Intelligent band-type brake detection system of railway wagon

Abstract

The invention discloses an intelligent band-type brake detection system of a railway wagon, and relates to the technical field of intelligent monitoring of railway wagons. Calculating a total graph gradient to generate a high thermal gradient region; calculating a total contact area and a region area of the closed polygon, calculating a local temperature rise and a temperature field, and calculating a predicted temperature field after local heat conduction; and calculating a band-type brake risk index, and judging a band-type brake state. The method comprises the steps of collecting thermal imaging and band-type brake state data for preprocessing, calculating a total graph gradient to generate a high thermal gradient region, calculating a total contact area and a region area of a closed polygon, calculating a local temperature rise and a temperature field, calculating a predicted temperature field after local heat conduction, calculating a band-type brake risk index, and judging the band-type brake state. The applicability and the generation quality of a complex image generation task are enhanced, the efficiency and the intelligent level of railway wagon band-type brake monitoring are improved, and the stability and the generalization ability of sensing loss in different tasks are ensured.

Description

Technical Field

[0001] The invention relates to the technical field of intelligent monitoring of railway freight cars, in particular to an intelligent brake detection system for railway freight cars. Background Art

[0002] Railway freight cars are an important part of the modern logistics and transportation system. Their operational safety and efficiency are crucial to transportation and economic development. The brake device is the core component of the railway freight car braking system, and its working state directly affects the operational safety of the train. In traditional technologies, the state monitoring of the brake device mainly relies on manual inspection and simple sensor monitoring based on mechanical properties. This method is not only inefficient, but also easily affected by human factors, and cannot meet the high requirements of modern railway transportation for real-time and accuracy. With the development of sensor technology, computer vision and artificial intelligence, the brake state detection system based on multi-source data fusion and intelligent analysis has gradually become a research hotspot. In particular, the key data is collected through technologies such as thermal imaging and pressure sensing, and data processing and fault prediction are combined with deep learning algorithms, which can significantly improve the automation level and detection accuracy of the monitoring system. The current technical solutions still have many deficiencies in the depth, refinement and real-time performance of data processing and fusion, and cannot fully meet the intelligent detection needs of railway freight car brake devices.

[0003] The shortcomings of existing technologies in the field of brake detection are mainly reflected in the following aspects: the existing signal analysis methods lack intelligence and real-time, and fail to fully utilize machine learning or deep learning technology to conduct in-depth analysis and optimization of detection data. In terms of image processing and data fusion, existing technologies are difficult to balance real-time and accuracy, resulting in delayed monitoring of the brake system status. Summary of the invention

[0004] In view of the above problems existing in the existing intelligent brake detection system of railway freight cars, the present invention is proposed.

[0005] Therefore, the problem to be solved by the present invention is that the existing signal analysis methods lack intelligence and real-time capabilities, and fail to fully utilize machine learning or deep learning technologies to conduct in-depth analysis and optimization of detection data. In terms of image processing and data fusion, the existing technologies are difficult to balance real-time and accuracy, resulting in delayed monitoring of the brake system status.

[0006] To solve the above technical problems, the present invention provides the following technical solutions: an intelligent brake detection system for a railway freight car, comprising: a collection and processing module, used to collect thermal imaging and brake status data, and pre-process the collected thermal imaging and brake status data; a filtering calculation module, used to calculate the full-image gradient to generate a high thermal gradient area; a calculation prediction module, used to calculate the total contact area and the area of ​​the closed polygon, calculate the local temperature rise and temperature field, and calculate the predicted temperature field after local heat conduction; an indicator judgment module, used to calculate the brake risk index and judge the brake status; a visualization storage module, used to build a visualization interface to display the brake status results, and store the thermal imaging and brake status data collected and analyzed.

[0007] As a preferred solution of the intelligent brake detection system for railway freight cars of the present invention, wherein: the collection of thermal imaging and brake status data refers to installing a thermal imaging sensor at the front end of the brake device to collect thermal imaging data, installing a pressure sensor at the joint of the brake hydraulic system, and installing an optical slip speed sensor at the side edge of the wheel to collect brake status data; Use the time protocol to synchronize the collected thermal imaging and brake status data; The brake state data includes: pressure data and collected slip speed data.

