A tire characterization method and system based on lateral force coefficient

Abstract

The present invention relates to the technical field of tire quality evaluation, specifically to a tire characterization method and system based on a lateral force coefficient, the method comprising the following steps: selecting a test tire and a test section, obtaining tire test data based on the test tire and the test section; setting a data quality analysis model, using the data quality analysis model to perform quality inspection on the tire test data, and obtaining target test data for the test tire; constructing a tire characterization index analysis mechanism, obtaining tire characterization parameters based on the tire characterization index analysis mechanism and the target test data; obtaining characterization parameter characteristic information of the tire characterization parameters based on the target test data; obtaining a lateral force friction coefficient based on the characterization parameter characteristic information, and analyzing and evaluating the tire performance of the test tire based on the lateral force friction coefficient. The present invention tests tires based on the lateral force friction coefficient and effectively understands the use performance of the tire.

Description

Technical Field

[0001] The present invention relates to the technical field of tire quality evaluation, and in particular to a tire characterization method and system based on lateral force coefficient. Background Art

[0002] Tire performance evaluation is a key issue in the tire manufacturing industry and vehicle engineering. Currently, tires are mainly identified and classified by their specifications and models. Related parameters such as size, tread width, tire height, etc. mainly reflect the basic physical characteristics and appearance of the tire. However, there are certain limitations in tire performance evaluation, especially when facing special tires that need to be accurately measured and evaluated.

[0003] Traditional tire characterization methods can no longer meet the classification requirements of modern vehicles for tire performance, especially in situations where precise measurement of special tires is required. Tire performance is directly related to the vehicle's handling stability and safety under dynamic driving conditions such as turning and changing lanes.

[0004] Therefore, it is necessary to develop a more accurate and comprehensive tire characterization method and system to make up for the shortcomings of traditional characterization methods, achieve accurate classification and evaluation of tire performance, provide strong support for tire research and development, production and quality control, and also help improve the overall performance and safety of the vehicle. Summary of the invention

[0005] In view of the shortcomings of existing methods and the needs of practical applications, in order to improve the efficiency and accuracy of tire performance evaluation and provide a basis for multiple key links such as tire research and development, production quality control and safety assessment, the present invention can accurately capture the key performance indicators of tires and analyze and evaluate tire performance based on the lateral force friction coefficient. On the one hand, the present invention provides a tire characterization method based on the lateral force coefficient, which includes the following steps: selecting a test tire and a test section, and obtaining tire test data based on the test tire and the test section; setting a data quality analysis model, using the data quality analysis model to perform quality inspection on the tire test data, and obtaining the target test data of the test tire; constructing a tire characterization index analysis mechanism, and obtaining tire characterization parameters based on the tire characterization index analysis mechanism and the target test data; obtaining the characterization parameter characteristic information of the tire characterization parameters based on the target test data; obtaining the lateral force friction coefficient based on the characterization parameter characteristic information, and analyzing and evaluating the tire performance of the test tire according to the lateral force friction coefficient. The present invention can systematically analyze the performance of tires under different conditions, provide a scientific basis for the quantitative evaluation of tire performance, and make the evaluation results more objective and accurate.

[0006] Optionally, the setting of the data quality analysis model includes: dividing the tire test data into a plurality of sample data sets, and establishing the data quality analysis model according to the plurality of sample data sets;

[0007] The data quality analysis model satisfies the following relationship:

[0008] ,

[0009] in, represents the loss value of tire test data, Indicates the capacity of tire test data, Indicates tire test data The predicted label of the sample, Indicates tire test data The present invention divides the tire test data into multiple sample data sets, and establishes a data quality analysis model based on the sample data sets, which can more comprehensively evaluate the overall quality of the test data.

[0010] Optionally, the method of using the data quality analysis model to perform quality inspection on tire test data to obtain target test data for the test tire includes: obtaining a loss value of the tire test data through the data quality analysis model; performing quality inspection on the tire test data according to the loss value and obtaining a data quality inspection result; and obtaining the target test data for the test tire according to the quality inspection result. The present invention uses the data quality analysis model to perform quality inspection on tire test data, which not only improves data quality and analysis efficiency, but also ensures the accuracy of the analysis results.

