Whole vehicle axle test method

Through artificial intelligence technology, the characteristics of air pressure and braking torque are extracted from the braking factor chart, and fine-grained matching and response representation are performed, which solves the calculation deviation of braking factor caused by linear assumptions in the existing technology, and improves the testing of the vehicle axle Accuracy and reliability.

CN118913720AInactive Publication Date: 2024-11-08QINGDAO KAMAN TRANSPORTATION EQUIP CO LTD
View PDF 0 Cites 2 Cited by

Patent Information

Application Number
CN202411398747.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-10-09
Publication Date
2024-11-08
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

The existing vehicle axle testing method assumes that there is a simple linear relationship between air pressure and braking torque, and it is impossible to accurately capture the complex nonlinear relationship, resulting in deviations in the calculation results of braking factors.

Method used

Using artificial intelligence-based data processing technology, air pressure sampling values ​​and braking torque sampling values ​​are extracted from the braking factor chart, and air pressure-braking torque is generated through the feature extraction of correlation modes between sampling points, fine-grained feature decoupling and optimal matching. The fine-grained optimal alignment response represents intelligent calculation of the braking factor value.

Benefits of technology

It improves the accuracy and reliability of the calculation of braking factor, captures the complex nonlinear interaction between air pressure and braking torque, and enhances the reflection of braking performance.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118913720A_ABST
    Figure CN118913720A_ABST
Patent Text Reader

Abstract

The invention relates to the field of intelligent data analysis, and provides a whole vehicle axle test method, which comprises the following steps of: extracting a set of air pressure sampling values and a set of braking torque sampling values from a braking factor chart; local correlation between sampling points of the air pressure sampling value and the braking torque sampling value is carried out by adopting a data processing technology based on artificial intelligence; and intelligently obtaining a braking factor value based on the fine-grained interaction response representation between the local correlation characteristics between the air pressure sampling points and the local correlation characteristics between the braking torque sampling points. Therefore, a more complex nonlinear interaction relation between the air pressure and the braking torque and subtle changes between the sampling points can be captured, so that the fitting precision is improved, the actual braking performance is reflected more accurately, and the automation level and reliability of braking factor calculation are improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of intelligent data analysis, and more specifically, to a vehicle axle testing method. Background Art

[0002] The vehicle axle is an important part of the vehicle chassis system, usually including the front axle and the rear axle. The main functions of the axle are to support the weight of the vehicle body, transmit driving force and braking force, and maintain good contact between the tires and the ground to ensure the stability of the vehicle. When the vehicle is moving, good braking performance is the key to driver safety. With the improvement of intelligent driving assistance technology, accurate braking factors (braking force generated by unit air pressure) are essential. If the braking factor data is inaccurate, it may cause the emergency braking system to respond improperly, resulting in insufficient or excessive braking, thereby endangering driving safety.

[0003] Patent CN115901287A proposes a vehicle axle testing method, device, equipment and storage medium, which simulates the braking conditions under actual road conditions on a roller stand, measures and analyzes the braking effect of the wheel at different air pressure values, and thus obtains an accurate braking factor value.

[0004] In the above patent, a braking factor graph is drawn based on the data in the test air pressure value and braking torque value data set, and a curve is fitted using a fitting method (such as the least squares method and the gradient descent method) based on the discrete points on the braking factor graph, and the slope is calculated based on the fitted curve, which is the braking factor value. This is a traditional data fitting method to find the linear relationship between the air pressure value and the braking torque value, and calculate the braking factor value accordingly. However, this fitting method assumes that there is a simple linear relationship between air pressure and braking torque, but in actual situations, this relationship may be more complex, and even present nonlinear characteristics or have multiple inflection points. A single linear fitting method cannot accurately capture these complex nonlinear relationships, especially when there is a significant nonlinear interaction between air pressure and braking torque. This fitting method may cause deviations in the calculation results of the braking factor. In addition, when the data points show a significant nonlinear trend in certain intervals, the linear fitting may underestimate or overestimate the changes in the data points, thereby affecting the accuracy of the braking factor value.

[0005] Therefore, an optimized whole vehicle axle testing solution is desired. Summary of the invention

[0006] In view of the shortcomings of the prior art, the present application provides a whole vehicle axle testing method.

[0007] According to one aspect of the present application, a whole vehicle axle testing method is provided, which includes: placing a vehicle to be tested on a roller stand, the air pipe of the roller stand is connected to the relay valve of the whole vehicle axle of the vehicle to be tested, and the wheels of the vehicle to be tested are driven to accelerate to a target speed through the acceleration roller of the roller stand; in response to the roller reaching the target speed, the air pressure of the roller stand is adjusted to decelerate the wheel, and the braking torque value of the roller stand is recorded to obtain a time series set of braking torque values; based on the time series set of braking torque values, a braking factor chart is drawn; based on the braking factor chart, a braking factor value is obtained; wherein, based on the braking factor chart, the braking factor value is obtained, including: Extracting a set of air pressure sampling values ​​and a set of braking torque sampling values ​​from the braking factor chart; Performing inter-sampling point correlation pattern feature extraction on the set of air pressure sampling values ​​and the set of braking torque sampling values ​​to obtain a local correlation feature vector between air pressure sampling points and a local correlation feature vector between braking torque sampling points; Performing an air pressure sampling-braking torque characteristic interactive response on the local correlation feature vector between the air pressure sampling points and the local correlation feature vector between the braking torque sampling points to obtain an air pressure sampling-braking torque sampling response representation; Based on the air pressure sampling-braking torque sampling response representation, obtaining the braking factor value; Among them, performing air pressure sampling-braking torque characteristic interactive response on the local correlation feature vector between the air pressure sampling points and the local correlation feature vector between the braking torque sampling points to obtain the air pressure sampling-braking torque sampling response representation includes: Based on a predetermined scale, fine-grained feature decoupling is performed on the local correlation feature vector between the air pressure sampling points and the local correlation feature vector between the braking torque sampling points to obtain sub-component features of the air pressure sampling points and sub-component features of the braking torque sampling points; The air pressure sampling point sub-component characteristics and the braking torque sampling point sub-component characteristics are optimally matched to obtain the air pressure sampling-braking torque sampling response representation.

[0008] In the above-mentioned vehicle axle testing method, the set of air pressure sampling values ​​and the set of braking torque sampling values ​​are subjected to sampling point correlation pattern feature extraction to obtain local correlation feature vectors between air pressure sampling points and local correlation feature vectors between braking torque sampling points, including: respectively inputting the set of air pressure sampling values ​​and the set of braking torque sampling values ​​into a sampling point correlation pattern feature extractor based on a one-dimensional convolutional layer to obtain the local correlation feature vectors between air pressure sampling points and the local correlation feature vectors between braking torque sampling points.