[0008] As a preferred solution of the intelligent brake detection system for railway freight cars of the present invention, the preprocessing of the collected thermal imaging and brake status data includes: The brake status data is smoothed and denoised using the sliding average method, the outliers of the brake status data are screened and deleted using the interquartile range method, the missing data of the brake status data are filled using the linear interpolation method, and the smoothed brake status data is normalized; The collected thermal imaging data is denoised using Gaussian filtering, the denoised thermal imaging data is converted into a pseudo-color image using Jet mapping, the segmentation window is set using a dynamic window adjustment method, and the pseudo-color image is segmented according to the segmentation window to obtain a segmented image; Calculate the gradient magnitude of the segmented image , use the pre-trained CycleGAN model to generate a blurred mask from the segmented image , extract the fuzzy area ; For fuzzy areas Perform fast Fourier transform to get the spectrum , calculate the spectrum High frequency energy distribution ; Calculate the fuzzy area The mean gradient of and standard deviation , based on the gradient mean and standard deviation Calculate blur kernel parameters ; Based on the fuzzy kernel parameters , construct a two-dimensional Gaussian blur kernel ; Based on high frequency energy distribution Optimizing 2D Gaussian blur kernel , the formula is: ; in is the optimization coefficient; Use optimized 2D Gaussian blur kernel Spectrum Perform compensation to obtain compensated frequency domain data ; For the compensated frequency domain data Use a bandpass Gaussian filter to smooth the frequency domain data. ; The smoothed frequency domain data is converted back to the spatial domain to obtain the compensated blurred area image. ; The compensated blurred area image Replace it with the original segmented image to get the initial segmented compensation image ; Concatenate the initial segmentation compensated image and blur mask into the input tensor , ; Collect thermal images with historical labels for preprocessing and extraction of historical deep features to generate a training set; Thermal imaging based on historical tags uses clear image mapping to obtain a clear image of the target ; Extract tensors using the pre-trained ResNet-50 model The deep features ; Build a generative adversarial network, including a generator and a discriminator; Set the input format of the generative adversarial network to deep features ; Input the deep features into the generator to generate optimized compensated images , using the discriminator to receive the generated optimized compensation image and the target clear image, and output the compensation image and the target clear score; Use the pre-trained VGG network to extract feature representations of the compensated image and the target clear image; Train the GAN using the training set; Calculate perceptual loss using spatial error , using cross entropy loss to calculate adversarial loss ; Based on perceptual loss and combat loss , construct the loss function of the generator , the formula is: ;

[0009] in Feature representation for target clear image; Use cross entropy loss to construct the discriminator loss function ; Use the loss change monitoring method to set the training stop condition when the generator's loss function And the loss function of the discriminator Stop training when the training stop condition is met; Real-time deep features Input into the trained generative adversarial network to obtain the final segmentation compensation image; The final segmented and compensated image is fused using overlapping weights to obtain the corrected thermal imaging data. .

[0010] As a preferred solution of the intelligent brake detection system for railway freight cars of the present invention, wherein: the calculation of the full image gradient to generate the high thermal gradient area refers to calculating the temperature gradients in the horizontal and vertical directions using the Sobel operator based on the corrected thermal imaging data; The full-image gradient is calculated using the gradient modulus method based on the temperature gradients in the horizontal and vertical directions; The mean standard deviation method is used to set the gradient threshold, and the full-image gradient greater than the gradient threshold is marked to generate a high thermal gradient area.

[0011] As a preferred solution of the intelligent brake detection system for railway freight cars of the present invention, wherein: the calculation of the total contact area and the area of ​​the closed polygon refers to setting the neighborhood radius using the k-nearest neighbor distance curve method and setting the minimum number of points using the data dimension method; For each pixel in the high thermal gradient area, the point set in its neighborhood is calculated. When the neighborhood point set is greater than or equal to the minimum number of points, it is determined to be a core point and marked as visited. When the neighborhood point set is less than the minimum number of points, it is determined to be a boundary point. All pixel points in the neighborhood are added to the core point until all pixel points in the high thermal gradient area are visited and the expansion stops. The contact area point set consisting of core points and boundary points is used to calculate the total contact area using discrete integration ; Use the convex hull algorithm to sort the boundary points in geometric order to generate a closed boundary polygon point set; Calculate the area of ​​a closed polygon using the Shoelace Formula .

[0012] As a preferred solution of the intelligent brake detection system for railway freight cars of the present invention, the calculation of the local temperature rise and temperature field and the calculation of the predicted temperature field after local heat conduction refer to obtaining the material parameters of the friction block from the material database and preprocessing them; Calculate the local heat power in the high heat area based on the friction force and sliding speed ; Based on local thermal power and material parameters to calculate the local temperature rise at time t ; Use explicit difference method to separate heat conduction equation and calculate temperature field ; According to the current local temperature rise and temperature field , calculate the predicted temperature field after local heat conduction ; The material parameters include: specific heat capacity, density, thickness and a temperature threshold allowed by the material.

[0013] As a preferred solution of the intelligent brake detection system for railway freight cars of the present invention, the calculation of the brake risk index refers to the calculation of the brake risk index based on the predicted temperature rise. Calculate temperature difference ; Compare the temperature difference with the temperature threshold allowed by the material to generate an over-temperature point , calculate the area of ​​over-temperature region; Based on the average temperature rise and over-temperature area Calculate the brake risk index at time t .

