size of scratches on the surface of the brake disc to be tested. Scratched areas are typically represented by regions with low grayscale values ​​because scratches reduce surface reflectivity. The scratch feature includes: the average grayscale value of the scratched area; and the area percentage of the scratched area. The pinhole feature is used to quantify the pinhole condition on the surface of the brake disc to be tested. Pinholes are typically represented by regions with low grayscale values. The pinhole feature includes pinhole area, pinhole length, and pinhole number. The shrinkage feature is used to quantify the contour changes of the side image of the brake disc to be tested, reflecting the shrinkage condition of the brake disc. This shrinkage feature includes: shrinkage area, shrinkage length, and shrinkage number.

[0134] S1.4: Input the scratch features, pinhole features, and shrinkage features of the brake disc to be tested into the image analysis model to obtain the degree of scratches, pinhole conditions, and shrinkage conditions of the brake disc to be tested.

[0135] The degree of scratches includes: mild, moderate, and severe; the degree of trachoma includes: none, mild, moderate, and severe; and the degree of shrinkage includes: none, mild, and severe.

[0136] In this step, the image analysis model is a convolutional neural network (CNN). A CNN is a deep learning model particularly well-suited for processing image data. This CNN automatically learns feature representations in images through a combination of convolutional layers, pooling layers, and fully connected layers. Convolutional layers are used to extract local features, pooling layers are used to reduce the spatial dimensionality of features, and fully connected layers are used for classification or regression.

[0137] In brake disc inspection, a convolutional neural network model can automatically identify the degree of scratches, pinholes, and shrinkage on brake discs by learning from a large amount of labeled image data. The model outputs the classification result and confidence score for each feature, providing a basis for subsequent safety performance evaluation.

[0138] Specifically, the model structure of a convolutional neural network is as follows:

[0139] Input layer: The input image size is 224×224×3.

[0140] Convolutional layer: Multiple convolutional layers extract local features of an image.

[0141] Pooling layers: Multiple pooling layers reduce the spatial dimensionality of features.

[0142] Fully connected layers: Multiple fully connected layers are used for classification.

[0143] Output layer: The output layer contains three categories of tasks (scratching degree, pinhole condition, shrinkage condition), and each task has multiple categories.

[0144] Loss function and optimizer: We chose the cross-entropy loss function and the Adam optimizer. During training, the learning rate was adjusted to 0.001 and the batch size to 32.

[0145] Training process: A convolutional neural network model is trained using labeled image data. During training, hyperparameters such as the learning rate and batch size are adjusted to optimize model performance. The training process consists of 50 epochs (one epoch requires 50 iterations), and each epoch includes forward propagation, loss calculation, backpropagation, and parameter update.

[0146] Model Evaluation: The model's performance is evaluated using a validation set. Evaluation metrics include accuracy, recall, precision, and F1 score. Based on the evaluation results, the model is optimized.

[0147] Model Deployment: Features extracted from the image are input into a trained CNN model. The model outputs the classification result and confidence score for each feature. Based on the model's output, the safety performance of the brake disc is comprehensively evaluated.

[0148] S1.5: Conduct a comprehensive evaluation of the appearance performance of the brake disc under test based on the degree of scratches, pinholes, and shrinkage.

[0149] To comprehensively evaluate the appearance performance of the brake disc under test, different features (scratches, pinholes, shrinkage) need to be assigned different weights, and then a comprehensive safety performance score is obtained by weighted summation. The weight allocation should be based on the degree of influence of each feature on the safety performance of the brake disc.

[0150] The weights are set as follows: scratch severity weight ω1=0.4; pinholes weight ω2=0.4; shrinkage weight ω3=0.2.

[0151] The degree of scratches, pinholes, and shrinkage on the brake disc to be tested are quantified into a score; in this step, an expert scoring method is used to determine the score;

[0152] Specifically: minor scratches (1 point), moderate scratches (2 points), severe scratches (3 points); no trachoma (1 point), minor trachoma (2 points), moderate trachoma (3 points), severe trachoma (4 points); no shrinkage (1 point), minor shrinkage (2 points), severe shrinkage (3 points).

[0153] S2: Perform dimensional inspection on the brake disc to be inspected;

[0154] As a core component of the automotive braking system, the dimensional accuracy of the brake disc directly affects braking performance, vehicle stability, and safety. The purpose of dimensional inspection is to ensure that the brake disc's geometry meets design requirements, avoiding problems such as reduced braking performance, brake noise, and brake vibration caused by dimensional deviations.

[0155] Specifically, a laser triangulation sensor array, a turntable station, and a positioning fixture are used to detect the dimensions of the brake disc to be inspected.

