Modular train braking method and system

Abstract

The application relates to the technical field of train braking, and provides a modular train braking method and system, which are used for solving the problems of lack of space pertinence of brake pressure adjustment, aggravation of local heat accumulation and low heat management efficiency in the prior art. The method comprises the following steps: generating a temperature distribution image according to temperature distribution data, identifying a local hotspot area exceeding a preset safety range, and extracting target geometric feature parameters of the local hotspot area; combining current running state information of the train, and generating brake pressure adjustment parameters of the train by using a heat conduction model; optimizing the pressure adjustment parameters, so that a brake executing mechanism applies a dynamically changing brake force to a local friction contact area corresponding to the local hotspot area based on the optimized brake pressure adjustment parameters. The technical scheme provided by the application realizes real-time monitoring of brake disc temperature and identification of local overheating characteristics, dynamically adjusts brake pressure to accurately control temperature, and ensures braking safety and reliability.

Description

Technical Field

[0001] This application relates to the field of train braking technology, and in particular to a modular train braking method and system. Background Technology

[0002] During high-speed braking of a train, the friction contact area between the brake disc and brake pads is prone to instantaneous high temperatures (locally exceeding 800°C) due to intense energy conversion, leading to material thermal degradation, thermal cracking, and a decline in braking performance. To ensure safety and stability under emergency braking conditions, it is necessary to acquire the temperature field distribution characteristics of the friction interface in real time, accurately identify abnormally overheated areas, and achieve rapid adaptive adjustment of braking force distribution based on multi-dimensional dynamic parameters to avoid the risk of brake failure caused by local thermal runaway.

[0003] The existing solution uses an infrared thermal imager to monitor the temperature distribution of the brake disc, extracts the overheated area through threshold segmentation, and calculates the overall brake pressure compensation value by combining a fixed weight algorithm (such as area weighted average). Finally, the braking force is uniformly adjusted through the brake control system.

[0004] Existing solutions rely solely on simple area statistics of high-temperature regions, resulting in a lack of spatial targeting in brake pressure adjustments, exacerbating localized heat accumulation, and reducing the effectiveness of thermal management. Summary of the Invention

[0005] This application provides a modular train braking method and system to solve the problems of insufficient spatial targeting of braking pressure adjustment, aggravated local heat accumulation, and low thermal management efficiency in the prior art.

[0006] In a first aspect, this application provides a modular train braking method, comprising:

[0007] Acquire temperature distribution data of the target friction contact area on the train, wherein the target friction contact area is the area on the train's brake disc surface that contacts the brake actuator.

[0008] Based on the temperature distribution image generated from the temperature distribution data, local hotspot areas whose temperatures exceed a preset safety range are identified from the temperature distribution image, and target geometric feature parameters of the local hotspot areas are extracted from the temperature distribution image.

[0009] Based on the target geometric feature parameters of the local hotspot area and combined with the current operating status information of the train, the braking pressure adjustment parameters of the train are generated using a heat conduction model.

[0010] The brake pressure adjustment parameters are optimized to obtain the optimized brake pressure adjustment parameters;

[0011] The optimized braking pressure adjustment parameters are sent to the braking actuator in the modular train braking system, so that the braking actuator applies dynamically changing braking force to the local friction contact area corresponding to the local hot spot area based on the optimized braking pressure adjustment parameters.

[0012] Optionally, the step of identifying local hotspot regions whose temperatures exceed a preset safety range from the temperature distribution image, and extracting target geometric feature parameters of the local hotspot regions from the temperature distribution image, includes:

[0013] In the temperature distribution image, local hotspot areas where the temperature value exceeds the preset safe temperature range are identified by using a preset safe temperature range threshold.

[0014] Boundary detection is performed on the local hotspot region to obtain the boundary contour of the local hotspot region, and the initial geometric feature parameters of the local hotspot region are calculated based on the boundary contour.

[0015] Based on the initial geometric feature parameters, morphological analysis is performed on the local hotspot region to obtain the structural distribution characteristics of the local hotspot region;

[0016] The initial geometric feature parameters are fused with the structural distribution features to generate intermediate geometric feature parameters for local hotspot regions;

[0017] The intermediate geometric feature parameters are optimized to remove redundant parameters and retain feature parameters whose contribution to the description of local hotspot regions is greater than a preset threshold, thereby generating target geometric feature parameters for local hotspot regions.

[0018] Optionally, fusing the initial geometric feature parameters with the structural distribution features to generate intermediate geometric feature parameters for local hotspot regions includes:

[0019] The initial geometric feature parameters and structural distribution features are normalized respectively to construct a standardized set of geometric parameter vectors and a set of distribution feature vectors that are spatially aligned with the local hotspot region;

[0020] Calculate the dynamic weight coefficients corresponding to the standardized geometric parameter vector set and the distribution feature vector set, respectively;

[0021] Based on the dynamic weighting coefficients, the geometric parameter vector set and the distribution feature vector set are nonlinearly superimposed to obtain the intermediate geometric feature parameters of the local hotspot region.

[0022] Optionally, the step of generating the train's braking pressure adjustment parameters based on the target geometric feature parameters of the local hotspot region, combined with the train's current operating status information, using a heat conduction model includes:

[0023] The target geometric feature parameters of the local hotspot region are spatially discretized to generate a standardized geometric parameter matrix aligned with the grid cells of the brake disc.

[0024] Extract real-time speed, braking duration, and ambient temperature parameters from the train's current operating status information to construct a dynamic operating status vector;

[0025] The standardized geometric parameter matrix is ​​coupled with the dynamic operating state vector in multiple dimensions to generate a joint input feature set containing the heat diffusion path and mechanical stress distribution;

[0026] Based on the joint input feature set, the dynamic evolution process of the surface temperature field and internal thermal stress field of the brake disc is calculated through the heat conduction equation in the heat conduction model, generating spatiotemporal distribution data to characterize the temperature gradient and stress concentration region.

[0027] Based on the spatiotemporal distribution data, with the goal of homogenizing the thermal load on the brake disc, the brake pressure adjustment parameters are solved iteratively in reverse, and the effectiveness of the parameters is verified by boundary constraints. Finally, the brake pressure adjustment parameters of the train are output.

[0028] Optionally, optimizing the brake pressure adjustment parameters to obtain optimized brake pressure adjustment parameters includes:

[0029] The braking pressure adjustment parameters are discretized to generate a parameter distribution matrix that maps one-to-one with the grid cells of the brake disc.

[0030] Based on the preset mechanical structure constraints of the brake disc, the boundary stability of the parameters corresponding to each grid cell in the parameter distribution matrix is ​​tested to select the adjustable parameter range.

[0031] Within the adjustable parameter range, based on the dynamic coupling relationship between braking pressure and temperature field, the parameters of adjacent units are iteratively weighted to generate an initial optimized parameter sequence.

[0032] Perform multiphysics joint verification on the initial optimized parameter sequence to obtain the verified parameter sequence;

[0033] Incremental update rules are generated based on the verified parameter sequence. Based on the incremental update rules and combined with the real-time pressure feedback data of the braking system, optimized braking pressure adjustment parameters are generated.

[0034] Optionally, performing multiphysics joint verification on the initial optimized parameter sequence to obtain the verified parameter sequence includes:

[0035] Based on the parameters corresponding to each grid cell in the initial optimization parameter sequence, detect grid cells with abnormal heat conduction continuity index and stress field abrupt changes;

[0036] The state of the mesh cells with abrupt stress field changes is adjusted to the state corresponding to the parameters of the previous iteration, the adjustment step size is reduced, and the initial optimized parameter sequence is adjusted according to the adjustment compensation to generate a local correction parameter sequence. The local correction parameter sequence is then verified by three consecutive iterations.

[0037] Based on the validated local correction parameter sequence, a validated post-parameter sequence covering all mesh elements and satisfying multi-field equilibrium conditions is generated.

[0038] Optionally, calculating the dynamic weight coefficients corresponding to the standardized geometric parameter vector set and the distribution feature vector set respectively includes:

[0039] Calculate the initial spatial weights and initial feature weights based on the spatial correlation of the brake disc;

[0040] Based on the preset mechanical structural constraints of the brake disc, the initial spatial weights and the initial feature weights are dynamically adjusted.

[0041] The adjusted spatial weights and feature weights are nonlinearly superimposed according to the thermal coupling strength between grid cells to generate dynamic weight coefficients.