[0014] As a preferred solution of the intelligent brake detection system for railway freight cars of the present invention, wherein: the judging of the brake state refers to setting a risk threshold using statistical distribution; Set the time t brake risk index Compare with the risk threshold. If the brake risk index is greater than or equal to the risk threshold, it is judged as abnormal, triggering an alarm signal and notifying the maintenance personnel via email; If the brake risk index is less than the risk threshold, it is judged to be normal and the thermal imaging and brake status data continue to be monitored.

[0015] As a preferred solution of the intelligent brake detection system for railway freight cars of the present invention, wherein: the construction of a visual interface to display the brake status result refers to using React.js to construct a visual interface to display the brake risk index and brake status; Users who have passed real-name verification are allowed to view the information.

[0016] As a preferred solution of the intelligent brake detection system for railway freight cars described in the present invention, the storage, collection and analysis of the thermal imaging and brake status data refers to storing the collected thermal imaging and brake status data and the brake risk indicators and brake status generated by the analysis in a central database, which is sorted in chronological order and marked with corresponding tags. The collected thermal imaging and brake status data and the brake risk indicators and brake status generated by the analysis are simultaneously backed up in the cloud, and the backup data is regularly checked for integrity, which significantly improves the data quality.

[0017] The beneficial effects of the present invention are as follows: the present invention collects thermal imaging and brake status data, pre-processes the collected thermal imaging and brake status data, calculates the full-image gradient to generate a high thermal gradient area, calculates the total contact area and the area of ​​the closed polygon, calculates the local temperature rise and temperature field, calculates the predicted temperature field after local heat conduction, calculates the brake risk index, and judges the brake status; the applicability and generation quality of complex image generation tasks are enhanced, the efficiency and intelligence level of railway freight car brake monitoring are improved, and the stability and generalization ability of perception loss in different tasks are ensured. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0019] Figure 1 This is a structural diagram of the intelligent brake detection system for railway freight cars.

[0020] Figure 2 The figure is a flow chart of the intelligent brake detection system for railway freight cars. DETAILED DESCRIPTION

[0021] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below in conjunction with the accompanying drawings.

[0022] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0023] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor is it a separate or selective embodiment that is mutually exclusive with other embodiments.

[0024] Example 1, reference Figure 1 and Figure 2 , which is the first embodiment of the present invention, and which provides an intelligent brake detection system for a railway freight car, the intelligent brake detection system for a railway freight car comprising: S1, a collection and processing module, used to collect thermal imaging and brake status data, and pre-process the collected thermal imaging and brake status data; Specifically, collecting thermal imaging and brake status data means installing a thermal imaging sensor at the front end of the brake device to collect thermal imaging data, installing a pressure sensor at the joint of the brake hydraulic system, and installing an optical slip speed sensor at the side edge of the wheel to collect brake status data; Use the time protocol to synchronize the collected thermal imaging and brake status data; The brake state data includes: pressure data and collected slip speed data.

[0025] By combining multiple data sources such as thermal imaging, pressure and slip speed, the invention can realize comprehensive monitoring of the brake status. In particular, the introduction of thermal imaging data provides a new dimension for abnormal detection of the brake status, greatly improving the detection accuracy and reliability. By detecting the slip speed between the wheel and the track under the action of the brake, it is possible to analyze whether the brake performance is normal and prevent excessive braking or slipping out of control. The time protocol is used to synchronize data, so that the system can realize fusion analysis of multi-source data under transient conditions. Compared with traditional detection methods that rely on a single data source, the system significantly improves real-time and intelligent levels.

[0026] Furthermore, preprocessing of the collected thermal imaging and brake status data includes: The brake status data is smoothed and denoised using the sliding average method, the outliers of the brake status data are screened and deleted using the interquartile range method, the missing data of the brake status data are filled using the linear interpolation method, and the smoothed brake status data is normalized; The collected thermal imaging data is denoised using Gaussian filtering, the denoised thermal imaging data is converted into a pseudo-color image using Jet mapping, the segmentation window is set using a dynamic window adjustment method, and the pseudo-color image is segmented according to the segmentation window to obtain a segmented image; Calculate the gradient magnitude of the segmented image ; Use the pre-trained CycleGAN model to generate a blurred mask from the segmented image , 1 means the pixel belongs to the blurred area, 0 means the pixel belongs to the non-blurred area, extract the blurred area ; For fuzzy areas Perform fast Fourier transform to get the spectrum , calculate the spectrum High frequency energy distribution ; Calculate the fuzzy area The mean gradient of and standard deviation , based on the gradient mean and standard deviation Calculate blur kernel parameters , the formula is: ; Based on the fuzzy kernel parameters , construct a two-dimensional Gaussian blur kernel , the formula is: ; Where u and v are the frequency domain coordinates after Fourier transformation, corresponding to the frequency components of the image in the horizontal and vertical directions respectively; Based on high frequency energy distribution Optimizing 2D Gaussian blur kernel , the formula is: ; in is the optimization coefficient; High-frequency information usually represents texture or detail information in image processing, and blur kernel optimization needs to accurately consider these high-frequency characteristics. By introducing high-frequency energy distribution, the processing intensity of the blur kernel on the high-frequency area can be effectively controlled. This targeted method enables blur kernel optimization to avoid over-smoothing details and maintain image texture features. The optimization coefficient is adaptively adjusted according to different high-frequency energy distributions. Traditional two-dimensional Gaussian blur kernel optimization methods are usually adjusted based on global characteristics, while this method adjusts the blur kernel through local high-frequency energy distribution, thereby more accurately optimizing different areas. Two-dimensional Gaussian blur kernels are usually based on fixed parameters or simple global optimization adjustments, and fail to fully consider the impact of high-frequency characteristics on blur effects. The present invention introduces , directly integrating high-frequency distribution information into the blur kernel optimization, significantly improving the ability to retain high-frequency characteristics. The optimization formula of the present invention is particularly suitable for image processing scenarios with high resolution and complex textures, such as medical imaging, remote sensing imaging, and industrial detection.