[0156] The laser triangulation sensor array employs multiple high-precision laser triangulation sensors (such as the KEYENCELJ-V series) arranged around the brake disc to form a 360° detection loop. The number of laser triangulation sensors is determined based on the required detection accuracy and speed, and is generally no less than six.

[0157] The turntable station is used to place the brake disc to be tested on a rotatable turntable. The turntable is driven by a high-precision servo motor to ensure rotational accuracy and speed stability. The rotation angle of the turntable is fed back by a high-precision rotary encoder, which is used to synchronize the measurement data of the laser sensor.

[0158] The positioning fixture is used to position and fix the brake disc to be tested on the turntable to ensure the stability of the brake disc during the measurement process.

[0159] The laser triangulation sensor emits a laser beam onto the surface of the brake disc to be tested. The laser beam is reflected off the surface of the brake disc and received by the laser triangulation sensor. By measuring the change in the laser reflection angle, the distance between the measured point on the brake disc and the sensor is calculated, thereby realizing the detection of the thickness and flatness of the brake disc.

[0160] S3: Perform material performance testing on the brake disc to be tested;

[0161] The material properties of brake discs directly determine their braking performance and service life under high-temperature and high-load conditions. Even minor changes in material properties can lead to decreased braking performance, accelerated heat fade, fatigue fracture, and other problems, thereby posing safety hazards.

[0162] The process involves using an eddy current heater, an infrared thermal imager, an electromagnetic ultrasonic transducer, and a data acquisition and processing system to perform material performance testing on the brake disc under test.

[0163] Furthermore, the eddy current heater uses high-frequency eddy current heating technology, which can uniformly heat the brake disc to a set temperature (such as 120℃±2℃) in a short time. The power and heating time of the eddy current heater can be adjusted according to the material and size of the brake disc.

[0164] Infrared thermal imagers are used to acquire the temperature field distribution on the surface of brake discs. The resolution and frame rate of infrared thermal imagers need to meet the requirements of high-precision measurement. For example, the FLIR A655sc thermal imager can achieve a resolution of 640×480 pixels and a frame rate of 120Hz.

[0165] Electromagnetic ultrasonic transducers (EMATs) are used to excite and receive ultrasonic signals to measure the acoustic properties of materials. The probes of EMAs excite transverse and longitudinal waves on the material surface through electromagnetic induction, requiring no coupling agent and making them suitable for high-temperature environments.

[0166] The data acquisition and processing system includes a high-speed data acquisition card, a signal processing module, and an industrial PC, which is used to acquire and process measurement data from infrared thermal imagers and electromagnetic ultrasonic transducers in real time.

[0167] After the eddy current heater completes the heating process, an infrared thermal imager acquires a temperature field image of the surface of the brake disc under test, and then calculates the thermal conductivity (λ) and specific heat capacity (c) of the material of the brake disc. The calculated thermal conductivity and specific heat capacity are then compared with standard values ​​to determine whether the material properties meet the requirements. For example, for gray cast iron HT250 brake discs, the standard value for thermal conductivity is 45±3 W / (m·K), and the standard value for specific heat capacity is 0.5±0.05 kJ / (kg·K).

[0168] After the brake disc under test is heated to a set temperature, a probe using an electromagnetic ultrasonic transducer excites transverse and longitudinal waves on the surface of the brake disc. The excitation frequency of the probe is optimized according to the material characteristics, typically within the range of 1-5 MHz. The probe receives the reflected ultrasonic signals and transmits them to the data processing system via a high-speed data acquisition card. The acquired signals include the time-of-flight (TOF) and amplitude of the transverse and longitudinal waves. Then, using the TOF and amplitude data, the Young's modulus (E) and Poisson's ratio (ν) of the brake disc material under test are calculated. The calculated Young's modulus and Poisson's ratio are compared with standard values ​​to test the material properties of the brake disc. For example, for gray cast iron HT250 brake discs, the standard value for Young's modulus is 110±5 GPa, and the standard value for Poisson's ratio is 0.25±0.02. If the detected material performance parameters deviate from the standard range, the material properties of the brake disc under test are deemed unqualified.

[0169] Example 2

[0170] The present invention also provides a brake disc safety performance testing system, which adopts a brake disc safety performance testing method of Embodiment 1. The brake disc safety performance testing system includes:

[0171] The appearance inspection module is used to perform appearance inspection on the brake disc to be inspected.

[0172] The dimension inspection module is used to inspect the dimensions of the brake disc to be inspected.

[0173] The material performance testing module is used to perform material performance testing on the brake disc to be tested.