[0009] In the above-mentioned whole vehicle axle test method, fine-grained feature decoupling is performed on the local correlation feature vectors between the air pressure sampling points and the local correlation feature vectors between the braking torque sampling points based on a predetermined scale to obtain air pressure sampling point sub-component features and braking torque sampling point sub-component features, including: fine-grained feature decoupling is performed on the local correlation feature vectors between the air pressure sampling points and the local correlation feature vectors between the braking torque sampling points based on a predetermined scale to obtain a set of air pressure sampling point sub-component feature vectors and a set of braking torque sampling point sub-component feature vectors as the air pressure sampling point sub-component features and the braking torque sampling point sub-component features.

[0010] In the above-mentioned vehicle axle test method, the air pressure sampling point sub-component characteristics and the braking torque sampling point sub-component characteristics are optimally matched to obtain the air pressure sampling-braking torque sampling response representation, including: calculating the air pressure-braking torque hyperbolic space distance measurement factor between any two air pressure sampling point sub-component feature vectors and the braking torque sampling point sub-component feature vectors in the set of the air pressure sampling point sub-component feature vectors and the set of the braking torque sampling point sub-component feature vectors; based on the air pressure-braking torque hyperbolic space distance measurement factor, the set of the air pressure sampling point sub-component feature vectors and the set of the braking torque sampling point sub-component feature vectors are optimally matched to obtain the air pressure sampling-braking torque sampling response representation. Obtain a set of optimal {air pressure sampling point sub-component feature vectors, braking torque sampling point sub-component feature vectors} pairs; input each optimal {air pressure sampling point sub-component feature vector, braking torque sampling point sub-component feature vector} pair in the set of optimal {air pressure sampling point sub-component feature vectors, braking torque sampling point sub-component feature vectors} pairs into a feature multi-dimensional interaction module to obtain a set of air pressure-braking torque optimal sub-component feature pairing interaction fusion feature vectors; cascade the set of air pressure-braking torque optimal sub-component feature pairing interaction fusion feature vectors to obtain an air pressure sampling-braking torque sampling fine-grained optimal alignment response representation vector as the air pressure sampling-braking torque sampling response representation.

[0011] In the above-mentioned vehicle axle test method, the air pressure-braking torque hyperbolic space distance measurement factor between any two air pressure sampling point sub-component feature vectors and the braking torque sampling point sub-component feature vectors in the set of the air pressure sampling point sub-component feature vectors and the set of the braking torque sampling point sub-component feature vectors is calculated, including: extracting a predetermined air pressure sampling point sub-component feature vector from the set of the air pressure sampling point sub-component feature vectors; calculating the hyperbolic space distance measurement factor between the predetermined air pressure sampling point sub-component feature vector and each braking torque sampling point sub-component feature vector in the set of the braking torque sampling point sub-component feature vector to obtain the set of air pressure-braking torque hyperbolic space distance measurement factors of the predetermined air pressure sampling point sub-component feature vectors.

[0012] In the above-mentioned vehicle axle test method, the hyperbolic space distance measurement factor between each braking torque sampling point sub-component feature vector in the set of the predetermined air pressure sampling point sub-component feature vector and the braking torque sampling point sub-component feature vector is calculated to obtain a set of air pressure-braking torque hyperbolic space distance measurement factors of the predetermined air pressure sampling point sub-component feature vector, including: calculating the position difference between the predetermined air pressure sampling point sub-component feature vector and the braking torque sampling point sub-component feature vector to obtain an air pressure-braking torque differential vector, and calculating the square of the norm of the air pressure-braking torque differential vector to obtain the air pressure-braking torque differential norm value; calculating the constant one and the air pressure sampling point sub-component feature vector; The method comprises the following steps: calculating the difference between the square of the one-norm of the sub-component eigenvector and the one-norm of the braking torque sampling point sub-component eigenvector to obtain the air pressure sampling point difference; calculating the difference between the constant one and the square of the one-norm of the braking torque sampling point sub-component eigenvector to obtain the braking torque sampling point difference; calculating the product between the air pressure sampling point difference and the braking torque sampling point difference to obtain the air pressure-braking torque product value; dividing the air pressure-braking torque difference norm value by the air pressure-braking torque product value, multiplying the obtained value by constant two and then adding it to constant one to obtain the air pressure-braking torque distance measurement value; inputting the air pressure-braking torque distance measurement value into the arccosh function to obtain the air pressure-braking torque hyperbolic space distance measurement factor.

[0013] In the above-mentioned vehicle axle test method, based on the air pressure-braking torque hyperbolic space distance measurement factor, the set of air pressure sampling point sub-component feature vectors and the set of braking torque sampling point sub-component feature vectors are screened for the best feature sub-component pairing to obtain the set of optimal {air pressure sampling point sub-component feature vectors, braking torque sampling point sub-component feature vectors} pairs, including: taking the braking torque sampling point sub-component feature vector corresponding to the smallest air pressure-braking torque hyperbolic space distance measurement factor in the set of air pressure-braking torque hyperbolic space distance measurement factors of the predetermined air pressure sampling point sub-component feature vectors and the predetermined air pressure sampling point sub-component feature vector as the optimal {air pressure sampling point sub-component feature vector, braking torque sampling point sub-component feature vector} pairing.

[0014] In the above-mentioned vehicle axle test method, each optimal {air pressure sampling point sub-component feature vector, braking torque sampling point sub-component feature vector} pair in the set of the optimal {air pressure sampling point sub-component feature vector, braking torque sampling point sub-component feature vector} pair is input into the feature multi-dimensional interaction module to obtain a set of air pressure-braking torque optimal sub-component feature pairing interactive fusion feature vectors, including: respectively calculating the position difference, position dot product and position addition between the optimal {air pressure sampling point sub-component feature vector, braking torque sampling point sub-component feature vector} pairs to obtain the air pressure-braking torque difference vector, the air pressure-braking torque dot product vector and an air pressure-braking torque sum vector; multiply the air pressure-braking torque differential vector, the air pressure-braking torque dot product vector and the air pressure-braking torque sum vector with the first weighted parameter, the second weighted parameter and the third weighted parameter respectively by position to obtain an air pressure-braking torque differential weighted vector, an air pressure-braking torque dot product weighted vector and an air pressure-braking torque sum weighted vector; calculate the air pressure-braking torque differential weighted vector, the air pressure-braking torque dot product weighted vector and the air pressure-braking torque sum weighted vector by position point to obtain the air pressure-braking torque optimal sub-component feature pairing interactive fusion feature vector.

[0015] In the above-mentioned vehicle axle test method, the braking factor value is obtained based on the air pressure sampling-braking torque sampling response representation, including: inputting the air pressure sampling-braking torque sampling fine-grained optimal alignment response representation vector into a braking factor value estimation module to obtain the braking factor value.