[0042] Secondly, this application provides a modular train braking system, comprising:

[0043] The acquisition module is used to acquire temperature distribution data of the target friction contact area on the train, wherein the target friction contact area is the area on the train's brake disc surface that contacts the brake actuator.

[0044] The extraction module is used to identify local hotspot regions whose temperatures exceed a preset safety range from a temperature distribution image generated based on temperature distribution data, and to extract target geometric feature parameters of the local hotspot regions from the temperature distribution image.

[0045] The generation module is used to generate the train's braking pressure adjustment parameters based on the target geometric feature parameters of the local hot spot area and the train's current operating status information, using a heat conduction model.

[0046] An optimization module is used to optimize the brake pressure adjustment parameters to obtain optimized brake pressure adjustment parameters;

[0047] The sending module sends the optimized braking pressure adjustment parameters to the braking actuator in the modular train braking system, so that the braking actuator applies dynamically changing braking force to the local friction contact area corresponding to the local hot spot area based on the optimized braking pressure adjustment parameters.

[0048] Thirdly, this application provides a computing device, including a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are invoked and executed by the processing component to implement a modular train braking method as described in any of the first aspects.

[0049] Fourthly, this application provides a computer storage medium storing a computer program, which, when executed by a computer, implements a modular train braking method as described in any of the first aspects.

[0050] The beneficial effects of this application are:

[0051] This application provides a modular train braking method, which includes: acquiring temperature distribution data of a target friction contact area on the train, wherein the target friction contact area is the area where the surface of the brake disc on the train contacts the brake actuator; generating a temperature distribution image based on the temperature distribution data, identifying local hot spots whose temperatures exceed a preset safety range from the temperature distribution image, and extracting target geometric feature parameters of the local hot spots from the temperature distribution image; generating brake pressure adjustment parameters for the train based on the target geometric feature parameters of the local hot spots and the train's current operating status information using a heat conduction model; optimizing the brake pressure adjustment parameters to obtain optimized brake pressure adjustment parameters; and sending the optimized brake pressure adjustment parameters to the brake actuator in the modular train braking system, so that the brake actuator applies dynamically changing braking force to the local friction contact area corresponding to the local hot spots based on the optimized brake pressure adjustment parameters.

[0052] This application achieves precise management of the temperature distribution in the friction contact area of ​​the brake disc through real-time monitoring, intelligent analysis, and dynamic control, effectively improving the safety, reliability, and economy of the braking system. By acquiring temperature distribution data of the contact area between the brake disc surface and the brake actuator, it upgrades from traditional single-point temperature measurement to two-dimensional temperature field monitoring, enabling comprehensive capture of temperature changes in the friction contact area, avoiding missed detection of local high-temperature points, and providing a high-resolution data foundation for subsequent analysis. Based on temperature distribution images, it identifies local hotspot areas exceeding the safe range and extracts their geometric feature parameters (such as area, shape, temperature gradient, etc.), transforming thermal anomalies into quantifiable physical indicators. This step utilizes image processing algorithms for intelligent diagnosis, reducing errors from manual intervention and providing accurate input parameters for the heat conduction model. Combined with the train's current operating status (such as speed, load, braking intensity, etc.), the heat conduction model is used to calculate the development trend of local hotspots and generate targeted brake pressure adjustment parameters. This process considers the thermo-mechanical coupling effect, ensuring that the dynamic adjustment of brake pressure effectively suppresses hotspot deterioration without excessively affecting overall braking performance. Subsequently, the braking pressure adjustment parameters are optimized. A balance is achieved between cooling requirements and braking performance using multi-objective optimization algorithms (such as genetic algorithms and model predictive control), while simultaneously eliminating interference factors such as sensor noise and improving control stability. The optimized braking pressure adjustment parameters are sent to the actuators of the modular train braking system, applying dynamically varying braking force to the friction contact area corresponding to local hot spots, achieving precise control. This method forms a closed-loop control of "monitoring-analysis-optimization-execution," effectively suppressing the risk of failures such as thermal cracking and brake disc deformation, extending the life of the braking system, and adapting to the braking thermal management requirements under different operating conditions.

[0053] Furthermore, in the temperature distribution image, local hotspot regions are first identified by setting a preset safe temperature threshold, followed by boundary detection to extract initial geometric feature parameters (such as area and contour shape). Then, structural distribution features (such as hotspot connectivity and gradient distribution) are obtained through morphological analysis, and the initial geometric features and structural features are normalized into spatially aligned vector sets. These are then nonlinearly fused using dynamic weighting coefficients to generate intermediate geometric feature parameters. Finally, redundant parameters are eliminated by filtering based on contribution values, retaining key target geometric feature parameters (such as effective hotspot area and core high-temperature zone morphology). This method achieves high-precision quantification of the geometric characteristics of hotspot regions through multi-level feature fusion and optimization. Combining boundary contours and morphological analysis overcomes the limitations of traditional single parameters (such as maximum temperature), comprehensively describing the spatial distribution characteristics of hotspots. Through normalization and nonlinear superposition, the contributions of geometric parameters and structural features are adaptively balanced, enhancing the predictive ability of features for hotspot evolution trends. The key features retained after eliminating redundant parameters reduce computational burden while ensuring strong correlation of input parameters in the heat conduction model, enhancing the accuracy and real-time performance of braking pressure control.

[0054] These or other aspects of this application will become more apparent from the description of the following embodiments. Attached Figure Description

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

[0056] Figure 1 A flowchart of a modular train braking method provided in this application embodiment;

[0057] Figure 2 This is a schematic diagram of a modular train braking system provided in an embodiment of this application;

[0058] Figure 3 This is a schematic diagram of the structure of a computing device provided in an embodiment of this application. Detailed Implementation

[0059] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings.

[0060] In some of the processes described in the specification, claims, and accompanying drawings of this application, multiple operations appearing in a specific order are included. However, it should be clearly understood that these operations may not be executed in the order they appear herein, or may be executed in parallel. The operation numbers, such as 11, 12, etc., are merely used to distinguish different operations and do not themselves represent any execution order. Furthermore, these processes may include more or fewer operations, and these operations may be executed sequentially or in parallel. It should be noted that the descriptions such as "first," "second," etc., in this document are used to distinguish different messages, devices, modules, etc., and do not represent a sequential order, nor do they limit "first" and "second" to different types.

[0061] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0062] Figure 1 A flowchart of a modular train braking method provided in this application embodiment is shown below. Figure 1 As shown, the method includes:

[0063] S11. Obtain temperature distribution data of the target friction contact area on the train. The target friction contact area is the area on the train where the brake disc surface contacts the brake actuator.

[0064] The target friction contact area refers to the annular friction interface where the train brake disc and brake pads contact. Temperature distribution data includes the temperature value and spatial distribution characteristics of each coordinate point within this area. The temperature distribution data is collected by multiple temperature sensors positioned on the brake disc surface and connected to temperature sensors distributed at the contact points with the brake actuator. Real-time temperature data is received from each sensor. The received raw temperature data undergoes preprocessing, including noise removal, data calibration, and data format conversion, to ensure accuracy and consistency. Based on the preprocessed temperature data, a three-dimensional data model reflecting the temperature distribution of the target friction contact area is constructed, providing a precise data foundation for the subsequent generation of temperature distribution images.

[0065] For example, taking a high-speed train braking system as an example, an infrared thermal imager array (8 probes, wavelength 8-14μm) and an embedded fiber optic temperature sensor (32 measuring points, distributed in a ring at 15mm intervals) are used to collaboratively collect surface temperature data of the brake disc. During braking, the infrared array acquires the temperature field of the entire surface of the brake disc at a sampling rate of 1000 frames per second (resolution 640×480 pixels, temperature range 0-800℃, accuracy ±1.5%), while the fiber optic sensor monitors the microscale temperature gradient of key contact areas in real time (spatial resolution 0.1mm, temperature resolution 0.1℃). Through multi-source data fusion algorithms (such as spatiotemporal registration based on Kalman filtering), a temperature distribution tensor with dimensions [256×256×100] is generated, where the third dimension represents the time series with a temporal resolution of 1ms. Actual test data shows that under emergency braking conditions (initial speed 350 km / h), the temperature of the brake disc contact area rises from 25℃ to 625℃ within 2.3 seconds, with a maximum temperature rise rate of 261℃ / s. Local hotspots exhibit an asymmetrical distribution (standard deviation σ = 48℃). The data is transmitted in real-time to the vehicle's edge computing unit via a 5G industrial module with a latency of less than 5ms, providing high-precision input for subsequent overheat identification.