[0027] Use optimized 2D Gaussian blur kernel Spectrum Perform compensation to obtain compensated frequency domain data , the formula is: ; in is the regularization parameter to prevent the denominator from being zero; For the compensated frequency domain data Use a bandpass Gaussian filter to smooth the frequency domain data. , the formula is: ; ; in is the Gaussian filter position in the frequency domain The value of To control the frequency range, it is dynamically adjusted according to the image resolution and blur level; Convert the smoothed frequency domain data back to the spatial domain to obtain the compensated blurred area image , the formula is: ; Where k is the dimension of the segmented image, t is the time, g is the imaginary unit, H is the height of the segmented image, and W is the width of the segmented image. is the contribution of the vertical frequency component, is the contribution of the horizontal frequency component, i and j are the pixel coordinates in the spatial domain, representing the vertical and horizontal positions of the image respectively; The compensated blurred area image Replace it with the original segmented image to get the initial segmented compensation image ; Concatenate the initial segmentation compensated image and blur mask into the input tensor , ; Collect thermal images with historical labels for preprocessing and extraction of historical deep features to generate a training set; Thermal imaging based on historical tags uses clear image mapping to obtain a clear image of the target ; Extract tensors using the pre-trained ResNet-50 model The deep features ; Build a generative adversarial network, including a generator and a discriminator; Set the input format of the generative adversarial network to deep features ; Input the deep features into the generator to generate optimized compensated images , using the discriminator to receive the generated optimized compensation image and the target clear image, and output the compensation image and the target clear score; Train the GAN using the training set; Use the pre-trained VGG network to extract feature representations of the compensated image and the target clear image; Calculate perceptual loss using spatial error , the formula is: ; in To compensate for the feature representation of the image at layer l, is the feature representation of the target clear image at the lth layer, where l is the feature layer; Calculate adversarial loss using cross entropy loss , the formula is: ; in is the score of the discriminator on the compensated image; Based on perceptual loss and combat loss , construct the loss function of the generator , the formula is: ; Perceptual loss can measure the similarity between the generated image and the target clear image in the high-level feature space in the image generation task. This high-level feature usually corresponds to the texture and content information of the image. It is difficult to capture these high-level characteristics by directly using pixel-level loss (such as mean square error). Perceptual loss is based on the feature representation of a specific layer in a pre-trained network (such as the VGG network), which can better optimize the generator to generate images with more natural textures. Adversarial loss encourages the generator to generate results closer to the real data distribution through the feedback of the discriminator. Using only perceptual loss may cause the generated image to be close to the target image in the feature space, but the overall visual effect (such as image clarity) is insufficient. Perceptual loss normalization ensures that the loss calculation will not be unbalanced due to the scale difference of the target feature values. Through the weighted combination of perceptual loss and adversarial loss, the two optimization goals of feature similarity and image realism are comprehensively considered. This joint optimization strategy has unique advantages in high-resolution image generation tasks and can achieve a better balance between generated image quality and feature retention. In many applications of perceptual loss, the target features are not optimized. Normalization processing may lead to unbalanced loss weights between different image data sets or feature layers. By normalizing the eigenvalues ​​of the target clear image, the stability and generalization ability of the perceptual loss in different tasks are ensured. Existing methods often use perceptual loss or adversarial loss alone, which leads to the generator optimization being too biased towards specific goals. This formula combines perceptual loss and adversarial loss to take into account both the matching of high-level features and the optimization of overall realism, thereby achieving high-quality output of generated images. The present invention utilizes the feature representation of a specific layer to ensure that the optimization process is adjusted for the key feature level, thereby improving the generation ability of the generator. The perceptual loss of this formula is based on high-level features and can capture complex textures and semantic information. At the same time, combined with adversarial loss, it further enhances the realism and detail expression of the generated image.