[0011] Optionally, the construction of a tire characterization index analysis mechanism and obtaining tire characterization parameters based on the tire characterization index analysis mechanism and the target test data include: constructing a tire characterization index analysis mechanism based on tire characterization requirements; screening a first tire characterization index according to the tire characterization index analysis mechanism; analyzing the basic properties, correlation, common factor variance and total variance of the first tire characterization index according to the tire characterization index analysis mechanism; and obtaining tire characterization parameters by combining the basic properties, the correlation, the common factor variance and the total variance. The present invention analyzes the screened first characterization index, which helps to more comprehensively understand the impact of related indicators on tire performance, thereby more accurately evaluating the overall performance of the tire.

[0012] Optionally, the tire characterization method based on the lateral force coefficient further comprises: analyzing the component coefficients of the tire characterization parameters according to the tire characterization index analysis mechanism. The component coefficient analysis results of the present invention can reveal the contribution of different components or factors in tire performance, which helps to understand tire performance more deeply.

[0013] Optionally, the step of obtaining the characteristic parameter feature information of the tire characterization parameter based on the target test data comprises: setting a characteristic information extraction model; the characteristic information extraction model obtains the characteristic parameter feature information of the tire characterization parameter based on the target test data. The characteristic information extraction model of the present invention can extract key information closely related to the tire characterization parameter from complex target test data, thereby improving the efficiency and accuracy of data processing.

[0014] Optionally, the feature information extraction model satisfies the following relationship:

[0015]

[0016] in, Indicates The feature matrix of layer data output, Indicates Layer data ResNet The feature map matrix of the path, represents the addition operation, It indicates the function mechanism of entering the ResNet path after feature extraction through the Bottlenec structure during the information extraction process. Indicates The feature matrix of layer data output, Represents a merge operation. Indicates The feature map matrix of the DenseNe path in the layer data, It represents the function mechanism of entering the DenseNet path after feature extraction through the Bottleneck structure during the information extraction process. The model of the present invention processes the RESNET path and the DENSENET path in parallel, and can capture feature information at different levels at the same time, which helps to more comprehensively understand the complexity and diversity of tire characterization parameters.

[0017] Optionally, obtaining the lateral force friction coefficient according to the characterization parameter characteristic information includes: analyzing the relationship between the tire characterization parameter and the lateral force friction coefficient according to the characterization parameter characteristic information; establishing a lateral force friction coefficient calculation model based on the relationship and the tire characterization parameter;

[0018] The lateral force friction coefficient calculation model satisfies the following relationship:

[0019]

[0020] in, represents the lateral force friction coefficient, Indicates tire hardness. Indicates tire wear. The invention quantifies the correlation between the tire characterization parameter and the lateral force friction coefficient, so that the lateral force friction coefficient can be predicted more accurately.

[0021] Optionally, obtaining the lateral force friction coefficient according to the characterization parameter characteristic information, and analyzing and evaluating the tire performance of the test tire according to the lateral force friction coefficient comprises: obtaining the lateral force friction coefficient through the lateral force friction coefficient calculation model; analyzing and evaluating the tire performance of the test tire according to the lateral force friction coefficient. The present invention can more accurately calculate the lateral force friction coefficient by characterizing the parameter characteristic information, further ensuring the accuracy of the tire performance evaluation result.

[0022] In the second aspect, in order to efficiently execute the tire characterization method based on the lateral force coefficient provided by the present invention, the present invention also provides a tire characterization system based on the lateral force coefficient, including a processor, an input device, an output device and a memory, wherein the processor, the input device, the output device and the memory are interconnected, wherein the memory is used to store a computer program, the computer program includes program instructions, and the processor is configured to call the program instructions to execute the tire characterization method based on the lateral force coefficient as described in the first aspect of the present invention. The tire characterization system based on the lateral force coefficient of the present invention has a compact structure and stable performance, and can stably execute the tire characterization method based on the lateral force coefficient provided by the present invention, thereby improving the overall applicability and practical application capability of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 It is a flow chart of the tire characterization method based on lateral force coefficient of the present invention;

[0024] Figure 2 It is a schematic diagram of the relationship between the target test data accuracy and the number of data optimization iterations in the tire characterization method based on the lateral force coefficient of the present invention;

[0025] Figure 3 A schematic diagram of gravels with different principal components in the tire characterization method based on lateral force coefficient of the present invention;

[0026] Figure 4 It is a schematic diagram of a space component after rotation in the tire characterization method based on lateral force coefficient of the present invention;

[0027] Figure 5 It is a structural diagram of the tire characterization system based on the lateral force coefficient of the present invention. DETAILED DESCRIPTION

[0028] See also Figure 1In order to improve the efficiency and accuracy of tire performance evaluation, the present invention screens and analyzes the key performance indicators of the tire, and finally conducts in-depth analysis and evaluation of the tire performance based on the lateral force friction coefficient. The present invention provides a tire characterization method based on the lateral force coefficient, and the tire characterization method based on the lateral force coefficient includes the following steps:

[0029] S1. Select a test tire and a test section, and obtain tire test data based on the test tire and the test section. The specific implementation steps and contents are as follows;

[0030] 1. Tire test preparation stage.