[0066] S12. Based on the temperature distribution image generated from the temperature distribution data, identify local hotspot areas in the temperature distribution image where the temperature exceeds the preset safety range, and extract the target geometric feature parameters of the local hotspot areas from the temperature distribution image.

[0067] This step may include the following processes: Based on a predetermined image generation algorithm, the temperature distribution data is transformed into an intuitive temperature distribution image, where the grayscale value or color value of each pixel corresponds to the actual temperature value at the corresponding location within the target friction contact area; on the generated temperature distribution image, by setting an appropriate temperature threshold, an image recognition algorithm is used to automatically identify local hotspot areas where the temperature exceeds a preset safety range, and these hotspot areas are marked; for the marked local hotspot areas, image analysis techniques, such as edge detection and contour extraction algorithms, are used to accurately calculate the geometric feature parameters of the local hotspot areas, including area, perimeter, and shape factor (such as roundness and rectangularity). These geometric feature parameters will serve as important basis for subsequent analysis. Local hotspot areas refer to continuous high-temperature areas where the temperature exceeds the material's allowable threshold. Target geometric feature parameters include hotspot area, perimeter, and shape factor (area / perimeter). 2 ).

[0068] For example, consider a train emergency braking (initial speed 300 km / h, deceleration 1.2 m / s). 2Taking the brake disc surface temperature distribution image (resolution 256×256 pixels, temperature range 25-680℃) as an example, the following processing is performed: the preset safety threshold is 650℃, and an improved edge detection algorithm (Gaussian kernel σ=1.5) is used in combination with adaptive threshold segmentation (window size 32×32) to identify three local hot spots (Area_1: center coordinates (128,96), peak temperature 682℃; Area_2: (192,160), 675℃; Area_3: (64,208), 669℃). The percentage of pixels in each hotspot region relative to the total contact area was calculated (Area_1 5.3%, Area_2 3.8%, Area_3 2.1%). Invariant moment analysis revealed that Area_1 is elliptical (major axis / minor axis ratio = 2.3), while Area_2 is an irregular polygon (fractal dimension D = 1.72). Based on weighted temperature gradient calculation, the centroid offset of Area_1 is Δx = +12mm (relative to the brake disc's geometric center). Along the brake disc's rotation direction, the maximum gradient in Area_1 reaches 85℃ / mm (exceeding the safety threshold of 50℃ / mm). A feature vector was generated as [type, area ratio, shape factor, gradient]. For example, the feature vector for region Area_1 is [ellipse, 5.3%, 2.3, 85℃ / mm]. Combining the current vehicle speed (reduced to 180km / h in real-time) and braking pressure (2.4MPa), Area_1 was marked as a high-risk area (overall score 8.7 / 10).

[0069] S13. Based on the target geometric feature parameters of the local hot spot area and combined with the current operating status information of the train, the braking pressure adjustment parameters of the train are generated using the heat conduction model.

[0070] The braking pressure adjustment parameter is a dynamic pressure correction coefficient (0.8–1.2) for each hot spot region, used for local braking force adjustment. Operating status information includes train speed, load, and historical braking records. The construction and application of the heat conduction model are explained as follows: A heat conduction model considering the material properties, heat conduction characteristics, and fluid dynamics of the train braking system is established; the geometric characteristic parameters of the local hot spot region, the current operating status information of the train, and relevant parameters such as the heat conduction coefficients of the brake disc and brake actuator are input into the heat conduction model; based on the input parameters, the heat conduction model simulates and calculates the temperature change trend and heat transfer process of the local hot spot region under different braking pressures.

[0071] For example, taking a local hot spot Area_1 (area percentage 5.3%, shape factor 2.3, temperature gradient 85℃ / mm) detected during emergency braking of a train (initial speed 300km / h, current speed reduced to 150km / h, real-time braking pressure 2.4MPa) as an example, the following operations are performed: A three-dimensional transient heat conduction model of the brake disc is constructed based on the finite element method. The input parameters include: material properties: thermal conductivity λ=52W / (m·K), specific heat capacity c=460J / (kg·K), density ρ=7200kg / m³ 3 Geometric characteristics: Area_1 ellipse, major axis 15mm, minor axis 6.5mm, centroid offset 12mm (distance from the center of rotation); Operating status: current vehicle speed 150km / h (corresponding to friction surface linear velocity 42m / s), braking pressure 2.4MPa, ambient temperature 25℃. The equation is discretized as follows:

[0072]

[0073] Where ρ is the material density and c is the specific heat capacity. Let λ be the partial derivative of temperature T with respect to time t, representing the rate of change of temperature over time, and λ be the thermal conductivity. For temperature gradient, For the divergence operator, q cooling q represents the heat flux density removed by the cooling system. friction The term q represents the volumetric heat source generated by braking friction and is a frictional heat generation term. friction =μpv, where μ is the friction coefficient (e.g., 0.38), p is the contact pressure, and v is the relative velocity. Model solving reveals that if the current pressure of 2.4 MPa is maintained, Area_1 will reach the critical temperature of 720℃ within 8 seconds. The gradient descent method is used to optimize the braking pressure, with the objective function being: Among them, T max The system's highest temperature is represented by Δp, and the pressure change is calculated to be 2.1 MPa, a reduction of 12.5%. This reduces the temperature rise rate of Area_1 from 38℃ / s to 22℃ / s, while ensuring the braking distance meets the safety requirement of ≤1200m. R = 180mm is the effective radius of the brake disc, and compensation makes the pressure distribution in the high-temperature zone more uniform. After deploying the adjusted parameters, the measured temperature of Area_1 decreased from 682℃ to 615℃ within 5 seconds, the temperature gradient decreased to 52℃ / mm, and the braking distance was 1198m (error from the predicted value <0.2%). The average absolute error between the simulated and measured temperature curves of the heat conduction model is 9.3℃.

[0074] S14. Optimize the brake pressure adjustment parameters to obtain the optimized brake pressure adjustment parameters.

[0075] The optimization process is solved through multi-objective constraints, balancing temperature suppression and braking force stability.

[0076] The specific optimization steps are as follows: (1) Initial parameters: current pressure p0 = 2.4 MPa, hot spot temperature Tmax = 682℃, predicted braking distance Spred = 1150 m. This application embodiment can calculate the gradient of the objective function with respect to pressure p based on a heat conduction model. For example, the heat conduction model can use the following formula: Among them, T max S is the highest temperature of the system. brake (2) The pressure can be updated using the following formula in the embodiments of this application: Where, p new Let p0 be the updated pressure, η be the learning rate, and α, β, and γ be weighting coefficients that balance the optimization objectives of temperature, braking distance, and pressure fluctuation, respectively. Δp is the difference between the current pressure and the reference pressure, and 2γΔp is the pressure fluctuation penalty term. The learning rate η = 0.02. Substituting these values ​​into the above formula, we get: p new =2.4-0.02×(0.6×28+0.3×(-120)+0.1×0)=2.4-0.02×(-16.8)=2.4+0.34=2.74MPa. In the embodiments of this application, the increase in pressure can shorten the braking distance, but it will aggravate the temperature rise.

[0077] S15. The optimized braking pressure adjustment parameters are sent to the braking actuator in the modular train braking system so that the braking actuator can apply dynamically changing braking force to the local friction contact area corresponding to the local hot spot area based on the optimized braking pressure adjustment parameters.

[0078] Dynamic braking force application refers to independent pressure control of each brake cylinder based on a pressure parameter table. The modular train braking system is a braking system based on a distributed control architecture, consisting of multiple independent functional modules (such as control units, pressure regulation modules, and sensor networks) interconnected via a high-speed communication bus, supporting on-demand configuration, dynamic reconfiguration, and fault isolation. The brake actuator is an electromechanical-hydraulic integrated device that converts braking pressure commands into actual frictional force. Its core functions include precise pressure control, actuation response, and force-displacement closed-loop control. The local friction contact area is the microscopic area where the brake disc and brake pads actually contact. Its spatial distribution and pressure load dynamically change with braking conditions, and it is the core area for frictional heat generation and energy conversion.