[0174] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the technical solutions of the embodiments of the present invention.

Claims

1. A method for testing the safety performance of a brake disc, characterized in that, Includes the following steps: S1: Perform visual inspection on the brake disc to be inspected; Step S1 includes: S1.1: Use a camera to acquire an image of the brake disc to be inspected; S1.2: Perform image preprocessing on the brake disc image to be detected; use an improved adaptive filtering algorithm to denoise the brake disc image to be detected, specifically: Sa: Decompose the brake disc image to be detected into sub-images of multiple scales; Sb: Extract local features from each sub-image, the local features including local variance, gradient intensity, and texture information; Sc: Dynamically assign weights based on the local features; Sd: Perform adaptive filtering on each scale sub-image based on the weights; Se: Reconstruct the adaptively filtered sub-images at multiple scales to obtain the final filtered image; In step Sc, the formula for dynamically allocating weights ω(x, y) is: ； Where α(x,y), β(x,y) and γ(x,y) are weighting coefficients used to balance the influence of different features; ε is a constant. The methods for determining the weighting coefficients α(x,y), β(x,y), and γ(x,y) include: Calculate the global noise level (Global Var) of the image of the brake disc to be detected; Calculate the global texture complexity (Global Texture) of the brake disc image to be detected; Calculate the local noise level Local Var(x,y) and the local texture complexity Local Texture(x,y) for each pixel (x,y); The weighting coefficients α(x, y), β(x, y), and γ(x, y) are determined based on the global noise level, global texture complexity, local noise level of each pixel, and local texture complexity of each pixel. The specific formula is as follows: ； ； ； In the formula, It is a constant; S1.3: Extract the scratch features, pinhole features, and shrinkage features of the brake disc to be tested from the preprocessed image of the brake disc to be tested; S1.4: Input the scratch features, pinhole features and shrinkage features of the brake disc to be tested into the image analysis model to obtain the degree of scratches, pinhole conditions and shrinkage conditions of the brake disc to be tested; S1.5: Conduct a comprehensive evaluation of the appearance performance of the brake disc under test based on the degree of scratches, pinholes, and shrinkage. S2: Perform dimensional inspection on the brake disc to be inspected; S3: Perform material performance testing on the brake disc to be tested.

2. The method for testing the safety performance of a brake disc according to claim 1, characterized in that, In step Sa, the brake disc image to be detected is decomposed into sub-images of multiple scales, and each sub-image corresponds to image information of different frequencies; Step Sa is expressed by the formula: ； Where I is the image of the brake disc to be detected, I s Let S represent the sub-image at the s-th scale, where S is the total number of scales.

3. The method for testing the safety performance of a brake disc according to claim 2, characterized in that, In step Sb, the specific formula is as follows: ； ； ； In the formula, (x, y) represents the pixel coordinates in the image; Window represents a local window region centered at pixel (x, y), used to calculate local statistics; N represents the number of pixels in the local window Window; μ s Represents a sub-image I in a local window (Window). s The average value; Var(I s ) represents the sub-image I in the local window Window. s Local variance; x I s Sub-image I s Gradient in the horizontal direction; y I s Sub-image I s Gradient in the vertical direction; Grad(I) s ) represents sub-image I s The gradient strength; Texture(Is) represents the texture information of the sub-image Is in the local window Window; Is(x, y) represents the sub-image I s The pixels in.

4. The method for testing the safety performance of a brake disc according to claim 1, characterized in that, Step S1.1 includes: S1.1.1: Install the camera at the monitoring station to ensure that the camera can capture images of the brake disc to be inspected from multiple angles; S1.1.2: Install an adjustable light source around the camera; S1.1.3: Obtain the image of the brake disc to be inspected.

5. The method for testing the safety performance of a brake disc according to claim 4, characterized in that, In step S1.1.3, acquiring the image of the brake disc to be detected includes: A surface image is taken from the front of the brake disc to be tested. A side image is taken from the side of the brake disc to be tested. The images were taken from the top, bottom, and oblique angles of the brake disc under test.

6. The method for testing the safety performance of a brake disc according to claim 1, characterized in that, In step S1.2, the image preprocessing further includes grayscale conversion.

7. A brake disc safety performance testing system, characterized in that, The brake disc safety performance testing system employs a brake disc safety performance testing method according to any one of claims 1-6, and the brake disc safety performance testing system comprises: The appearance inspection module is used to perform appearance inspection on the brake disc to be inspected. A size detection module is used to detect the size of the brake disc to be tested. The material performance testing module is used to perform material performance testing on the brake disc to be tested.

Images