[0031] According to the test requirements, representative tire samples are selected so that the test tires can cover different specifications, materials and designs: road sections that meet the test standards are selected to ensure that the road sections can fully reflect the performance of the tires in actual use.

[0032] 2. Tire testing and data collection phase.

[0033] The physical and mechanical properties of the tire are measured. The key performance indicators such as the wear and elasticity of the tire are mainly measured. In an optional embodiment, a wear test device is introduced, which can simulate the wear of the tire in actual use. By accurately controlling the test conditions such as load, rotation speed, wear medium, etc., the wear amount of the tire is accurately measured, so as to evaluate its wear resistance; an elastic test subsystem is set up to accurately measure the key elasticity indicators such as the elastic modulus and rebound rate of the tire. The above indicators directly reflect the recovery ability and deformation characteristics of the tire after being stressed, which are crucial for evaluating the durability and comfort of the tire; at the same time, in order to gain a deeper understanding of the mechanical properties of the tire, a multifunctional material testing machine is also used to test the mechanical properties of the test tire such as stretching, compression, and tearing, which helps to better obtain information such as the strength, toughness and durability of the tire.

[0034] In order to ensure the accuracy and comparability of the test results, a standardized test process was developed in the examples. From sample preparation, test condition setting to data recording and analysis, each step follows the established standards and specifications. On the other hand, big data and machine learning technology are used to record and store test data in a timely manner to provide data support for subsequent tire performance analysis.

[0035] In this embodiment, not only attention is paid to the collection and preservation of tire test data, but also a data collection feedback mechanism is set up. By regularly reviewing and analyzing tire test data, potential problems can be discovered in time and corresponding improvement measures can be taken to ensure real-time collection and effective storage of tire test data.

[0036] The specification parameters and material properties of the tire are recorded. In an optional embodiment, the specification parameters and material properties of the tire such as section width, rim diameter, ply, cord material and rubber properties are recorded in detail, where the specification parameters are as follows: the section width determines the contact area between the tire and the ground, which in turn affects the vehicle's handling, stability and grip; the width variation is due to the fact that tires of different models and uses have different section widths, which helps to adapt to different driving needs and road conditions; the rim diameter is the key dimension of the tire when installed on the wheel, which determines the tire's compatible model and wheel size; the adaptability range is because tires with different rim diameters are suitable for wheels of different sizes, which in turn affects the vehicle's appearance design and driving performance; the ply level of the tire usually indicates its load-bearing capacity and durability. The higher the ply level of the tire, the stronger the load-bearing capacity and is suitable for heavier vehicles or worse road conditions.

[0037] The material properties are as follows: Tire rubber is usually a mixture of natural rubber and synthetic rubber in material composition, and has excellent elasticity and wear resistance; rubber performance indicators such as hardness, tensile strength, tear strength and wear resistance determine the tire's grip, anti-skid and service life; modern tire rubber materials also focus on environmental performance, such as low rolling resistance, low noise, etc., to reduce the negative impact on the environment. When comprehensively describing the tire, you can also include its structural characteristics, design concepts, applicable scenarios and other information, so that you can have a more comprehensive and detailed understanding of the tire's specifications and material properties.

[0038] Testing the load and air pressure of the tire. In an optional embodiment, testing the load capacity and air pressure requirements of the tire can ensure the safety and stability of the tire during driving, wherein the load capacity test includes but is not limited to a static load test, a dynamic load test, and a load distribution test.

[0039] Static load test: Under specific temperature and pressure conditions, the tire is mounted on a standard test rim and the load is gradually increased until the rated load of the tire is reached or the tire is significantly deformed. At the same time, the deformation of the tire under different loads is recorded, including but not limited to radial and axial deformation, to evaluate the structural strength and stability of the tire.