[0028] Use cross entropy loss to construct the loss function of the discriminator , the formula is: ; in is the score of the discriminator for the target clear image, is the score of the discriminator on the compensated image; Use the loss change monitoring method to set the training stop condition when the generator's loss function And the loss function of the discriminator Stop training when the training stop condition is met; Real-time deep features Input into the trained generative adversarial network to obtain the final segmentation compensation image; The final segmented and compensated image is fused using overlapping weights to obtain the corrected thermal imaging data. .

[0029] Through the combined use of sliding average method, interquartile range method and linear interpolation method, the brake status data can be efficiently cleaned and repaired to ensure its high stability and accuracy, improve the reliability of subsequent analysis, and avoid the result deviation caused by abnormal or missing data. Through Gaussian filtering denoising and Jet mapping pseudo-color processing, the noise in the thermal imaging data can be significantly reduced, and the visual effect of the image is more intuitive. The CycleGAN model is used to generate blur masks and extract blur areas. The high-frequency energy distribution is analyzed by fast Fourier transform, and the two-dimensional Gaussian blur kernel is optimized. The blur area in the thermal imaging data can be accurately identified and effectively compensated. The compensated image is clearer and the information expression is more complete. The constructed generative adversarial network can generate a compensated image close to the target clear image through deep feature extraction and perceptual loss optimization. The final segmented compensated image is corrected by overlapping weight fusion, which can significantly improve the overall accuracy and reliability of the thermal imaging data.

[0030] S2, filtering calculation module, used to calculate the full image gradient to generate high thermal gradient areas; Specifically, calculating the full image gradient to generate a high thermal gradient area refers to calculating the temperature gradients in the horizontal and vertical directions using the Sobel operator based on the corrected thermal imaging data; The full-image gradient is calculated using the gradient modulus method based on the temperature gradients in the horizontal and vertical directions; The mean standard deviation method is used to set the gradient threshold, and the full-image gradient greater than the gradient threshold is marked to generate a high thermal gradient area.

[0031] The Sobel operator can efficiently calculate the temperature gradients in the horizontal and vertical directions of the corrected thermal imaging data, thereby accurately capturing the directional information of the temperature change on the surface of the brake device, helping the system to identify possible overheating edges and abnormal heat source locations. The temperature gradients in the horizontal and vertical directions are integrated into a scalar value through the gradient modulus method, which can intuitively reflect the gradient intensity of each pixel point. The generation of the full-image gradient can effectively show the area with the most significant temperature change in the thermal imaging image. The gradient threshold is set by the mean standard deviation method, and the threshold range can be dynamically adjusted according to the data distribution, thereby improving the extraction accuracy of the high thermal gradient area. Marking the high thermal gradient area can significantly highlight the overheating parts that need to be paid attention to in the thermal imaging image, and intuitively display the areas where abnormalities may exist. This dynamic adjustment mechanism significantly improves the accuracy of high thermal gradient area extraction and avoids missed detection or false detection problems caused by fixed thresholds. Compared with the traditional overall thermal map analysis method, this area marking method can more accurately lock the problem area and improve the diagnostic efficiency. By analyzing the thermal imaging data in real time, the system can identify possible overheating problems of the equipment in advance, thereby preventing safety hazards caused by brake failure or wear.

[0032] S3, calculation prediction module, used to calculate the total contact area and the area of ​​the closed polygon, calculate the local temperature rise and temperature field, and calculate the predicted temperature field after local heat conduction; Specifically, calculating the total contact area and the area of ​​the closed polygon refers to setting the neighborhood radius using the k-nearest neighbor distance curve method and setting the minimum number of points using the data dimension method; For each pixel in the high thermal gradient area, the point set in its neighborhood is calculated. When the neighborhood point set is greater than or equal to the minimum number of points, it is determined to be a core point and marked as visited. When the neighborhood point set is less than the minimum number of points, it is determined to be a boundary point. All pixel points in the neighborhood are added to the core point until all pixel points in the high thermal gradient area are visited and the expansion stops. The contact area point set consisting of core points and boundary points is used to calculate the total contact area using discrete integration ; Use the convex hull algorithm to sort the boundary points in geometric order to generate a closed boundary polygon point set; Calculate the area of ​​a closed polygon using the Shoelace Formula .