[0016] In the above-mentioned vehicle axle testing method, the air pressure sampling-braking torque sampling fine-grained optimal alignment response representation vector is input into a braking factor value estimation module to obtain the braking factor value, including: inputting the air pressure sampling-braking torque sampling fine-grained optimal alignment response representation vector into a brake factor value estimation module based on a decoder to obtain a decoded value of the brake factor value.

[0017] This application has significant technical effects due to the adoption of the above technical solutions: The vehicle axle test method provided by the present application extracts a set of air pressure sampling values ​​and a set of braking torque sampling values ​​from a braking factor chart, and uses artificial intelligence-based data processing technology to perform local correlation between the sampling points of the air pressure sampling values ​​and the braking torque sampling values, thereby intelligently obtaining the braking factor value based on the fine-grained interactive response representation between the local correlation features between the air pressure sampling points and the local correlation features between the braking torque sampling points. In this way, the more complex nonlinear interactive relationship between air pressure and braking torque and the subtle changes between sampling points can be captured, thereby improving the fitting accuracy, thereby more accurately reflecting the actual braking performance, and improving the automation level and reliability of the braking factor calculation. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0019] Figure 1 Flow chart of a vehicle axle testing method according to an embodiment of the present application.

[0020] Figure 2 The present invention is a flow chart for obtaining a braking factor value based on the braking factor chart in a vehicle axle testing method according to an embodiment of the present application.

[0021] Figure 3 The present invention is a data flow diagram for obtaining a braking factor value based on the braking factor chart in a vehicle axle testing method according to an embodiment of the present application.

[0022] Figure 4 In the vehicle axle testing method according to an embodiment of the present application, an air pressure sampling-braking torque characteristic interactive response is performed on the local correlation feature vectors between the air pressure sampling points and the local correlation feature vectors between the braking torque sampling points to obtain a flowchart representing the air pressure sampling-braking torque sampling response.

[0023] Figure 5 The present invention is a flowchart for optimally matching the air pressure sampling point sub-component characteristics and the braking torque sampling point sub-component characteristics in a vehicle axle testing method according to an embodiment of the present application to obtain the air pressure sampling-braking torque sampling response representation. DETAILED DESCRIPTION

[0024] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0025] As a key component of the vehicle chassis system, the vehicle axle is responsible for supporting the vehicle body, transmitting driving force, braking force and maintaining tire contact with the ground, which is crucial to vehicle stability. The quality of braking performance is directly related to driving safety, so the precise braking factor (i.e. the braking force generated by unit air pressure) is crucial for intelligent driving assistance technology.

[0026] Patent CN115901287A introduces an axle testing method, device, equipment and storage medium, which performs braking tests on a roller stand to simulate actual road conditions to obtain accurate braking factor values.

[0027] In the above patent, a braking factor graph is drawn based on the data in the test air pressure value and braking torque value data set, and the linear relationship between the air pressure and the braking torque is determined by a fitting method (such as the least squares method or the gradient descent method), thereby calculating the braking factor value. However, this fitting method assumes that there is a simple linear relationship between the air pressure and the braking torque, but in fact this relationship may be more complex, showing nonlinear characteristics or having multiple inflection points. A single linear fitting method cannot accurately capture complex nonlinear relationships, and when there is a significant nonlinear interaction between the air pressure and the braking torque, it may cause deviations in the braking factor calculation results. In addition, when the data points show a significant nonlinear trend in certain intervals, the linear fitting may underestimate or overestimate the changes in the data points, thereby affecting the accuracy of the braking factor value.

[0028] Based on this, the present application proposes a whole vehicle axle testing method. Figure 1 FIG. 1 is a flow chart of a vehicle axle testing method according to an embodiment of the present application. Figure 1 As shown, the whole vehicle axle testing method according to the embodiment of the present application includes: S110, placing the vehicle to be tested on a roller stand, the air pipe of the roller stand is connected to the relay valve of the whole vehicle axle of the vehicle to be tested, and the wheels of the vehicle to be tested are driven to accelerate to a target speed through the acceleration roller of the roller stand; S120, in response to the roller reaching the target speed, adjusting the air pressure of the roller stand to decelerate the wheel, and recording the braking torque value of the roller stand to obtain a time series set of braking torque values; S130, drawing a braking factor chart based on the time series set of braking torque values; S140, obtaining a braking factor value based on the braking factor chart.

[0029] It should be understood that the roller stand is a device for testing the axle of a whole vehicle, which can make the front wheel or rear wheel of the vehicle be on the roller of the roller stand. When the roller on the roller stand starts to rotate, the wheels of the whole vehicle parked on the roller stand will rotate together under the drive of the roller. In this application, it is first necessary to park the vehicle to be tested on the roller stand, and the air pipe of the roller stand equipment is connected to the relay valve of the whole vehicle axle of the vehicle to be tested. When the wheel of the vehicle to be tested is located on the roller stand, the roller of the roller stand is accelerated so that the roller drives the wheel of the vehicle to be tested to accelerate to the target speed. It is particularly noted that the target speed refers to the target speed of the roller, and the target speed here can be flexibly adjusted according to the actual test requirements. Then when the speed of the roller reaches the target speed, the air pressure of the roller stand is adjusted so that the wheel decelerates under the action of the air pressure, and the braking torque value of the roller stand is recorded to obtain the time series set of the braking torque value. Specifically, when the rolling speed of the roller reaches the target speed, the pneumatic device connected to the roller stand opens the air brake circuit through the driving relay valve of the front axle or rear axle of the whole vehicle, so that the speed of the wheel begins to decelerate under the action of air pressure. During the deceleration process, when the wheel deceleration is stable, the braking torque value of the current roller stand is collected. The braking torque value here refers to the braking torque value generated by the roller stand during the wheel braking process. Then, based on the time series set of the braking torque value, a braking factor chart is drawn. The braking factor chart in this application refers to a collection of data for recording the air pressure value applied by the pneumatic device connected to the roller stand, and the corresponding braking torque value under the action of the air pressure value. Finally, based on the braking factor chart, the braking factor value is obtained.

[0030] Accordingly, in obtaining the braking factor value based on the braking factor chart, the technical concept of the present application is to extract a set of air pressure sampling values ​​and a set of braking torque sampling values ​​from the braking factor chart, and use an artificial intelligence-based data processing and analysis algorithm to perform local correlation between the sampling points of the air pressure sampling values ​​and the braking torque sampling values, thereby intelligently obtaining the braking factor value based on the fine-grained interactive response representation between the local correlation features between the air pressure sampling points and the local correlation features between the braking torque sampling points. In this way, the more complex nonlinear interactive relationship between the air pressure and the braking torque and the subtle changes between the sampling points can be captured, thereby improving the fitting accuracy, thereby more accurately reflecting the actual braking performance, and improving the automation level and reliability of the braking factor calculation.