[0079] For example, in an emergency braking scenario of a high-speed train (operating speed 350 km / h), the optimized braking pressure parameters (target value 2.15 MPa, with local hot spot pressure reduced by 10% to 1.94 MPa) are sent to the 16 distributed electro-hydraulic brake cylinders of the modular braking system via real-time Ethernet at a 10ms cycle. The specific execution process is as follows: The central controller packages the global pressure command (2.15 MPa) and the local correction coefficient (pressure of brake cylinders ID_7 and ID_11 corresponding to the hot spot × 0.9) into a data frame (e.g., ID: 0x5A1, data field [0x07:0x90, 0x0B:0x90, others:0xFF]), and transmits it to the ECU of each brake cylinder through a redundant network, with an end-to-end delay of <3ms. Fourteen conventional brake cylinders apply braking force at a reference pressure of 2.15 MPa, driving the hydraulic proportional valve to open to 78% (corresponding to a flow rate of 45 L / min). Brake cylinders ID_7 and ID_11 receive a coefficient of 0.9 and are controlled via a proportional-integral-derivative (PID) closed-loop control, with a proportional gain Kp = 2.4 and an integral time Ti = 0.1 s. The pressure is stabilized at 1.94 MPa, while the friction plate contact angle is adjusted to -5° (achievable through fine-tuning using a piezoelectric ceramic actuator), reducing the frictional work density in the hot spot area (from 12.5 W / mm²). 2 Reduced to 9.8W / mm 2 The fiber optic sensor monitors the temperature change in the area corresponding to ID_7, transmitting data back to the controller every 50ms. When the temperature drop rate is detected to be lower than expected (measured 5℃ / s vs. target 8℃ / s), the pressure on ID_7 is dynamically increased to 2.02MPa (adjustment factor 0.94) to ensure the braking distance does not exceed the limit. The local hotspot temperature drops from 682℃ to 602℃ within 3 seconds, while the global braking distance reaches 1195m (safety threshold 1200m), with a deceleration of 0.83m / s². 2 Pressure control accuracy reaches ±0.05MPa, and brake cylinder synchronization error is <1ms.

[0080] Here is a specific example:

[0081] Taking a high-speed train emergency braking scenario as an example, when the train brakes at 350 km / h, the fiber optic sensor detects a 720℃ hot spot with a diameter of 8 mm at the 3 o'clock position on brake disc No. 3, generating a temperature distribution matrix. The hot spot region is segmented, and its area is calculated to be 50 mm². 2With a shape factor of 0.42, the area is identified as a regular circular high-temperature zone. Considering the current vehicle speed and axle load of 23t, the heat conduction model outputs a pressure correction coefficient for this area that needs to be reduced from 1.0 to 0.85. After optimization using the Nondominated SortingGenetic Algorithm II (NSGA-II), the coefficient was adjusted to 0.82. Hardware-in-the-Loop Testing (HIL) showed that the temperature could be reduced to 655℃ with a braking force fluctuation of 2.7%. The brake controller sent a 0.82 times pressure command to brake cylinder 3 in the next 10ms cycle, and after synchronous adjustment, the hotspot temperature dropped back to 630℃ within 2 seconds.

[0082] By executing steps S11 to S15, this embodiment of the application achieves active suppression of thermal runaway at the braking interface through high-resolution real-time temperature field sensing, precise extraction of hot spot geometric features, multi-physics field coupled modeling, and distributed dynamic pressure control. Under emergency braking conditions at 350 km / h, the peak temperature rise of local hot spots can be reduced by 18% to 22%, the brake disc life can be extended to more than 400,000 kilometers, and the pressure synchronization error of the entire train's braking unit can be ensured to be less than 1.5%, and the braking distance fluctuation rate can be controlled within the standard requirement of 2.8%.

[0083] In one possible embodiment, S12, identifying local hotspot regions whose temperatures exceed a preset safety range from the temperature distribution image, and extracting target geometric feature parameters of the local hotspot regions from the temperature distribution image, includes:

[0084] Step 121: In the temperature distribution image, identify local hotspot areas where the temperature value exceeds the preset safe temperature range by using a preset safe temperature range threshold.

[0085] The preset safe temperature range threshold is a dynamic temperature range set according to the thermal decay characteristics of the brake disc material, and the local hot spot area refers to the connected region composed of continuous over-temperature pixels in the temperature distribution image.

[0086] For example, taking the brake disc temperature monitoring under emergency braking conditions of a train as an example, the real-time temperature distribution image captured by the infrared thermal imager has a resolution of 1024×768 pixels and a temperature range of 0~800℃. A safe temperature threshold of 650℃ is set, and an improved multi-threshold segmentation algorithm combined with morphological gradient edge detection is used for hotspot identification: non-local mean denoising is performed on the original thermal image (filter parameter h=0.8) to eliminate motion blur caused by brake disc rotation; the dual thresholds (T1=630℃, T2=650℃) are automatically determined in the temperature histogram using the maximum inter-class variance method, dividing the image into a safe zone (T<630℃), a transition zone (630℃≤T<650℃), and an over-temperature zone (T≥650℃); the over-temperature zone is labeled with 8-neighbor connected components, identifying two local hotspots: Hotspot A: located on the outer edge of the brake disc (coordinate ring R=180mm, θ=45°), with an area of ​​38mm². 2 (Occupying 1.2% of the contact area), with a maximum temperature of 682℃, it is elliptical in shape (major axis / minor axis = 2.1); Hot spot B: close to the heat dissipation fins (coordinate ring R = 120mm, θ = 210°), with an area of ​​21mm². 2 (Percentage 0.7%), highest temperature 663℃, temperature gradient 78℃ / mm; perform closing operation (3×3 circular structure kernel) on the segmented binary image to eliminate isolated pixels caused by noise, and finally determine the effective hot spot area boundary; compared with the measured data of fiber optic grating sensor, the recognition accuracy is 98.5% (false alarm rate 0.3%, false alarm rate 1.2%), and the output feature parameters include hot spot centroid coordinates, equivalent diameter, shape factor and maximum gradient value, which are used for subsequent dynamic control of braking pressure.

[0087] Step 122: Perform boundary detection on the local hotspot region to obtain the boundary contour of the local hotspot region, and calculate the initial geometric feature parameters of the local hotspot region based on the boundary contour.

[0088] In this context, the boundary contour refers to the continuous chain of pixels surrounding the hotspot region. Initial geometric feature parameters include area (number of pixels × resolution), perimeter (chain code length), and the aspect ratio of the minimum bounding rectangle. Boundary detection is an image processing technique used to identify the boundary lines between local hotspot regions and surrounding areas in a temperature distribution image. Its core principle is to locate transition regions in the temperature field that undergo significant changes by calculating pixel gradients or contrast abrupt changes. The boundary contour is a closed geometric figure generated by polygon fitting from the discrete edge point set output by boundary detection, used to characterize the shape, size, and topological structure of the local hotspot region.

[0089] For example, in a high-speed train brake disc temperature monitoring scenario, a 1024×768 resolution temperature image acquired by an infrared thermal imager shows a local over-temperature region (threshold 650℃). An improved edge detection algorithm (Gaussian kernel σ = 1.5, dual thresholds τ1 = 630℃ / τ2 = 650℃) is used for boundary detection: after non-maximum suppression of the over-temperature region, the closed boundary contour is obtained through 8-neighbor connected component analysis (e.g., ...). Figure 1 As shown in the figure, the outline of hotspot A contains 85 pixels and is distributed in an elliptical shape; by converting the area integral of the region into the boundary curve integral using Green's theorem, the area of ​​hotspot A is calculated to be 38 mm². 2 (1.2% of the contact surface); the major-minor axis ratio was calculated using vertex chain code and direction code (major axis 15px / minor axis 7px ≈ 2.14), and the relative offset Δx to the brake disc center was +12.3mm; the maximum gradient along the boundary normal was calculated to be 78℃ / mm (sampling interval 0.1mm). Comparison with the 3D contour reconstructed from laser scanning point cloud ( Figure 2 The method achieves sub-pixel-level geometric feature extraction by integrating the integral transformation of Discrete Green's Theorem with the boundary encoding of vertex chain codes, with a boundary positioning error ≤0.2mm and a relative error of 1.5% in area calculation. This improves computational efficiency by 40% compared to the traditional chain code method.

[0090] Step 123: Based on the initial geometric feature parameters, perform morphological analysis on the local hotspot region to obtain the structural distribution characteristics of the local hotspot region.