[0040] Dynamic load test: simulates the actual driving conditions on the test section, and applies dynamic load to the tire at a certain speed and acceleration on the test equipment. At the same time, the temperature change, wear and stress distribution of the internal structure of the tire under dynamic load are monitored to evaluate the durability and fatigue resistance of the tire.

[0041] Load distribution test: Evaluate the performance of tires under different load distribution conditions, such as single-sided load, uneven load, etc. Through relevant tests, we can understand the performance of tires under complex load conditions and provide data support for their design and optimization.

[0042] The above pressure requirement tests include but are not limited to standard pressure test, pressure change test and long-term pressure retention test. Standard pressure test: At the rated pressure of the tire, check the tire's expansion, sealing performance and pressure stability to ensure that the tire can maintain a good shape and size at the standard pressure, and has sufficient rigidity to support the weight of the vehicle.

[0043] Air pressure change test: simulates the air pressure changes that may be encountered in actual driving, such as increased or decreased air pressure, and monitors the performance changes of the tire under the change of air pressure, including temperature, deformation and wear. Through the above test, the adaptability and stability of the tire under different air pressure conditions can be evaluated.

[0044] Long-term air pressure retention test: Over a certain period of time, which can be weeks or months, the tire's air pressure changes are continuously monitored to check the tire's air tightness, material aging, and the stability of its internal structure. Related tests help understand the tire's air pressure retention performance in long-term use and provide guidance for its maintenance and replacement.

[0045] Based on the above process and specific implementation steps, tire test data can be obtained. Furthermore, the method for obtaining tire test data in this embodiment is only an optional condition of the present invention. In other embodiments, the method for obtaining tire test data can be replaced according to the tire performance analysis requirements and tire data detection conditions, which helps to more accurately evaluate the performance of the tire, optimize the test process, and improve test efficiency and cost-effectiveness.

[0046] S2. Setting a data quality analysis model, using the data quality analysis model to perform quality inspection on tire test data, and obtaining target test data for the test tire. The specific implementation steps and contents are as follows:

[0047] Firstly, the tire test data is divided into multiple sample data sets.

[0048] In this embodiment, the tire test data is divided to form a plurality of sample data sets with different characteristics and attributes. The sample data sets not only include the performance data of the tire under different test conditions, but also cover a variety of key indicators such as load, air pressure, temperature, wear, etc. In order to ensure the validity and accuracy of the data, data preprocessing and cleaning are performed on each sample data set.

[0049] Next, data processing is performed based on the specific characteristics and requirements of the sample data set. For sample data sets with clear labeling or classification requirements, multiple machine learning algorithms such as supervised learning algorithms, support vector machines (SVMs) and random forests are selected to train based on known sample data sets, and learn to classify new data into known sample data sets, which helps detect and obtain target data sets.

[0050] For sample data sets without clear labels or classification requirements, unsupervised learning algorithms such as principal component analysis (PCA) and K-Means are used. These algorithms can extract useful features and attributes from sample data without prior knowledge, reduce data dimensions while retaining as much information as possible, and thus reveal the potential structure and relationships in tire data.

[0051] The embodiment constructs multiple sample data sets that satisfy the following relationship , where different samples represent different feature vector samples. Then, based on the sample data set for training and testing, the potential information and rules in the tire test data can be obtained, providing strong support for tire performance analysis.

[0052] Then, a data quality analysis model is established based on multiple sample data sets. The above data quality analysis model satisfies the following relationship:

[0053]

[0054] in, represents the loss value of tire test data, Indicates the capacity of tire test data, Indicates tire test data The predicted label of the sample, Indicates tire test data The detection label of the sample.

[0055] In the field of tire performance evaluation and testing technology, tire test data loss value refers to the quantitative index of performance degradation or loss of the tire under specific test conditions. It is a comprehensive index involving multiple aspects of the tire, which helps to further understand the performance characteristics of the tire, optimize tire design and improve tire quality.

[0056] Tire test data capacity refers to the total amount or scale of data collected, recorded or processed during tire performance testing or experiments. The above data includes but is not limited to the size, load capacity, air pressure, wear resistance, skid resistance, rolling resistance, etc. of the test tire.

[0057] The predicted labels of different samples in tire test data refer to the predicted results or classifications for specific samples when tire test data is processed and analyzed using machine learning or deep learning algorithms.