[0031] Figure 2 The present invention is a flow chart for obtaining a braking factor value based on the braking factor chart in a vehicle axle testing method according to an embodiment of the present application. Figure 3 The data flow diagram of obtaining the braking factor value based on the braking factor chart in the vehicle axle testing method according to the embodiment of the present application is as follows. Figure 2 and Figure 3 As shown, based on the braking factor chart, a braking factor value is obtained, including: S141, extracting a set of air pressure sampling values ​​and a set of braking torque sampling values ​​from the braking factor chart; S142, performing correlation pattern feature extraction between sampling points on the set of air pressure sampling values ​​and the set of braking torque sampling values ​​to obtain a local correlation feature vector between air pressure sampling points and a local correlation feature vector between braking torque sampling points; S143, performing air pressure sampling-braking torque feature interactive response on the local correlation feature vector between the air pressure sampling points and the local correlation feature vector between the braking torque sampling points to obtain an air pressure sampling-braking torque sampling response representation; S144, obtaining the braking factor value based on the air pressure sampling-braking torque sampling response representation.

[0032] In step S141, a set of air pressure sampling values ​​and a set of braking torque sampling values ​​are extracted from the braking factor chart. It should be understood that the braking factor value refers to the parameter that affects the braking effect of different air pressure values ​​when the wheel is braking during air braking. Considering that the braking factor chart is an original data set and only includes simple data correspondences, in order to more accurately identify and analyze the deep correlation between air pressure and braking torque, and then ensure the reliability of the final braking factor value, in the technical solution of the present application, it is first necessary to extract a set of air pressure sampling values ​​and a set of braking torque sampling values ​​from the braking factor chart.

[0033] In step S142, the set of air pressure sampling values ​​and the set of braking torque sampling values ​​are subjected to sampling point correlation pattern feature extraction to obtain local correlation feature vectors between air pressure sampling points and local correlation feature vectors between braking torque sampling points. Specifically, in an embodiment of the present application, the set of air pressure sampling values ​​and the set of braking torque sampling values ​​are subjected to sampling point correlation pattern feature extraction to obtain local correlation feature vectors between air pressure sampling points and local correlation feature vectors between braking torque sampling points, including: inputting the set of air pressure sampling values ​​and the set of braking torque sampling values ​​into a sampling point correlation pattern feature extractor based on a one-dimensional convolution layer to obtain local correlation feature vectors between air pressure sampling points and local correlation feature vectors between braking torque sampling points. It should be understood that there are different degrees of local correlation between the sampling points in the set of air pressure sampling values ​​and the set of braking torque sampling values, and the convolution kernel of the one-dimensional convolution layer will slide on the sampling point data, considering only a small data window at a time, which helps to identify local correlations in the data. Based on this, in the technical solution of the present application, the set of air pressure sampling values ​​and the set of braking torque sampling values ​​are respectively input into a sampling point correlation pattern feature extractor based on a one-dimensional convolutional layer to obtain a local correlation feature vector between air pressure sampling points and a local correlation feature vector between braking torque sampling points.

[0034] In step S143, the local correlation feature vectors between the air pressure sampling points and the local correlation feature vectors between the braking torque sampling points are subjected to an air pressure sampling-braking torque feature interactive response to obtain an air pressure sampling-braking torque sampling response representation. Accordingly, it is considered that there is a microscopic interaction detail information and relationship between the local sampling points between the local correlation features between the air pressure sampling points and the local correlation features between the braking torque sampling points. In order to understand and model the complex relationship between air pressure and braking torque at a deeper level, thereby improving the accuracy and reliability of the braking factor value, in the technical solution of the present application, the local correlation feature vectors between the air pressure sampling points and the local correlation feature vectors between the braking torque sampling points are subjected to an air pressure sampling-braking torque feature interactive response to obtain an air pressure sampling-braking torque sampling fine-grained optimal alignment response representation vector as an air pressure sampling-braking torque sampling response representation.

[0035] In detail, first, the local correlation feature vectors between the air pressure sampling points and the local correlation feature vectors between the braking torque sampling points are decoupled based on a predetermined scale to obtain a set of sub-component feature vectors of the air pressure sampling points and a set of sub-component feature vectors of the braking torque sampling points. That is, separating and analyzing the entire feature vector can obtain simpler, lower-level features, such as the air pressure value or braking torque value at a certain time point, the rate of change of the air pressure or braking torque between adjacent sampling points, etc., so as to obtain more fine-grained features. Then, the hyperbolic space distance measurement factor between any two sub-component feature vectors of the air pressure sampling points and the sub-component feature vectors of the braking torque sampling points is calculated to more effectively measure the similarity or difference between the two feature vectors to obtain the air pressure-braking torque hyperbolic space distance measurement factor. In particular, the hyperbolic space is a non-Euclidean geometric space, which is suitable for processing data sets with hierarchical structures or long-tail distributions. That is, since the hyperbolic space expands faster in the direction away from the origin, this measurement method is more conducive to maintaining the hierarchical structure of the data and is more suitable for processing data distributions that grow exponentially. This is extremely critical for identifying the relationship between features, especially when analyzing complex or nonlinear interactions, and can more accurately reflect the actual differences between features. Then, based on the air pressure-braking torque hyperbolic space distance measurement factor, the set of sub-component feature vectors of the air pressure sampling points and the set of sub-component feature vectors of the braking torque sampling points are screened for the best feature sub-component pairing to obtain the best set of {air pressure sampling point sub-component feature vectors, braking torque sampling point sub-component feature vectors} pairs. That is, through the best feature sub-component pairing screening, a pair of feature vectors with the closest distance in the hyperbolic space can be identified to achieve the best correspondence. The core of this process is to ensure that each pair of matches can maximize the similarity of the two sets of feature vectors, so as to maximize the retention of the most valuable information in the subsequent feature fusion. Subsequently, the optimal pairing of each group is input into the feature multi-dimensional interaction module to further explore and understand the deep-level association and interaction relationship between the air pressure sampling points and the braking torque sampling points to obtain a set of air pressure-braking torque optimal sub-component feature pairing interaction fusion feature vectors. Finally, the set of interactive fusion features is cascaded to fuse the new information generated by fine-grained matching and multi-dimensional interaction, and the air pressure sampling-braking torque sampling fine-grained optimal alignment response representation vector is obtained.