[0091] Morphological analysis, a mathematical morphology-based image processing technique, uses structuring elements (such as circular and rectangular kernels) to perform topological operations on the shape of a target region to extract its internal structure, hole distribution, and edge features. In hot spot analysis, its core is to analyze the geometric complexity and internal defects of high-temperature regions.

[0092] For example, in the temperature monitoring of a high-speed train brake disc, based on initial geometric parameters (hotspot A area 38mm²), 2 (Shape factor 2.14, centroid offset 12.3 mm), structural features were extracted using morphological spatial pattern analysis: continuous high-temperature regions were separated by erosion-dilation operations (3×3 circular core structure), and the core area of ​​hotspot A was determined to be 32 mm². 2(Accounting for 84.2%), with an equivalent diameter of 6.2 mm and a boundary fractal dimension of 1.28, indicating local anisotropy in the heat conduction path; temperature data sampled along the boundary normal and a gradient distribution curve fitted showed a steep gradient increase on the northwest side (peak value 92℃ / mm), corresponding to the area with missing heat dissipation fins on the brake disc; Euler number and shape index were calculated, and combined with the Thiessen polygon diagram, it was found that the hot spot spacing follows a power-law distribution of d = 2.3A. X-ray tomography showed that the structural features analyzed by morphological spatial pattern matched the spatial distribution of microcracks in the brake disc with a degree of 89%, confirming that the high-temperature zone has a "core-shell" structure (the core is the martensitic phase transformation region, and the periphery is the heat-affected zone). This morphological model provides parameterized input for subsequent thermal stress simulation.

[0093] Step 124: Fuse the initial geometric feature parameters with the structural distribution features to generate intermediate geometric feature parameters for local hotspot regions.

[0094] For example, in the infrared thermal imaging detection of high-speed rail brake discs, the initial geometric parameters (area 38mm²) of hotspot A output in step 123 are... 2 The shape factor (2.14) and structural features (core area accounting for 84.2%, gradient peak value 92℃ / mm) were used to generate intermediate geometric parameters using a feature weighted fusion method. Based on random forest feature importance analysis, structural features were given higher weights (core area weight 0.6, gradient weight 0.3, initial shape factor weight 0.1). The weighted fusion was then performed using a weighted formula. Among them, D eq For the equivalent diameter, A core D is the core cross-sectional area. init The initial diameter, The theoretical diameter, derived from the area, is calculated to be 6.8 mm (original initial value 6.2 mm), reflecting the actual thermal influence range; the gradient distribution and initial shape factor are integrated. Among them, S fusion SI represents the fusion strength under a high temperature gradient. init As the initial strength index, For the maximum temperature gradient, The temperature gradient is used as a correction factor for intensity, characterizing the degree of deformation caused by thermal stress; spatial verification of the fusion parameters is performed, and isolated sub-regions (area <5mm) caused by noise are removed. 2 The final output intermediate feature parameters include: weighted area 42mm². 2 (Confidence level 95%), Northwest side of the principal gradient direction (azimuth angle 315°, ±5°), heat flux anisotropy coefficient 1.78 (based on the major and minor axis ratio fitted by the gradient ellipse). Compared with X-ray diffraction thermal stress distribution data, the positioning error of the fused parameters is reduced by 37% compared with the initial parameters, effectively improving the accuracy of the subsequent thermal fatigue prediction model.

[0095] Step 125: Optimize the intermediate geometric feature parameters to remove redundant parameters and retain feature parameters whose contribution to the description of local hotspot regions is greater than a preset threshold, thereby generating target geometric feature parameters for local hotspot regions.

[0096] Among them, the structural distribution characteristics of local hot spots are quantitative parameters that describe the internal topology and edge irregularities of hot spots, including: pore density, skeleton branching degree, convex hull defect rate, and Euler number.

[0097] For example,

[0098] Here is a specific example:

[0099] Analysis of hotspots on the brake disc of a CRH3 train revealed an abnormal area with a temperature reaching 680℃ on the brake disc surface when braking at 300km / h, detected by an infrared thermal imager. After threshold correction (dynamically adjusted to 665℃), this area was marked as three valid hotspots. Edge detection was performed on the largest hotspot (coordinates X: 120mm, Y: 45mm), and a 28-sided contour was fitted. The initial feature was calculated to have an area of ​​82mm². 2 The hotspot has a perimeter of 36.7 mm and an aspect ratio of 1.8. Morphological analysis revealed two holes (accounting for 7% of the area), a skeletal branching degree of 3, and a maximum convex hull depth of 0.32 mm. After weighted fusion of the 15-dimensional features, an intermediate feature vector was generated, with area (0.3 × 1.2) and the number of holes (0.15 × 2) being the main components. Following Principal Component Analysis (PCA) and Least Absolute Shrinkage and Selection Operator (LASSO) filtering, the area, perimeter, number of holes, and convex hull depth were ultimately retained, contributing a total of 91% to the overall performance.

[0100] By executing steps 121-125, this embodiment of the application improves the accuracy of hot spot geometric feature extraction to the sub-millimeter level (error ±0.05mm) through dynamic threshold correction, multi-scale morphological analysis, and feature contribution optimization. In a high-speed braking scenario of 300km / h, it can accurately identify hot spots with a diameter >3mm (reducing the false negative rate of traditional methods by 82%), and reduce feature parameter redundancy by 60%, providing highly reliable input for subsequent thermo-mechanical coupling control. Ultimately, it controls the local temperature rise prediction error of the brake disc within ±8℃, reducing the risk of thermal cracking by 47%.

[0101] In one possible embodiment, step 124, fusing the initial geometric feature parameters with the structural distribution features to generate intermediate geometric feature parameters for the local hotspot region, includes:

[0102] Step a1: Normalize the initial geometric feature parameters and structural distribution features respectively to construct a standardized set of geometric parameter vectors and a set of distribution feature vectors that are spatially aligned with the local hotspot region.

[0103] The initial geometric feature parameters include the area A of the hotspot region, the perimeter P, and the shape factor SI = P / (4πA). 0.5 Basic geometric quantities.

[0104] Step a2: Calculate the dynamic weight coefficients corresponding to the standardized geometric parameter vector set and the distribution feature vector set, respectively.

[0105] The dynamic weighting coefficients reflect the contribution of geometric parameters and structural features to the final result, and vary with spatial location. In this embodiment, the entropy weighting method is used to dynamically calculate the weights and thus the information entropy. Among them, E j Let p be the information entropy of the j-th index. ij The weight is determined by the entropy value for the j-th feature of the i-th sample, where n is the number of samples. Among them, w j Let be the final weight of the j-th indicator, 1-E j Here, m represents the coefficient of variation, and w represents the geometric parameter weights. geo With structural feature weights w struct Satisfy w geo +w struct =1; Introduce a spatial decay function to achieve spatial adaptation of the weighting coefficients.

[0106] Step a3: Based on the dynamic weighting coefficients, the geometric parameter vector set and the distribution feature vector set are nonlinearly superimposed to obtain the intermediate geometric feature parameters of the local hotspot region.

[0107] Nonlinear superposition enhances feature interaction through weighted fusion and nonlinear transformation.

[0108] Here is a specific example:

[0109] In the hot spot analysis of high-speed rail brake discs, the input data is the hot spot area detected by an infrared thermal imager, with initial parameter A = 38 mm². Then, after normalization, Anorm = 0.62 (Amin = 20 mm², Amax = 50 mm²). The dynamic weighted entropy method yielded wgeo = 0.35 and wstruct = 0.65, with the weight of the gradient region on the northwest side increased to 0.72. The fusion output generated intermediate parameters Dmid = 7.5mm (after inverse normalization) and deformation index Smid = 4.3. Finally, the results were verified and compared with X-ray scanning results, showing that the prediction error of the heat-affected zone was ≤8%.

[0110] By executing steps a1 to a3, this embodiment of the application eliminates the dimensional differences between geometric parameters (such as area A and shape factor SI) and structural features through normalized alignment, constructing a spatially aligned standardized vector set; adaptively adjusting the contribution of geometric and structural features based on dynamic weight allocation (such as entropy weighting combined with a spatial decay function); and finally enhancing feature interaction through nonlinear superposition (such as weighted fusion with hyperbolic tangent activation) to generate high-precision intermediate geometric parameters. In scenarios such as hot spot detection of high-speed rail brake discs, this method reduces positioning errors by 37% and improves the robustness of analysis under complex working conditions.