[0058] The detection labels of different samples in the tire test data refer to an identification or classification assigned to a specific sample during the tire detection or test process, which is used to indicate the performance or status of the sample in a certain aspect. In the embodiment, the above detection labels are derived based on the actual detection data and the corresponding detection standards, and have clear meanings and classification standards.

[0059] Next, the loss value of the tire test data is obtained through the data quality analysis model, and the tire test data is quality tested based on the loss value to obtain the data quality test result; and the target test data of the test tire is obtained based on the quality test result.

[0060] The data quality analysis model is used to evaluate the quality of the collected tire test data. The data loss value is calculated through a series of algorithms and rules. The above loss value, as a key quantitative indicator, can intuitively reflect the deviation or deficiency of the tire test data in terms of completeness, accuracy, consistency, etc.

[0061] Based on the tire test data loss value, the tire test data is quality checked. In the embodiment, corresponding data quality check standards and thresholds are set to ensure accurate identification of potential problems in the test data. By comparing the loss value with the quality check standard, the data quality check result can be obtained, and the data with poor quality can be corrected and processed based on the relevant check results.

[0062] After collecting, organizing and processing the data, we can obtain the target test data of the test tire. The above target test data is not only highly accurate and reliable, but also can comprehensively and truly reflect the performance of the tire under specific conditions, providing a solid foundation and strong support for subsequent tire performance analysis, optimization design and quality control.

[0063] Furthermore, in order to ensure the accuracy and validity of the target test data of the tested tires, data evaluation indicators are also set in the tire characterization method based on the lateral force coefficient to verify the target test data. The above data evaluation indicators mainly include precision (P), recall rate (R) and F1 score (F1).

[0064] The precision (P) satisfies the following relationship:

[0065]

[0066] in, Indicates the accuracy of the data. Indicates the number of named entities that are positive in the label that are predicted to be positive. Indicates the number of named entities predicted as positive that are labeled as negative.

[0067] The recall rate (R) satisfies the following relationship:

[0068]

[0069] in, represents the recall rate of the data, Indicates the number of named entities that are positive in the label that are predicted to be positive. Indicates the number of named entities that are predicted as negative in the positive class in the label.

[0070] The F1 score (F1) satisfies the following relationship:

[0071]

[0072] Among them, the F1 score of F1 data is Indicates the number of named entities that are positive in the label that are predicted to be positive. Indicates the number of named entities with negative labels that are predicted to be positive. Indicates the number of named entities that are predicted as negative in the positive class in the label.

[0073] The embodiment shows the relationship between the accuracy and the number of data optimization iterations. Figure 2 .based on Figure 2 It can be seen that the test data has significant changes after 50 iterations of training. As the number of iterations continues to increase, the gap between the test data and the preset target data gradually narrows, showing a trend of gradual approach.

[0074] After 50 data iterations, the data accuracy has gradually stabilized, indicating that the data quality analysis model has fully learned the characteristics and rules in the training data and can perform well on new, unseen data. After data iteration training and optimization updates, the test accuracy of the target test data reached an optimal value of 94%, which not only verified the effectiveness of the data quality analysis model, but also provided technical support for subsequent tire test data analysis and prediction.

[0075] Furthermore, in the present embodiment, the optimization method of tire target test data is only an optional condition of the present invention. In one or some other embodiments, the data optimization method can be replaced according to the tire performance evaluation objectives and actual application requirements. According to the tire performance evaluation standards and actual application scenarios, the data optimization method can be appropriately replaced or adjusted, so that the present invention can be widely used in the processing and analysis of various tire test data.

[0076] S3. Construct a tire characterization index analysis mechanism, and obtain tire characterization parameters based on the tire characterization index analysis mechanism and target test data. The specific implementation steps and contents are as follows:

[0077] A tire characterization index analysis mechanism is constructed based on tire characterization requirements, and the first tire characterization index is selected according to the tire characterization index analysis mechanism.

[0078] The tire characterization index analysis mechanism mainly includes the determination of tire characterization indexes and the construction of an index system, and uses a hierarchical analysis method to determine the weight of each index and analyze the score of the tire characterization index. In an optional embodiment, first, the tire characterization index is determined according to the characteristics and evaluation requirements of the tire. The above index is a key factor in evaluating tire quality. Then, a systematic and hierarchical index system is constructed based on the tire characterization index, which helps to evaluate tire performance more clearly and systematically.