[0036] Specifically, Figure 4 The present invention is a flow chart of performing air pressure sampling-braking torque characteristic interactive response on the local correlation feature vector between the air pressure sampling points and the local correlation feature vector between the braking torque sampling points in the vehicle axle testing method according to the embodiment of the present application to obtain the air pressure sampling-braking torque sampling response representation. Figure 4As shown, the local correlation feature vector between the air pressure sampling points and the local correlation feature vector between the braking torque sampling points are subjected to an air pressure sampling-braking torque feature interactive response to obtain an air pressure sampling-braking torque sampling response representation, including: S1431, based on a predetermined scale, fine-grained feature decoupling is performed on the local correlation feature vector between the air pressure sampling points and the local correlation feature vector between the braking torque sampling points to obtain an air pressure sampling point sub-component feature and a braking torque sampling point sub-component feature; S1432, optimally matching the air pressure sampling point sub-component feature and the braking torque sampling point sub-component feature to obtain the air pressure sampling-braking torque sampling response representation.

[0037] More specifically, in an embodiment of the present application, fine-grained feature decoupling is performed on the local correlation feature vectors between the air pressure sampling points and the local correlation feature vectors between the braking torque sampling points based on a predetermined scale to obtain air pressure sampling point sub-component features and braking torque sampling point sub-component features, including: fine-grained feature decoupling is performed on the local correlation feature vectors between the air pressure sampling points and the local correlation feature vectors between the braking torque sampling points based on a predetermined scale to obtain a set of air pressure sampling point sub-component feature vectors and a set of braking torque sampling point sub-component feature vectors as the air pressure sampling point sub-component features and the braking torque sampling point sub-component features.

[0038] More specifically, Figure 5 The present invention is a flow chart of optimally matching the air pressure sampling point sub-component characteristics and the braking torque sampling point sub-component characteristics in the vehicle axle testing method according to the embodiment of the present application to obtain the air pressure sampling-braking torque sampling response representation. Figure 5As shown, the air pressure sampling point sub-component features and the braking torque sampling point sub-component features are optimally matched to obtain the air pressure sampling-braking torque sampling response representation, including: S14321, calculating the air pressure-braking torque hyperbolic space distance measurement factor between any two air pressure sampling point sub-component feature vectors and the braking torque sampling point sub-component feature vectors in the set of the air pressure sampling point sub-component feature vectors and the set of the braking torque sampling point sub-component feature vectors; S14322, based on the air pressure-braking torque hyperbolic space distance measurement factor, the set of the air pressure sampling point sub-component feature vectors and the set of the braking torque sampling point sub-component feature vectors are optimally matched to obtain the optimal { A set of {air pressure sampling point sub-component feature vectors, braking torque sampling point sub-component feature vectors} pairs; S14323, inputting each optimal {air pressure sampling point sub-component feature vector, braking torque sampling point sub-component feature vector} pair in the set of optimal {air pressure sampling point sub-component feature vectors, braking torque sampling point sub-component feature vectors} pairs into a feature multi-dimensional interaction module to obtain a set of air pressure-braking torque optimal sub-component feature pairing interaction fusion feature vectors; S14324, cascading the set of air pressure-braking torque optimal sub-component feature pairing interaction fusion feature vectors to obtain an air pressure sampling-braking torque sampling fine-grained optimal alignment response representation vector as the air pressure sampling-braking torque sampling response representation.

[0039] More specifically, in an embodiment of the present application, calculating the air pressure-braking torque hyperbolic space distance measurement factor between any two air pressure sampling point sub-component feature vectors and the braking torque sampling point sub-component feature vectors in the set of the air pressure sampling point sub-component feature vectors and the set of the braking torque sampling point sub-component feature vectors includes: extracting a predetermined air pressure sampling point sub-component feature vector from the set of the air pressure sampling point sub-component feature vectors; calculating the hyperbolic space distance measurement factor between the predetermined air pressure sampling point sub-component feature vector and each braking torque sampling point sub-component feature vector in the set of the braking torque sampling point sub-component feature vectors to obtain a set of air pressure-braking torque hyperbolic space distance measurement factors of the predetermined air pressure sampling point sub-component feature vectors.

[0040] More specifically, in an embodiment of the present application, the hyperbolic space distance measurement factor between each braking torque sampling point sub-component feature vector in the set of the predetermined air pressure sampling point sub-component feature vector and the braking torque sampling point sub-component feature vector is calculated to obtain a set of air pressure-braking torque hyperbolic space distance measurement factors of the predetermined air pressure sampling point sub-component feature vector, including: calculating the position difference between the predetermined air pressure sampling point sub-component feature vector and the braking torque sampling point sub-component feature vector to obtain an air pressure-braking torque differential vector, and calculating the square of a norm of the air pressure-braking torque differential vector to obtain an air pressure-braking torque differential norm value; calculating a constant one and the air pressure sampling point sub-component feature vector. The method comprises the following steps: calculating the difference between the square of the one-norm eigenvector of the point component eigenvector and the braking torque sampling point component eigenvector to obtain the air pressure sampling point difference value; calculating the difference between the constant one and the square of the one-norm eigenvector of the braking torque sampling point component eigenvector to obtain the braking torque sampling point difference value; calculating the product between the air pressure sampling point difference value and the braking torque sampling point difference value to obtain the air pressure-braking torque product value; dividing the air pressure-braking torque difference norm value by the air pressure-braking torque product value, multiplying the obtained value by constant two and then adding it to constant one to obtain the air pressure-braking torque distance measurement value; inputting the air pressure-braking torque distance measurement value into the arccosh function to obtain the air pressure-braking torque hyperbolic space distance measurement factor.

[0041] More specifically, in an embodiment of the present application, based on the air pressure-braking torque hyperbolic space distance measurement factor, the set of air pressure sampling point sub-component feature vectors and the set of braking torque sampling point sub-component feature vectors are screened for optimal feature sub-component pairing to obtain a set of optimal {air pressure sampling point sub-component feature vectors, braking torque sampling point sub-component feature vectors} pairs, including: taking the braking torque sampling point sub-component feature vector corresponding to the smallest air pressure-braking torque hyperbolic space distance measurement factor in the set of air pressure-braking torque hyperbolic space distance measurement factors of the predetermined air pressure sampling point sub-component feature vectors and the predetermined air pressure sampling point sub-component feature vector as the optimal {air pressure sampling point sub-component feature vector, braking torque sampling point sub-component feature vector} pairing.