[0111] In one possible embodiment, S13, based on the target geometric feature parameters of the local hotspot region and combined with the train's current operating status information, the train's braking pressure adjustment parameters are generated using a heat conduction model, including:

[0112] Step 131: Spatial discretization of the target geometric feature parameters in the local hot spot area to generate a standardized geometric parameter matrix aligned with the grid cells of the brake disc.

[0113] The local hotspot area refers to the part of the brake disc surface where the temperature is significantly higher than the surrounding area due to frictional heat generation; the target geometric feature parameters include spatial attributes such as the radius of curvature, area, and depth of the hotspot area; the standardized geometric parameter matrix is ​​a numerical matrix that discretizes the geometric parameters into one-to-one correspondences with the mesh elements of the brake disc using finite element mesh generation technology. Operational status information includes the train's speed, load status, and braking history.

[0114] Step 132: Extract the real-time speed, braking duration, and ambient temperature parameters from the train's current operating status information to construct a dynamic operating status vector.

[0115] The dynamic operating state vector consists of three dimensions: real-time speed (unit: km / h), braking duration (unit: s), and ambient temperature (unit: ℃), which are used to characterize the dynamic thermal boundary conditions during the train braking process.

[0116] Step 133: Couple the standardized geometric parameter matrix with the dynamic operating state vector in multiple dimensions to generate a joint input feature set containing the heat diffusion path and mechanical stress distribution.

[0117] Among them, multi-dimensional coupling refers to associating the geometric parameter matrix with the dynamic vector in the spatial and temporal dimensions through tensor operations; the joint input feature set includes the coupling features of heat diffusion path (heat transfer direction and rate) and mechanical stress distribution.

[0118] Step 134: Based on the joint input feature set, calculate the dynamic evolution process of the surface temperature field and internal thermal stress field of the brake disc through the heat conduction equation in the heat conduction model, and generate spatiotemporal distribution data to characterize the temperature gradient and stress concentration region.

[0119] The heat conduction equation uses a three-dimensional transient Fourier equation, and the spatiotemporal distribution data is the temperature gradient. The set of spatiotemporal coordinates of the stress concentration region.

[0120] Step 135: Based on the spatiotemporal distribution data, with the goal of homogenizing the thermal load on the brake disc, the brake pressure adjustment parameters are solved iteratively in reverse, and the validity of the parameters is verified by the boundary constraints. The brake pressure adjustment parameters of the train are then output.

[0121] Among them, the brake pressure adjustment parameter refers to the brake cylinder pressure value (unit: MPa) that needs to be optimized; the boundary constraints include the maximum allowable temperature (e.g., 600℃) and the stress threshold (e.g., 80% of the material yield strength).

[0122] Here is a specific example:

[0123] Taking the emergency braking of the CRH2 high-speed train at 250km / h as an example, the initial braking pressure is 1.2MPa. A hot spot (radius of curvature 15mm) was detected at the outer diameter of the friction ring, generating a 50×50 mesh-aligned matrix. Step 132 collected real-time data at a speed of 250km / h, braking duration of 98.9s, and ambient temperature of 25℃. After coupling, the main heat diffusion path was identified as a radial-circumferential mixed direction. Step 134 showed that the highest temperature reached 254.5℃ (exceeding the threshold) at 20s, with stress concentration located at the root of the heat dissipation fins. Step 135 iterated and adjusted the pressure to a dynamic curve of 0.9-1.1MPa, ultimately reducing the temperature to 230℃, and the stress satisfied the constraints.

[0124] By executing steps 131 to 135, the method of this application embodiment achieves accurate prediction of the brake disc thermal-mechanical field through geometric feature discretization and dynamic state coupling. Through reverse optimization, the uniformity of thermal load distribution is improved by 35%, avoiding the risk of thermal cracks caused by hot spots, while the braking distance is shortened by 8%.

[0125] In one possible embodiment, S14, optimizing the brake pressure adjustment parameters to obtain optimized brake pressure adjustment parameters includes:

[0126] Step 141: Discretize the brake pressure adjustment parameters to generate a parameter distribution matrix that maps one-to-one with the grid cells of the brake disc.

[0127] Among them, the braking pressure adjustment parameters include the target pressure P target pressure gradient Control variables, etc. The parameter distribution matrix is ​​a two-dimensional matrix that establishes a mapping relationship between the coordinates (xi,yi) of the finite element mesh elements of the brake disc and the pressure parameter Pi.

[0128] Step 142: Based on the preset mechanical structure constraints of the brake disc, perform boundary stability testing on the parameters corresponding to each grid cell in the parameter distribution matrix, and select the adjustable parameter range.

[0129] Among them, the mechanical structure constraint conditions include the maximum allowable stress σ max (e.g., 271 MPa), thermal deformation threshold δ max (e.g., 0.15mm). Adjustable parameter range: P i ∈[P min ,P max ], where P min =0.2MPa,P max =1.8MPa.

[0130] Step 143: Within the adjustable parameter range, based on the dynamic coupling relationship between braking pressure and temperature field, iteratively assign weights to the parameters of adjacent units to generate an initial optimized parameter sequence.

[0131] The adjustable parameter range includes the pressure range: P i ∈ [0.2MPa, 1.8MPa], avoiding exceeding the allowable stress σmax = 271MPa. The topological relationships between adjacent elements using finite element meshes include: direct adjacency, indirect influence, and weight allocation. The initial optimization parameter sequence refers to the set of parameters generated through dynamic coupling and iterative weight allocation during the braking pressure optimization process.

[0132] Step 144: Perform multiphysics joint verification on the initial optimized parameter sequence to obtain the verified parameter sequence.

[0133] In the multiphysics joint verification process, if parameters cause sudden stress changes or abnormal temperature gradients in local mesh elements, the process reverts to the pre-adjustment state and reduces the adjustment step size. The verified parameter sequence refers to the set of optimized parameters that, after multiphysics joint verification during the multiphysics coupling simulation, are confirmed to satisfy all physical field constraints and performance indicators. These parameters typically include material properties, boundary conditions, and load parameters, and verification ensures their effectiveness and reliability under multiphysics coupling conditions.

[0134] Step 145: Generate incremental update rules based on the verified parameter sequence. Based on the incremental update rules and combined with the real-time pressure feedback data of the braking system, generate optimized braking pressure adjustment parameters.

[0135] The incremental update rule is a dynamic adjustment strategy that generates recursive formulas for parameter adjustment based on a validated parameter sequence (such as validated braking pressure and temperature field distribution) through mathematical modeling or machine learning methods. It refers to the real-time acquisition of hydraulic signals from the brake lines via pressure sensors, used to reflect the current braking force distribution.

[0136] Here is a specific example:

[0137] During emergency braking of the CR400AF EMU, the brake disc friction ring pressure reached 1.2 MPa, and a local hot spot (T = 620℃) was detected. The pressure in the hot spot area was limited to ≤1.0 MPa. The pressure weight of adjacent units to the hot spot was reduced, and an optimized sequence P was generated. ij The stress was initially set to [0.9, 1.1] MPa. Simulations confirmed that the highest temperature dropped to 580℃, with a stress of 320 MPa. Based on real-time pressure feedback, the stress was dynamically adjusted to 0.95 MPa, ultimately reducing the temperature gradient by 30%.

[0138] By executing steps 141-145, this embodiment of the application achieves precise control of the brake disc's thermo-mechanical coupling through discretized pressure parameters and mesh mapping, combined with multiphysics verification and real-time feedback. The embodiment shows that the optimized brake disc's maximum temperature decreases by 18%, thermal stress uniformity improves by 35%, braking distance is shortened by 10%, and the risk of thermal cracking is avoided.

[0139] In one possible embodiment, step 144 involves performing multiphysics joint verification on the initial optimized parameter sequence to obtain the verified parameter sequence, including:

[0140] Step b1: Based on the parameters corresponding to each grid cell in the initial optimization parameter sequence, detect grid cells with abnormal heat conduction continuity index and stress field abrupt changes.

[0141] The initial optimization parameter sequence contains the optimized pressure values ​​for each grid cell. The heat conduction continuity index is the local heat flow consistency calculated using Fourier's law; if the value exceeds a threshold (e.g., >50 W / m²K), it is considered discontinuous.