[0033] The neighborhood radius is set by the k-nearest neighbor distance curve method, and the minimum number of points is set by combining the data dimension method, so that the pixel clustering in the high thermal gradient area has higher robustness and adaptability, effectively distinguishes the core points and the boundary points, and reduces the interference of noise on the area extraction. By recursively expanding the neighborhood set of each pixel point in the high thermal gradient area, all relevant pixels can be completely covered to form a contact area point set. The generation method ensures the coherence and integrity of the high thermal area extraction. The area of ​​the contact area point set is calculated by the discrete integral method, which can accurately quantify the total contact area of ​​the high thermal area, providing an intuitive basis for evaluating the braking effect and equipment status of the brake device. The boundary points are geometrically sorted by the convex hull algorithm to generate a closed boundary polygon point set, ensuring that the boundary definition of the high thermal gradient area has high geometric accuracy. This boundary information is convenient for subsequent area calculation and shape analysis. The area of ​​the closed polygon is quickly calculated by the Shoelace Formula, which can effectively quantify the shape characteristics and area size of the high thermal area. Combined with the total contact area analysis, the performance of the brake device can be more comprehensively evaluated.

[0034] Further, calculating the local temperature rise and temperature field, and calculating the predicted temperature field after local heat conduction means obtaining material parameters of the friction block from a material database and performing preprocessing; Calculate the local heat power in the high heat area based on the friction force and sliding speed , the formula is: ; in is the pressure at time t, is the slip velocity at time t; Based on local thermal power and material parameters to calculate the local temperature rise at time t , the formula is: ; in is the time interval, is the specific heat capacity, is the density, is the friction block thickness; Use explicit difference method to separate heat conduction equation and calculate temperature field ; According to the current local temperature rise and temperature field Calculate the predicted temperature field after local heat conduction , the formula is: ; The material parameters include: specific heat capacity, density, thickness and a temperature threshold allowed by the material.

[0035] The material parameters of the friction block are obtained from the material database and preprocessed to make the calculation more accurate and efficient. Material parameters such as specific heat capacity and density directly affect the rate of heat conduction and the amplitude of local temperature rise, while the temperature threshold helps to determine whether the material is in a safe working range. The local thermal power is calculated based on the friction force and the slip speed, which can quantitatively describe the heat generation rate in the high-heat area, provide initial conditions for the subsequent local temperature rise and temperature field calculation, and ensure the accuracy of the heat conduction analysis. By combining the local thermal power with the material parameters, the local temperature rise change can be accurately predicted. The local temperature rise data can be used to identify possible overheating areas in the brake device, which helps to discover potential material failure risks in advance. The explicit difference method discretizes the continuous heat conduction equation and can quickly calculate the dynamic changes of temperature in time and space. By combining the local temperature rise and the temperature distribution after heat conduction, the predicted temperature field can intuitively present the diffusion state of heat after time t. This prediction capability can identify the heat conduction characteristics and possible failure risks of high-heat areas in advance.

[0036] S4, index judgment module, used to calculate the brake risk index and judge the brake state; Specifically, the calculation of the brake risk index refers to the predicted temperature rise Calculate temperature difference ; Compare the temperature difference with the temperature threshold allowed by the material to generate an over-temperature point , calculate the over-temperature area , the formula is: ; in is the physical spacing of pixels in the horizontal direction, is the physical spacing of pixels in the vertical direction; Calculate the average temperature rise in the overheated area ; Based on the average temperature rise and over-temperature area Calculate the brake risk index at time t , the formula is: ; in is the total contact area.

[0037] The predicted temperature rise can directly reflect the local thermal response of the brake device. By calculating the temperature difference with the initial temperature, the thermal stress changes that the material is subjected to during operation can be clearly identified, which helps to discover potential thermal runaway risks in advance. The predicted temperature rise can directly reflect the local thermal response of the brake device. By calculating the temperature difference with the initial temperature, the thermal stress changes that the material is subjected to during operation can be clearly identified, which helps to discover potential thermal runaway risks in advance. High-temperature points that may exceed the material's tolerance limit can be quickly screened out to provide guidance for subsequent thermal management measures. The spatial distribution characteristics of high-temperature risks can be quantified by calculating the over-temperature area through spatial integration of over-temperature points. The larger the over-temperature area, the higher the potential risk of thermal runaway of the equipment. The average temperature rise and the over-temperature area are combined to calculate the brake risk index, which can comprehensively evaluate the thermodynamic performance and risk status of the equipment. The risk index is a comprehensive parameter that can quantify the safety margin in equipment operation.

[0038] Further, judging the brake state means setting a risk threshold using statistical distribution; Set the time t brake risk index Compare with the risk threshold. If the brake risk index is greater than or equal to the risk threshold, it is judged as abnormal, triggering an alarm signal and notifying the maintenance personnel via email; If the brake risk index is less than the risk threshold, it is judged to be normal and the thermal imaging and brake status data continue to be monitored.

[0039] By analyzing the historical statistical distribution of the brake risk index, the risk threshold that adapts to the current working conditions can be dynamically set. Compared with the fixed threshold method, this dynamic adjustment method can better adapt to different environmental conditions and equipment conditions, and significantly improve the accuracy of the monitoring system. By comparing the real-time calculated brake risk index with the dynamically set risk threshold, the operating status of the equipment can be quickly judged. This process can efficiently identify potential high temperatures or abnormal conditions, and trigger the corresponding response mechanism based on the results. When the brake risk index is greater than or equal to the risk threshold, the system automatically triggers an alarm signal and notifies maintenance personnel via email. The notification method shortens the equipment failure response time and reduces the risk caused by failure delays. When the brake risk index is lower than the risk threshold, the system continues to monitor the thermal imaging data and brake status data, thereby achieving continuous equipment operation health management. Through continuous monitoring, the system can timely capture the changing trend of potential risks.