[0042] More specifically, in an embodiment of the present application, each optimal {air pressure sampling point sub-component feature vector, braking torque sampling point sub-component feature vector} pair in the set of the optimal {air pressure sampling point sub-component feature vector, braking torque sampling point sub-component feature vector} pair is input into a feature multi-dimensional interaction module to obtain a set of air pressure-braking torque optimal sub-component feature pairing interaction fusion feature vectors, including: respectively calculating the position difference, position point multiplication and position addition between the optimal {air pressure sampling point sub-component feature vector, braking torque sampling point sub-component feature vector} pairs to obtain an air pressure-braking torque difference vector, an air pressure-braking torque point multiplication vector, and an air pressure-braking torque point multiplication vector. The method comprises the steps of: calculating a pressure-braking torque differential vector, a pressure-braking torque point multiplication vector and a pressure-braking torque sum vector; multiplying the pressure-braking torque differential vector, the pressure-braking torque point multiplication vector and the pressure-braking torque sum vector with the first weighted parameter, the second weighted parameter and the third weighted parameter respectively by position to obtain a pressure-braking torque differential weighted vector, a pressure-braking torque point multiplication weighted vector and a pressure-braking torque sum weighted vector; calculating the pressure-braking torque differential weighted vector, the pressure-braking torque point multiplication weighted vector and the pressure-braking torque sum weighted vector by position point to obtain the pressure-braking torque optimal sub-component feature pairing interactive fusion feature vector.

[0043] In the embodiment of the present application, specifically, the characteristic interaction response is performed on the local correlation feature vector between the air pressure sampling points and the local correlation feature vector between the braking torque sampling points, which can be expressed by the formula: ; in, is the local correlation feature vector between the pressure sampling points, are the first, second, ..., and third components of the pressure sampling point in the set of sub-component feature vectors. ,..., The sub-component feature vectors of the pressure sampling points, is the local correlation feature vector between the braking torque sampling points, are respectively the first, second, ..., first ,..., The sub-component feature vectors of the braking torque sampling points, Decouple operations for fine-grained features, For calculation and The air pressure-braking torque hyperbolic space distance measurement factor between them, To return the minimum value value, is the minimum approximate matching value, To calculate the square of the one-norm of a vector, is the inverse hyperbolic cosine function, and They are position difference, position dot product and position addition respectively. and is the corresponding weighting parameter, is the first in the set of sub-component feature vectors of the braking torque sampling points The sub-component feature vectors of the braking torque sampling points, is the set of the air pressure-braking torque optimal sub-component feature pairing interactive fusion feature vectors The optimal sub-component feature pairing of air pressure and braking torque interacts and fuses the feature vector, are the first, second and third in the set of the air pressure-braking torque optimal sub-component feature pairing interactive fusion feature vectors. The optimal sub-component feature pairing of air pressure and braking torque interacts and fuses the feature vector, For cascade processing, is the air pressure sampling-braking torque sampling fine-grained optimal alignment response representation vector.

[0044] In step S144, based on the air pressure sampling-braking torque sampling response representation, the braking factor value is obtained. Specifically, in an embodiment of the present application, based on the air pressure sampling-braking torque sampling response representation, the braking factor value is obtained, including: inputting the air pressure sampling-braking torque sampling fine-grained optimal alignment response representation vector into the braking factor value estimation module to obtain the braking factor value. More specifically, in an embodiment of the present application, inputting the air pressure sampling-braking torque sampling fine-grained optimal alignment response representation vector into the braking factor value estimation module to obtain the braking factor value, including: inputting the air pressure sampling-braking torque sampling fine-grained optimal alignment response representation vector into the brake factor value estimation module based on the decoder to obtain the decoded value of the braking factor value. That is, the air pressure sampling-braking torque sampling response representation obtained by the feature interaction response of the local correlation feature vector between the air pressure sampling points and the local correlation feature vector between the braking torque sampling points is decoded and processed, so as to intelligently obtain the braking factor value. In this way, the more complex nonlinear interaction between air pressure and braking torque and the subtle changes between sampling points can be captured, thereby improving the fitting accuracy, thereby more accurately reflecting the actual braking performance and improving the automation level and reliability of braking factor calculation.

[0045] In particular, in the technical solution of the present application, the local correlation feature vector between the air pressure sampling points and the local correlation feature vector between the braking torque sampling points represent the local correlation features between samples of the air pressure sampling value and the braking torque sampling value, respectively. Accordingly, when the local correlation feature vector between the air pressure sampling points and the local correlation feature vector between the braking torque sampling points are subjected to the air pressure sampling-braking torque feature interactive response, the source data modality difference and the data representation magnitude difference between the local correlation feature vector between the air pressure sampling points and the local correlation feature vector between the braking torque sampling points will cause an offset in the feature fine-grained optimal matching, and the air pressure sampling-braking torque sampling fine-grained optimal alignment response representation vector will have a fine-grained response aggregation difference caused by the optimal matching offset, so it is expected to improve the detailed semantic aggregation expression effect of the fine-grained response aggregation difference of the air pressure sampling-braking torque sampling fine-grained optimal alignment response representation vector.

[0046] Preferably, inputting the air pressure sampling-braking torque sampling fine-grained optimal alignment response representation vector into a decoder-based braking factor value estimation module to obtain a decoded value of the braking factor value includes: calculating the absolute value sum of each eigenvalue of the air pressure sampling-braking torque sampling fine-grained optimal alignment response representation vector to obtain a first air pressure sampling-braking torque sampling fine-grained optimal alignment response and modulation value, and calculating the square root of the sum of squares of each eigenvalue of the air pressure sampling-braking torque sampling fine-grained optimal alignment response representation vector to obtain a second air pressure sampling-braking torque sampling fine-grained optimal alignment response and modulation value; performing dot-wise subtraction on the air pressure sampling-braking torque sampling fine-grained optimal alignment response representation vector and the second air pressure sampling-braking torque sampling fine-grained optimal alignment response and modulation value, respectively performing dot-wise multiplication with the number of eigenvalues ​​of the air pressure sampling-braking torque sampling fine-grained optimal alignment response representation vector and the reciprocal of the first air pressure sampling-braking torque sampling fine-grained optimal alignment response and modulation value, and taking the reciprocal of each eigenvalue to obtain the first air pressure sampling-braking torque sampling fine-grained optimal alignment response and modulation value. The method comprises the following steps: obtaining a phase conversion vector of a second air pressure sampling-braking torque sampling fine-grained optimal alignment response; performing dot-wise subtraction on the air pressure sampling-braking torque sampling fine-grained optimal alignment response representation vector and the first air pressure sampling-braking torque sampling fine-grained optimal alignment response and modulation value, and then performing dot-wise multiplication on the square root of the number of eigenvalues ​​of the air pressure sampling-braking torque sampling fine-grained optimal alignment response representation vector and the reciprocal of the second air pressure sampling-braking torque sampling fine-grained optimal alignment response and modulation value, and taking the reciprocal of each eigenvalue to obtain a second air pressure sampling-braking torque sampling fine-grained optimal alignment response phase conversion vector; performing dot-wise subtraction on the first air pressure sampling-braking torque sampling fine-grained optimal alignment response phase conversion vector and the dot product vector of the second air pressure sampling-braking torque sampling fine-grained optimal alignment response phase conversion vector and a weighted hyperparameter to obtain an optimized air pressure sampling-braking torque sampling fine-grained optimal alignment response representation vector; and inputting the optimized air pressure sampling-braking torque sampling fine-grained optimal alignment response representation vector into the decoder-based braking factor value estimation module to obtain a decoded value of the braking factor value.