[0079] In order to evaluate the quality of tires more scientifically, the analytic hierarchy process is used in the tire characterization index analysis mechanism to determine the weight of each index in the index system. As an effective multi-attribute decision-making method, the analytic hierarchy process can reasonably allocate weights according to the importance of each index and their relationship, thereby ensuring the objectivity and accuracy of the evaluation results.

[0080] Finally, based on the determined weights, each characterization indicator is scored and analyzed. The above process not only takes into account the actual performance of each indicator, but also fully combines the influence of the weights, so that the scoring results can truly and comprehensively reflect the quality and performance of the tire.

[0081] The tire characterization index analysis mechanism in the embodiment is a systematic, scientific and practical evaluation tool, which covers multiple links such as the determination of tire characterization indexes, the construction of an index system, the application of the hierarchical analysis method and the scoring analysis of characterization indexes, providing support for the tire performance characterization method.

[0082] Based on the tire characterization index analysis mechanism, the first characterization index of the tire was further screened out. The above key indicators mainly refer to the physical and mechanical properties of the tire, such as tensile strength, hardness, abrasion, tear strength and elasticity, which together constitute an important basis for evaluating tire performance.

[0083] Next, a corresponding indicator system was constructed with the tire's tensile strength, hardness, abrasion, tear strength and elasticity as the core. It not only covers the key elements of tire quality evaluation, but also makes the evaluation process clearer and more organized through scientific classification and hierarchy.

[0084] The basic properties, correlation, common factor variance and total variance of the first tire characterization index are analyzed according to the tire characterization index analysis mechanism; and the tire characterization parameters are obtained by combining the basic properties, correlation, common factor variance and total variance.

[0085] In an optional embodiment, the SPSS method is used to perform principal component analysis on key characterization indicators of the physical and mechanical properties of the tire. The analysis results are shown in Table 1.

[0086] Table 1 Statistical table of characterization indicators

[0087] Characterization index of physical and mechanical properties average value Standard Deviation Number of cases analyzed Tensile Strength 18.720 2.1248 15 hardness 58.400 4.0143 15 wear and tear 0.20180 0.133354 15 Tear strength 68.800 17.2511 15 elasticity 37.600 9.5229 15

[0088] Based on Table 1, it can be seen that the embodiment conducts comprehensive data statistics on the physical and mechanical properties of the tire, which mainly involves five key characterization indicators: tensile strength, hardness, abrasion, tear strength and elasticity. All indicators are analyzed based on 15 identical test tires, which helps to ensure the reliability and comparability of the analysis results.

[0089] 1. Average level analysis

[0090] The average value of tensile strength is 18.720, which reflects the overall strength level of the tire in the tensile test.

[0091] The average hardness value is 58.400, which reveals the hardness characteristics of the tire material.

[0092] The average value of wear is 0.20180, which reflects the durability of the tire in the wear test.

[0093] The average tear strength value was 68.800, indicating the tire's resistance to tearing in the test.

[0094] The average value of elasticity is 37.600, which reflects the elasticity and recovery ability of the tire.

[0095] 2. Discreteness analysis

[0096] The standard deviation of the tensile strength is 2.1248, which shows that the data of the tensile strength are relatively concentrated and the difference in tensile strength between different tested tires is small.

[0097] The standard deviation of hardness is 4.0143. Compared with tensile strength, the data dispersion of hardness has increased, but it is still within a relatively controllable range.

[0098] The standard deviation of tear strength is as high as 17.2511, which shows that the tear strength data is relatively scattered and the tear strength between different test tires varies significantly.

[0099] The standard deviation of wear is 0.133354, indicating that the data dispersion of wear is low and the performance of each test tire in the wear test is relatively consistent.

[0100] The standard deviation of elasticity is 9.5229, and its dispersion is between tensile strength and tear strength, indicating that the differences in elasticity among the tested tires are moderate.

[0101] Based on the statistical information of the above tire physical and mechanical performance indicators, we can fully understand the average level and dispersion of each indicator. The data of tensile strength and abrasion are relatively concentrated, indicating that the above indicators are highly consistent among the tested tires; while the data of hardness and elasticity are moderately dispersed, and the data of tear strength are relatively scattered. The relevant statistical results provide a solid foundation for the subsequent principal component analysis and data interpretation of tires, and help to have a deeper understanding of the physical and mechanical properties of tires and their differences between different test tires.

[0102] Then, a correlation matrix analysis was conducted. That is, a correlation matrix was constructed for each index of the physical and mechanical properties of the tire, including tensile strength, hardness, abrasion, tear strength and elasticity. For specific matrix information, please refer to Table 2.