[0040] S5, a visualization storage module, used to construct a visualization interface to display the brake status results, store the thermal imaging and brake status data generated by the collection and analysis; Specifically, building a visual interface to display the brake status result refers to using React.js to build a visual interface to display the brake risk indicator and the brake status; Users who have passed real-name verification are allowed to view the information.

[0041] By using React.js to build a visual interface, the page can be quickly updated and responded to dynamically, ensuring that the brake risk indicators and status information can be presented to users in real time. The componentized design of React.js also makes interface development more efficient and facilitates subsequent functional expansion. By visually displaying the brake risk indicators and status information, users can intuitively understand the operation of the equipment. Through the real-name verification mechanism, the system can effectively limit data access rights to ensure that only authorized users can view the brake status information.

[0042] Furthermore, storing the thermal imaging and brake status data generated by the collection and analysis means storing the collected thermal imaging and brake status data and the brake risk indicators and brake status generated by the analysis in a central database, sorting the data in chronological order and marking corresponding tags in the central database, and synchronously backing up the collected thermal imaging and brake status data and the brake risk indicators and brake status generated by the analysis in the cloud, and regularly performing integrity checks on the backup data.

[0043] By storing the collected thermal imaging and brake status data as well as the brake risk indicators generated by analysis in a central database, centralized data management can be achieved. Time sorting ensures the storage logic and retrieval efficiency of data, while label marking enhances the flexibility of data classification. The cloud backup function provides additional data security protection and can quickly restore data when a central database fails. By regularly performing integrity checks on backup data, it can ensure that the data has not been damaged or tampered with, providing accurate and reliable basic data for subsequent analysis and decision-making.

Claims

1. An intelligent brake detection system for railway freight cars, characterized in that: include: A collection and processing module, used for collecting thermal imaging and brake status data, and pre-processing the collected thermal imaging and brake status data; The filtering calculation module is used to calculate the full-image gradient to generate high thermal gradient areas; The calculation prediction module is used to calculate the total contact area and the area of ​​the closed polygon, calculate the local temperature rise and temperature field, and calculate the predicted temperature field after local heat conduction; Index judgment module, used to calculate the brake risk index and judge the brake status; The visualization storage module is used to build a visualization interface to display the brake status results, and to store, collect and analyze the generated thermal imaging and brake status data.

2. The intelligent brake detection system for railway freight cars according to claim 1, characterized in that: The collecting of thermal imaging and brake status data refers to installing a thermal imaging sensor at the front end of the brake device to collect thermal imaging data, installing a pressure sensor at the joint of the brake hydraulic system, and installing an optical slip speed sensor at the side edge of the wheel to collect brake status data; Use the time protocol to synchronize the collected thermal imaging and brake status data; The brake state data includes: pressure data and collected slip speed data.

3. The intelligent brake detection system for railway freight cars according to claim 2, characterized in that: The preprocessing of the collected thermal imaging and brake status data includes: The brake status data is smoothed and denoised using the sliding average method, the outliers of the brake status data are screened and deleted using the interquartile range method, the missing data of the brake status data are filled using the linear interpolation method, and the smoothed brake status data is normalized; The collected thermal imaging data is denoised using Gaussian filtering, the denoised thermal imaging data is converted into a pseudo-color image using Jet mapping, the segmentation window is set using a dynamic window adjustment method, and the pseudo-color image is segmented according to the segmentation window to obtain a segmented image; Calculate the gradient magnitude of the segmented image , use the pre-trained CycleGAN model to generate a blurred mask from the segmented image , extract the fuzzy area ; For fuzzy areas Perform fast Fourier transform to get the spectrum , calculate the spectrum High frequency energy distribution ; Calculate the fuzzy area The mean gradient of and standard deviation , based on the gradient mean and standard deviation Calculate blur kernel parameters ; Based on the fuzzy kernel parameters , construct a two-dimensional Gaussian blur kernel ; Based on high frequency energy distribution Optimizing 2D Gaussian blur kernel , the formula is: ； in is the optimization coefficient; Use optimized 2D Gaussian blur kernel Spectrum Perform compensation to obtain compensated frequency domain data ; For the compensated frequency domain data Use a bandpass Gaussian filter to smooth the frequency domain data. ; The smoothed frequency domain data is converted back to the spatial domain to obtain the compensated blurred area image. ; The compensated blurred area image Replace it with the original segmented image to get the initial segmented compensation image ; Concatenate the initial segmentation compensated image and the blur mask into the input tensor , ; Collect thermal images with historical labels for preprocessing and extraction of historical deep features to generate a training set; Thermal imaging based on historical tags uses clear image mapping to obtain a clear image of the target ; Extract tensors using the pre-trained ResNet-50 model The deep features ; Build a generative adversarial network, including a generator and a discriminator; Set the input format of the generative adversarial network to deep features ; Input the deep features into the generator to generate optimized compensated images , using the discriminator to receive the generated optimized compensation image and the target clear image, and output the compensation image and the target clear score; Use the pre-trained VGG network to extract feature representations of the compensated image and the target clear image; Train the GAN using the training set; Calculate perceptual loss using spatial error , using cross entropy loss to calculate adversarial loss ; Based on perceptual loss and combat loss , construct the loss function of the generator , the formula is: ； in Feature representation for target clear image; Use cross entropy loss to construct the discriminator loss function ; Use the loss change monitoring method to set the training stop condition when the generator's loss function And the loss function of the discriminator Stop training when the training stop condition is met; Real-time deep features Input into the trained generative adversarial network to obtain the final segmentation compensation image; The final segmented and compensated image is fused using overlapping weights to obtain the corrected thermal imaging data. .