[0047] Here, the optimized representation of the air pressure sampling-braking torque sampling fine-grained optimal alignment response representation vector is: ; in, represents the air pressure sampling-braking torque sampling fine-grained optimal alignment response representation vector, represents the eigenvalue of the position in the air pressure sampling-braking torque sampling fine-grained optimal alignment response representation vector, represents the number of eigenvalues ​​of the air pressure sampling-braking torque sampling fine-grained optimal alignment response representation vector, represents the first air pressure sampling-braking torque sampling fine-grained optimal alignment response and modulation value, represents the second air pressure sampling-braking torque sampling fine-grained optimal alignment response and modulation value, For point-by-position multiplication, It means to subtract by position point. Take the reciprocal of each eigenvalue in the vector, represents the first air pressure sampling-braking torque sampling fine-grained optimal alignment response phase conversion vector, represents the second air pressure sampling-braking torque sampling fine-grained optimal alignment response phase conversion vector, represents the weighted hyperparameter, The optimized air pressure sampling-braking torque sampling fine-grained optimal alignment response representation vector is represented.

[0048] In this preferred embodiment, the difference in the feature value of the air pressure sampling-braking torque sampling fine-grained optimal alignment response representation vector relative to the feature set of the entire vector of the air pressure sampling-braking torque sampling fine-grained optimal alignment response representation vector and the difference in modulation representation are used as semantic change intensity information, and a phase-like conversion corresponding to the position-based intensity modulation is performed through different and modulation representation forms, so that the aggregation enhancement of semantic change phase perception can enhance the axial aggregation receptive field along the feature aggregation direction by performing a spatial translation operation based on alternating stacking under the scale balance of the vector set of the air pressure sampling-braking torque sampling fine-grained optimal alignment response representation vector, thereby enhancing the perception effect of the aggregated semantics of the air pressure sampling-braking torque sampling fine-grained optimal alignment response representation vector on the detailed semantic changes, thereby improving the expression effect of the air pressure sampling-braking torque sampling fine-grained optimal alignment response representation vector, and improving the accuracy of the decoded value of the braking factor value obtained by its input to the brake factor value estimation module based on the decoder. In this way, the more complex nonlinear interaction between air pressure and braking torque and the subtle changes between sampling points can be captured, thereby improving the fitting accuracy, thereby more accurately reflecting the actual braking performance and improving the automation level and reliability of braking factor calculation.

[0049] In summary, the vehicle axle test method based on the embodiment of the present application is explained, which extracts a set of air pressure sampling values ​​and a set of braking torque sampling values ​​from the braking factor chart, and uses artificial intelligence-based data processing technology to perform local correlation between the sampling points of the air pressure sampling values ​​and the braking torque sampling values, thereby intelligently obtaining the braking factor value based on the fine-grained interactive response representation between the local correlation features between the air pressure sampling points and the local correlation features between the braking torque sampling points. In this way, the more complex nonlinear interactive relationship between air pressure and braking torque and the subtle changes between sampling points can be captured, thereby improving the fitting accuracy, thereby more accurately reflecting the actual braking performance, and improving the automation level and reliability of the braking factor calculation.

[0050] It should be noted that in the method of the present application, each step can be decomposed and / or recombined. These decompositions and / or recombinations should be regarded as equivalent solutions of the present application.

[0051] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use the present application. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other aspects without departing from the scope of the present application. Therefore, the present application is not intended to be limited to the aspects shown herein, but rather to the widest scope consistent with the principles and novel features of the present application.

[0052] The above description has been given for the purpose of illustration and description. In addition, this description is not intended to limit the embodiments of the present application to the forms disclosed herein. Although multiple example aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, changes, additions and sub-combinations thereof.

Claims

1. A vehicle axle testing method, comprising: The vehicle to be tested is placed on a roller stand, the air pipe of the roller stand is connected to the relay valve of the whole vehicle axle of the vehicle to be tested, and the wheels of the vehicle to be tested are driven to accelerate to a target speed through the acceleration roller of the roller stand; in response to the roller reaching the target speed, the air pressure of the roller stand is adjusted to decelerate the wheels, and the braking torque value of the roller stand is recorded to obtain a time series set of braking torque values; based on the time series set of braking torque values, a braking factor chart is drawn; Based on the braking factor chart, a braking factor value is obtained; characterized in that, based on the braking factor chart, the braking factor value is obtained, including: Extracting a set of air pressure sampling values ​​and a set of braking torque sampling values ​​from the braking factor chart; Performing inter-sampling point correlation pattern feature extraction on the set of air pressure sampling values ​​and the set of braking torque sampling values ​​to obtain a local correlation feature vector between air pressure sampling points and a local correlation feature vector between braking torque sampling points; Performing air pressure sampling-braking torque feature interactive response on the local correlation feature vector between the air pressure sampling points and the local correlation feature vector between the braking torque sampling points to obtain air pressure sampling-braking torque sampling response representation, including: performing fine-grained feature decoupling on the local correlation feature vector between the air pressure sampling points and the local correlation feature vector between the braking torque sampling points based on a predetermined scale to obtain air pressure sampling point sub-component features and braking torque sampling point sub-component features; performing optimal matching on the air pressure sampling point sub-component features and the braking torque sampling point sub-component features to obtain the air pressure sampling-braking torque sampling response representation; The braking factor value is obtained based on the air pressure sampling-braking torque sampling response representation.

2. The vehicle axle testing method according to claim 1, characterized in that: The set of air pressure sampling values ​​and the set of braking torque sampling values ​​are subjected to sampling point correlation pattern feature extraction to obtain local correlation feature vectors between air pressure sampling points and local correlation feature vectors between braking torque sampling points, including: respectively inputting the set of air pressure sampling values ​​and the set of braking torque sampling values ​​into a sampling point correlation pattern feature extractor based on a one-dimensional convolutional layer to obtain the local correlation feature vectors between air pressure sampling points and the local correlation feature vectors between braking torque sampling points.

3. The vehicle axle testing method according to claim 2, characterized in that: Based on a predetermined scale, fine-grained feature decoupling is performed on the local correlation feature vectors between the air pressure sampling points and the local correlation feature vectors between the braking torque sampling points to obtain air pressure sampling point sub-component features and braking torque sampling point sub-component features, including: based on a predetermined scale, fine-grained feature decoupling is performed on the local correlation feature vectors between the air pressure sampling points and the local correlation feature vectors between the braking torque sampling points to obtain a set of air pressure sampling point sub-component feature vectors and a set of braking torque sampling point sub-component feature vectors as the air pressure sampling point sub-component features and the braking torque sampling point sub-component features.