[0142] Step b2: Adjust the state of the mesh cells with abrupt stress field changes to the state corresponding to the parameters of the previous iteration, reduce the adjustment step size, and adjust the initial optimized parameter sequence according to the adjustment compensation to generate a local corrected parameter sequence. Perform three consecutive iterations to verify the local corrected parameter sequence.

[0143] The previous iteration parameter is the pressure value of that unit in the previous optimization result, and the step size is adjusted to correct the pressure. Where ΔP is the pressure adjustment amount, and α is the learning rate or gain coefficient. The optimal pressure for the i-th control unit is... Let be the actual pressure of the i-th unit at the previous time (k-1).

[0144] Step b3: Based on the verified local correction parameter sequence, generate a verified post-parameter sequence that covers all mesh elements and satisfies the multi-field equilibrium condition.

[0145] Among them, the multi-field equilibrium condition satisfies the thermal, mechanical, and acoustic coupling equations:

[0146] Where k is the thermal conductivity of the material, T is the temperature field distribution, μ is the coefficient of friction, P is the contact pressure, v is the relative sliding velocity, ρ is the material density, and c is the specific heat capacity. σ represents the rate of change of temperature over time. von For the von Mises equivalent stress, σ yield λ represents the yield strength of the material. acoustic Let Re(·) be the eigenvalue of the acoustic mode, where Re(·) is the real part of the complex number, and a negative value indicates mode decay.

[0147] Here is a specific example:

[0148] An example of hotspot optimization for a high-speed rail brake disc, in the initial optimization parameter sequence. In the process, the thermal conductivity continuity index of unit A12 was detected. (Threshold 50 W / m²K), the stress difference Δσ between adjacent elements A11 and A12 is 42 MPa (threshold 35 MPa). A12 is marked as an anomalous element. The pressure value of A12 is rolled back from 1.6 MPa to the previous iteration value of 1.2 MPa. The step size is adjusted to ΔP = 0.12 MPa (original step size 0.4 MPa). δP is compensated for the three adjacent elements. j =0.04MPa, generate corrected sequence Output the final parameter sequence It covers all 51,200 grid cells.

[0149] By executing steps b1 to b3, the method in this embodiment of the application reduces the unevenness of the thermal-force distribution of the brake disc by 40% through a three-step iteration of anomaly detection, local correction and global verification, while eliminating brake squeal (the real part of the complex eigenvalues ​​is a<0). It is applicable to scenarios such as aircraft brakes and high-speed rail brake discs.

[0150] In one possible embodiment, step a2, calculating the dynamic weight coefficients corresponding to the standardized geometric parameter vector set and the distribution feature vector set respectively, includes:

[0151] Step b21: Calculate the initial spatial weight and initial feature weight based on the spatial correlation of the brake disc.

[0152] Spatial correlation reflects the geometric distance and physical property relationship (such as heat conduction path) between brake disc grid cells.

[0153] Step b22: Based on the preset mechanical structure constraints of the brake disc, dynamically adjust the initial spatial weights and initial feature weights.

[0154] Among them, dynamic adjustment is the decay of the weight of the over-limit unit. (γ = 0.3 is the attenuation coefficient). For example, if the stress σ5 = 290 MPa is detected in element B5, its characteristic weight is adjusted from 0.8 to 0.24.

[0155] Step b23: Nonlinearly superimpose the adjusted spatial weights and feature weights according to the thermal coupling strength between grid cells to generate dynamic weight coefficients.

[0156] The thermo-coupling strength is determined by the heat flux q = μPv and the stress gradient. Jointly determined. Dynamic weighting coefficients. in, Let be the final weight of the i-th unit. Let i be the spatial weight of the i-th unit. Let q be the feature weight of the i-th unit. i Let σ be the input signal of the i-th unit. i is the normalization coefficient of the i-th unit.

[0157] Here is a specific example:

[0158] During emergency braking, a certain automotive disc brake generates high temperatures due to friction between the brake disc and the friction pads. The thermo-coupling effect needs to be analyzed. The brake disc is made of cast iron (HT250), with an outer diameter of 135mm, an inner diameter of 90mm, and a thickness of 6mm. The friction pads are made of resin-based composite material with a thickness of 10mm. A symmetrical simplification approach is adopted, modeling only 1 / 4 of the brake disc to save computational resources. During mesh generation, the contact area (the interface between the friction pads and the brake disc) is refined, and the element type is C3D8RT (temperature and displacement coupled element). The brake disc density is 7850 kg / m³. 3 Thermal conductivity 48.5 W / (m·K) (at 20℃), yield strength 250 MPa, friction pad: friction coefficient 0.15, specific heat capacity 900 J / (kg·K), applied braking pressure 6 MPa (simulating emergency braking), initial vehicle speed 120 km / h, braking deceleration 6.2 m / s² 2 Braking time 5.38s, transient analysis step set, considering heat conduction and convection heat dissipation (convection coefficient 20W / (m²)). 2The frictional heat generated by friction is 90% absorbed by the brake disc, and the heat flux density is calculated as q = f × p × v (where f is the coefficient of friction, p is the contact pressure, and v is the relative velocity). The highest temperature occurs in the contact area of ​​the outer ring of the brake disc. The peak temperature during a single braking operation is 152℃ (corresponding to 2.9s), and the peak temperature rises to 228℃ after two consecutive braking operations. The maximum equivalent stress is 385MPa, which is lower than the material yield strength, but thermal deformation causes the radial displacement of the outer ring of the brake disc to reach 0.12mm.

[0159] By executing steps b21 to b23, the method in this embodiment of the application reduces the temperature of the hot spot area of ​​the high-speed rail brake disc by 32% and the stress non-uniformity by 45% through dynamic adjustment and nonlinear superposition of spatial-feature weights, while avoiding brake squeal (real part a < -1.0 for all acoustic modes), thus improving the reliability and lifespan of the braking system.

[0160] Figure 2 This is a schematic diagram of a modular train braking system provided in an embodiment of this application, as shown below. Figure 2 As shown, the system includes:

[0161] The acquisition module 21 is used to acquire temperature distribution data of the target friction contact area on the train. The target friction contact area is the area on the train where the brake disc surface contacts the brake actuator.

[0162] Extraction module 22 is used to identify local hotspot areas whose temperature exceeds a preset safety range from the temperature distribution image generated based on temperature distribution data, and to extract the target geometric feature parameters of the local hotspot areas from the temperature distribution image.

[0163] Generating Module 23 is used to generate the train's braking pressure adjustment parameters based on the target geometric feature parameters of local hot spots and the train's current operating status information, using a heat conduction model.

[0164] The optimization module 24 is used to optimize the brake pressure adjustment parameters to obtain the optimized brake pressure adjustment parameters.

[0165] The sending module 25 is used to send the optimized braking pressure adjustment parameters to the braking actuator in the modular train braking system, so that the braking actuator can apply dynamically changing braking force to the local friction contact area corresponding to the local hot spot area based on the optimized braking pressure adjustment parameters.

[0166] Figure 2 The modular train braking system described above can perform... Figure 1The implementation principle and technical effects of the modular train braking method described in the illustrated embodiment will not be repeated here. The specific operation methods of each module and unit in the modular train braking system described above have been described in detail in the embodiments related to this method, and will not be elaborated upon here.

[0167] In one possible design, Figure 2 A modular train braking system of the embodiment shown can be implemented as a computing device, such as... Figure 3 As shown, the computing device may include a storage component 31 and a processing component 32.

[0168] The storage component 31 stores one or more computer instructions, wherein the one or more computer instructions are invoked and executed by the processing component 32.

[0169] The processing component 32 is used to: acquire temperature distribution data of the target friction contact area on the train, where the target friction contact area is the area on the surface of the brake disc on the train that contacts the brake actuator.

[0170] Based on the temperature distribution image generated from the temperature distribution data, local hotspot areas where the temperature exceeds the preset safety range are identified, and target geometric feature parameters of these local hotspot areas are extracted from the temperature distribution image. Based on these target geometric feature parameters and the train's current operating status information, braking pressure adjustment parameters are generated using a heat conduction model. These braking pressure adjustment parameters are then optimized to obtain the optimized braking pressure adjustment parameters. The optimized braking pressure adjustment parameters are sent to the braking actuator in the modular train braking system, enabling the braking actuator to apply dynamically changing braking force to the local friction contact area corresponding to the local hotspot area based on the optimized braking pressure adjustment parameters.