4. The intelligent brake detection system for railway freight cars as claimed in claim 3, characterized in that: The calculation of the full image gradient to generate the high thermal gradient area refers to calculating the temperature gradients in the horizontal and vertical directions using the Sobel operator based on the corrected thermal imaging data; The full-image gradient is calculated using the gradient modulus method based on the temperature gradients in the horizontal and vertical directions; The mean standard deviation method is used to set the gradient threshold, and the full-image gradient greater than the gradient threshold is marked to generate a high thermal gradient area.

5. The intelligent brake detection system for railway freight cars according to claim 4, characterized in that: The calculation of the total contact area and the area of ​​the closed polygon refers to setting the neighborhood radius using the k-nearest neighbor distance curve method and setting the minimum number of points using the data dimension method; For each pixel in the high thermal gradient area, the point set in its neighborhood is calculated. When the neighborhood point set is greater than or equal to the minimum number of points, it is determined to be a core point and marked as visited. When the neighborhood point set is less than the minimum number of points, it is determined to be a boundary point. All pixel points in the neighborhood are added to the core point until all pixel points in the high thermal gradient area are visited and the expansion stops. The contact area point set consisting of core points and boundary points is used to calculate the total contact area using discrete integration ; Use the convex hull algorithm to sort the boundary points in geometric order to generate a closed boundary polygon point set; Calculate the area of ​​a closed polygon using the Shoelace Formula .

6. The intelligent brake detection system for railway freight cars according to claim 5, characterized in that: The calculation of the local temperature rise and temperature field, and the calculation of the predicted temperature field after the local heat conduction refers to obtaining the material parameters of the friction block from the material database and performing preprocessing; Calculate the local heat power in the high heat area based on the friction force and sliding speed ; Based on local thermal power and material parameters to calculate the local temperature rise at time t ; Use explicit difference method to separate heat conduction equation and calculate temperature field ; According to the current local temperature rise and temperature field , calculate the predicted temperature field after local heat conduction ; The material parameters include: specific heat capacity, density, thickness and a temperature threshold allowed by the material.

7. The intelligent brake detection system for railway freight cars according to claim 6, characterized in that: The calculation of the brake risk index refers to the predicted temperature rise Calculate temperature difference ; Compare the temperature difference with the temperature threshold allowed by the material to generate an over-temperature point , calculate the area of ​​over-temperature region; Based on the average temperature rise and over-temperature area Calculate the brake risk index at time t .

8. The intelligent brake detection system for railway freight cars according to claim 7, characterized in that: Determining the brake state refers to setting a risk threshold using statistical distribution; Set the time t brake risk index Compare with the risk threshold. If the brake risk index is greater than or equal to the risk threshold, it is judged as abnormal, triggering an alarm signal and notifying the maintenance personnel via email; If the brake risk index is less than the risk threshold, it is judged to be normal and the thermal imaging and brake status data continue to be monitored.

9. The intelligent brake detection system for railway freight cars according to claim 8, characterized in that: The constructing a visual interface to display the brake status result refers to using React.js to construct a visual interface to display the brake risk index and the brake status; Users who have passed real-name verification are allowed to view the information.

10. The intelligent brake detection system for railway freight cars according to claim 9, characterized in that: The storing, collecting and analyzing the thermal imaging and brake status data refers to storing the collected thermal imaging and brake status data and the brake risk indicators and brake status generated by the analysis in a central database, sorting the central database in chronological order and marking corresponding tags, synchronously backing up the collected thermal imaging and brake status data and the brake risk indicators and brake status generated by the analysis in the cloud, and regularly performing integrity checks on the backup data.

Images