4. The vehicle axle testing method according to claim 3, characterized in that: Optimally matching the air pressure sampling point sub-component features and the braking torque sampling point sub-component features to obtain the air pressure sampling-braking torque sampling response representation, including: Calculating an air pressure-braking torque hyperbolic space distance metric factor between any two air pressure sampling point sub-component feature vectors and braking torque sampling point sub-component feature vectors in the set of the air pressure sampling point sub-component feature vectors and the set of the braking torque sampling point sub-component feature vectors; Based on the air pressure-braking torque hyperbolic space distance measurement factor, the set of the air pressure sampling point sub-component feature vectors and the set of the braking torque sampling point sub-component feature vectors are screened for optimal feature sub-component pairing to obtain an optimal {air pressure sampling point sub-component feature vector, braking torque sampling point sub-component feature vector} pairing set; Input each optimal {air pressure sampling point sub-component feature vector, braking torque sampling point sub-component feature vector} pair in the set of the optimal {air pressure sampling point sub-component feature vector, braking torque sampling point sub-component feature vector} pair into the feature multi-dimensional interaction module to obtain a set of air pressure-braking torque optimal sub-component feature pairing interaction fusion feature vectors; The set of the air pressure-braking torque optimal sub-component feature pairing interactive fusion feature vectors is cascaded to obtain an air pressure sampling-braking torque sampling fine-grained optimal alignment response representation vector as the air pressure sampling-braking torque sampling response representation.

5. The vehicle axle testing method according to claim 4, characterized in that: Calculating the air pressure-braking torque hyperbolic space distance measurement factor between any two air pressure sampling point sub-component feature vectors and braking torque sampling point sub-component feature vectors in the set of the air pressure sampling point sub-component feature vectors and the set of the braking torque sampling point sub-component feature vectors, comprising: Extracting a predetermined air pressure sampling point sub-component feature vector from the set of air pressure sampling point sub-component feature vectors; The hyperbolic space distance measurement factor between the predetermined air pressure sampling point sub-component feature vector and each braking torque sampling point sub-component feature vector in the set of the braking torque sampling point sub-component feature vector is calculated to obtain a set of air pressure-braking torque hyperbolic space distance measurement factors of the predetermined air pressure sampling point sub-component feature vector.

6. The vehicle axle testing method according to claim 5, characterized in that: Calculating the hyperbolic space distance measurement factor between the predetermined air pressure sampling point sub-component feature vector and each braking torque sampling point sub-component feature vector in the set of the braking torque sampling point sub-component feature vector to obtain a set of air pressure-braking torque hyperbolic space distance measurement factors of the predetermined air pressure sampling point sub-component feature vector, comprising: Calculating the position difference between the predetermined air pressure sampling point sub-component feature vector and the braking torque sampling point sub-component feature vector to obtain an air pressure-braking torque differential vector, and calculating the square of a norm of the air pressure-braking torque differential vector to obtain an air pressure-braking torque differential norm value; Calculating the difference between the constant 1 and the square of the norm of the sub-component eigenvector of the air pressure sampling point to obtain the air pressure sampling point difference; Calculating the difference between the constant 1 and the square of the norm of the braking torque sampling point sub-component feature vector to obtain the braking torque sampling point difference; Calculating the product of the air pressure sampling point difference value and the braking torque sampling point difference value to obtain an air pressure-braking torque product value; After dividing the air pressure-braking torque differential norm value by the air pressure-braking torque product value, the obtained value is first multiplied by constant 2 and then added to constant 1 to obtain the air pressure-braking torque distance measurement value; The air pressure-braking torque distance measurement value is input into an arccosh function to obtain the air pressure-braking torque hyperbolic space distance measurement factor.

7. The vehicle axle testing method according to claim 6, characterized in that: Based on the air pressure-braking torque hyperbolic space distance measurement factor, the set of air pressure sampling point sub-component feature vectors and the set of braking torque sampling point sub-component feature vectors are screened for optimal feature sub-component pairing to obtain an optimal set of {air pressure sampling point sub-component feature vectors, braking torque sampling point sub-component feature vectors} pairs, including: taking the braking torque sampling point sub-component feature vector corresponding to the smallest air pressure-braking torque hyperbolic space distance measurement factor in the set of air pressure-braking torque hyperbolic space distance measurement factors of the predetermined air pressure sampling point sub-component feature vectors and the predetermined air pressure sampling point sub-component feature vector as the optimal {air pressure sampling point sub-component feature vector, braking torque sampling point sub-component feature vector} pairing.

8. The vehicle axle testing method according to claim 7, characterized in that: Inputting each optimal {air pressure sampling point sub-component feature vector, braking torque sampling point sub-component feature vector} pair in the set of the optimal {air pressure sampling point sub-component feature vector, braking torque sampling point sub-component feature vector} pair into the feature multi-dimensional interaction module to obtain a set of air pressure-braking torque optimal sub-component feature pairing interaction fusion feature vectors, including: Respectively calculating the position difference, position dot product and position addition between the optimal {air pressure sampling point sub-component feature vector, braking torque sampling point sub-component feature vector} pairs to obtain an air pressure-braking torque difference vector, an air pressure-braking torque dot product vector and an air pressure-braking torque sum vector; The air pressure-braking torque difference vector, the air pressure-braking torque dot product vector and the air pressure-braking torque sum vector are respectively multiplied by position with the first weighting parameter, the second weighting parameter and the third weighting parameter to obtain an air pressure-braking torque difference weighted vector, an air pressure-braking torque dot product weighted vector and an air pressure-braking torque sum weighted vector; The air pressure-braking torque differential weighted vector, the air pressure-braking torque point multiplication weighted vector and the air pressure-braking torque sum weighted vector are calculated point by point to obtain the air pressure-braking torque optimal sub-component feature pairing interactive fusion feature vector.

9. The vehicle axle testing method according to claim 8, characterized in that: Based on the air pressure sampling-braking torque sampling response representation, the braking factor value is obtained, including: inputting the air pressure sampling-braking torque sampling fine-grained optimal alignment response representation vector into a braking factor value estimation module to obtain the braking factor value.

10. The vehicle axle testing method according to claim 9, characterized in that: The air pressure sampling-braking torque sampling fine-grained optimal alignment response representation vector is input into a braking factor value estimation module to obtain the braking factor value, including: inputting the air pressure sampling-braking torque sampling fine-grained optimal alignment response representation vector into a brake factor value estimation module based on a decoder to obtain a decoded value of the brake factor value.

Citation Information

Cited By

  • Preparation method of seasoning

    CN119236779A

  • System exception handling method, device, equipment, medium and product

    CN119759723A