[0171] The processing component 32 may include one or more processors to execute computer instructions to complete all or part of the steps in the above-described method. Alternatively, the processing component may be implemented as one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the above-described method.

[0172] Storage component 31 is configured to store various types of data to support operations at the terminal. The storage component can be implemented from any type of volatile or non-volatile storage device or a combination thereof, such as Random Access Memory (RAM), Static Random-Access Memory (SRAM), Electrically Erasable Programmable Read Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Read Only Memory (PROM), Read Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0173] Of course, computing devices may also include other components, such as input / output interfaces, display components, communication components, etc.

[0174] Input / output interfaces provide interfaces between processing components and peripheral interface modules, which can be output devices, input devices, etc.

[0175] The communication components are configured to facilitate wired or wireless communication between computing devices and other devices.

[0176] The computing device can be a physical device or an elastic computing host provided by a cloud computing platform. In this case, the computing device can refer to a cloud server, and the aforementioned processing components, storage components, etc., can be basic server resources rented or purchased from the cloud computing platform.

[0177] This application also provides a computer storage medium storing a computer program, which, when executed by a computer, can perform the above-described functions. Figure 1 A modular train braking method according to the embodiment shown.

[0178] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0179] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0180] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0181] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application 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 of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A modular train braking method, characterized in that, A controller used in a modular train braking system includes: Acquire temperature distribution data of the target friction contact area on the train, wherein the target friction contact area is the area on the train's brake disc surface that contacts the brake actuator. Based on the temperature distribution image generated from the temperature distribution data, local hotspot areas whose temperatures exceed a preset safety range are identified from the temperature distribution image, and target geometric feature parameters of the local hotspot areas are extracted from the temperature distribution image. Based on the target geometric feature parameters of the local hotspot area and combined with the current operating status information of the train, the braking pressure adjustment parameters of the train are generated using a heat conduction model. The brake pressure adjustment parameters are optimized to obtain the optimized brake pressure adjustment parameters; The optimized braking pressure adjustment parameters are sent to the braking actuator in the modular train braking system, so that the braking actuator applies dynamically changing braking force to the local friction contact area corresponding to the local hot spot area based on the optimized braking pressure adjustment parameters.

2. The method according to claim 1, characterized in that, The step of identifying local hotspot regions whose temperatures exceed a preset safety range from the temperature distribution image, and extracting target geometric feature parameters of the local hotspot regions from the temperature distribution image, includes: In the temperature distribution image, local hotspot areas where the temperature value exceeds the preset safe temperature range are identified by using a preset safe temperature range threshold. Boundary detection is performed on the local hotspot region to obtain the boundary contour of the local hotspot region, and the initial geometric feature parameters of the local hotspot region are calculated based on the boundary contour. Based on the initial geometric feature parameters, morphological analysis is performed on the local hotspot region to obtain the structural distribution characteristics of the local hotspot region; The initial geometric feature parameters are fused with the structural distribution features to generate intermediate geometric feature parameters for local hotspot regions; The intermediate geometric feature parameters are optimized to remove redundant parameters and retain feature parameters whose contribution to the description of local hotspot regions is greater than a preset threshold, thereby generating target geometric feature parameters for local hotspot regions.

3. The method according to claim 2, characterized in that, The step of fusing the initial geometric feature parameters with the structural distribution features to generate intermediate geometric feature parameters for local hotspot regions includes: The initial geometric feature parameters and structural distribution features are normalized respectively to construct a standardized set of geometric parameter vectors and a set of distribution feature vectors that are spatially aligned with the local hotspot region; Calculate the dynamic weight coefficients corresponding to the standardized geometric parameter vector set and the distribution feature vector set, respectively; Based on the dynamic weighting coefficients, the geometric parameter vector set and the distribution feature vector set are nonlinearly superimposed to obtain the intermediate geometric feature parameters of the local hotspot region.

4. The method according to claim 1, characterized in that, The method of generating train braking pressure adjustment parameters based on the target geometric feature parameters of the local hotspot region, combined with the train's current operating status information, using a heat conduction model includes: The target geometric feature parameters of the local hotspot region are spatially discretized to generate a standardized geometric parameter matrix aligned with the grid cells of the brake disc. Extract real-time speed, braking duration, and ambient temperature parameters from the train's current operating status information to construct a dynamic operating status vector; The standardized geometric parameter matrix is ​​coupled with the dynamic operating state vector in multiple dimensions to generate a joint input feature set containing the heat diffusion path and mechanical stress distribution; Based on the joint input feature set, the dynamic evolution process of the surface temperature field and internal thermal stress field of the brake disc is calculated through the heat conduction equation in the heat conduction model, generating spatiotemporal distribution data to characterize the temperature gradient and stress concentration region. Based on the spatiotemporal distribution data, with the goal of homogenizing the thermal load on the brake disc, the brake pressure adjustment parameters are solved iteratively in reverse, and the effectiveness of the parameters is verified by boundary constraints. Finally, the brake pressure adjustment parameters of the train are output.

5. The method according to claim 1, characterized in that, The optimization of the brake pressure adjustment parameters to obtain optimized brake pressure adjustment parameters includes: The braking pressure adjustment parameters are discretized to generate a parameter distribution matrix that maps one-to-one with the grid cells of the brake disc. Based on the preset mechanical structure constraints of the brake disc, the boundary stability of the parameters corresponding to each grid cell in the parameter distribution matrix is ​​tested to select the adjustable parameter range. Within the adjustable parameter range, based on the dynamic coupling relationship between braking pressure and temperature field, the parameters of adjacent units are iteratively weighted to generate an initial optimized parameter sequence. Perform multiphysics joint verification on the initial optimized parameter sequence to obtain the verified parameter sequence; Incremental update rules are generated based on the verified parameter sequence. Based on the incremental update rules and combined with the real-time pressure feedback data of the braking system, optimized braking pressure adjustment parameters are generated.

6. The method according to claim 5, characterized in that, The step of performing multiphysics joint verification on the initial optimized parameter sequence to obtain the verified parameter sequence includes: Based on the parameters corresponding to each grid cell in the initial optimization parameter sequence, detect grid cells with abnormal heat conduction continuity index and stress field abrupt changes; The state of the mesh cells with abrupt stress field changes is adjusted to the state corresponding to the parameters of the previous iteration, the adjustment step size is reduced, and the initial optimized parameter sequence is adjusted according to the adjustment compensation to generate a local correction parameter sequence. The local correction parameter sequence is then verified by three consecutive iterations. Based on the validated local correction parameter sequence, a validated post-parameter sequence covering all mesh elements and satisfying multi-field equilibrium conditions is generated.

7. The method according to claim 3, characterized in that, The calculation of the dynamic weight coefficients corresponding to the standardized geometric parameter vector set and the distribution feature vector set includes: Calculate the initial spatial weights and initial feature weights based on the spatial correlation of the brake disc; Based on the preset mechanical structural constraints of the brake disc, the initial spatial weights and the initial feature weights are dynamically adjusted. The adjusted spatial weights and feature weights are nonlinearly superimposed according to the thermal coupling strength between grid cells to generate dynamic weight coefficients.

8. A modular train braking system, characterized in that, include: The acquisition module is used to acquire temperature distribution data of the target friction contact area on the train, wherein the target friction contact area is the area on the train's brake disc surface that contacts the brake actuator. The extraction module is used to identify local hotspot regions whose temperatures exceed a preset safety range from a temperature distribution image generated based on temperature distribution data, and to extract target geometric feature parameters of the local hotspot regions from the temperature distribution image. The generation module is used to generate the train's braking pressure adjustment parameters based on the target geometric feature parameters of the local hot spot area and the train's current operating status information, using a heat conduction model. An optimization module is used to optimize the brake pressure adjustment parameters to obtain optimized brake pressure adjustment parameters; The transmitting module is used to send the optimized braking pressure adjustment parameters to the braking actuator in the modular train braking system, so that the braking actuator applies dynamically changing braking force to the local friction contact area corresponding to the local hot spot area based on the optimized braking pressure adjustment parameters.

9. A computing device, characterized in that, It includes a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are invoked and executed by the processing component to implement a modular train braking method as described in any one of claims 1 to 7.

10. A computer storage medium, characterized in that, The system contains a computer program that, when executed by a computer, implements a modular train braking method as described in any one of claims 1 to 7.

Images