[0103] Table 2 Correlation matrix of characterization indicators

[0104]

[0105] The linear correlation between the indicators can be clearly observed from Table 2. From the correlation matrix, it can be seen that the indicators of tire physical and mechanical properties generally show weak correlation, specifically, the correlation coefficients between tensile strength and hardness, abrasion, tear strength and elasticity are all low, indicating that the above indicators lack obvious linear correlation in the data. Similarly, there is no significant linear correlation between hardness, abrasion, tear strength and elasticity.

[0106] The above weak correlation means that each indicator is independent to a certain extent, that is, different indicators reflect different aspects of the physical and mechanical properties of the tire, and there is no obvious interdependence between these aspects, which provides a basis for subsequent tire performance analysis. At the same time, when analyzing tire performance, it is necessary to comprehensively consider multiple indicators to comprehensively evaluate the physical and mechanical properties of the test tire.

[0107] Then, the common factor variance was calculated, see Table 3 for details.

[0108] Table 3 Common factor variance of characterization indicators

[0109]

[0110] From the common factor variance results in Table 3, we can clearly see the information retention of each index of tire physical and mechanical properties during the principal component extraction process.

[0111] Tensile strength: In the initial state, the common factor variance of tensile strength is 1.000, indicating that its information is intact. After principal component extraction, the common factor variance drops to 0.710, which means that 71% of the information of tensile strength is successfully extracted, retaining most of the key information.

[0112] Hardness: For the hardness index, the initial common factor variance and the common factor variance after extraction are both 1.000, which further shows that all the information of hardness is fully retained in the principal component extraction process without any information loss or omission.

[0113] Wear: The initial common factor variance of wear is also 1.000, showing the integrity of its information. After principal component extraction, the common factor variance is reduced to 0.936, that is, 93.6% of the wear information is effectively extracted, and the information retention is high.

[0114] Tear strength: The common factor variance of tear strength after principal component extraction is 0.938, indicating that 93.8% of its information is successfully extracted. This result shows that the information of tear strength is well retained in principal component analysis.

[0115] Elasticity: For the elasticity index, the common factor variance after extraction is 0.886, which means that 88.6% of the elasticity information is extracted by the principal component. Although it is slightly lower than other indicators, it still maintains a high information retention rate.

[0116] In summary, most of the information of tire characterization indicators has been effectively extracted through the principal component analysis method. Among them, the information of hardness is completely retained, and the information retention rate of tensile strength, abrasion, tear strength and elasticity is also more than 88%. The above results further provide a strong basis for understanding the physical and mechanical properties of tires and their principal component structure.

[0117] The principal component analysis method used in the embodiment shows a remarkable effect in extracting information on the physical and mechanical performance indicators of the tire, especially for the hardness indicator, whose information is almost completely extracted, with an extraction rate of nearly 100%. The relevant results show that the principal component analysis method is very effective in capturing and retaining key information of the hardness indicator.

[0118] At the same time, for other indicators such as tensile strength, abrasion, tear strength and elasticity, the principal component analysis method also showed a high ability to extract information. The information extraction ratio of key characterization indicators exceeded 88%, showing the powerful function of principal component analysis in summarizing and explaining the main information of indicators.

[0119] Next, the total variance was explained, see Table 4 for details.

[0120] Table 4 Total variance information table of characterization indicators

[0121]

[0122] Through the total variance explanation in Table 4, we can deeply analyze the explanatory power of the principal components on the data.

[0123] 1. Initial eigenvalue analysis

[0124] The first component: Its initial eigenvalue is 1.846, accounting for 36.925% of the total variance, indicating that the first component dominates the data and can explain more than one-third of the data variation.

[0125] Second component: The initial eigenvalue is 1.597, and the variance percentage is 31.932%. The second component also has significant data explanatory power, and together with the first component, it explains more than two-thirds of the data variation.

[0126] The third component: The initial eigenvalue is 1.027 and the variance percentage is 20.546%. Although the explanatory power of the third component is slightly weaker than that of the first two components, it can still explain more than one-fifth of the data variation.

[0127] Based on this, we can see that the first three components together explain most of the variation in the data, showing strong data interpretation capabilities.

[0128] 2. Extraction of Sum of Squares of Loadings Analysis

[0129] The results of the extracted sum of squares of loadings were similar to the initial eigenvalues, further validating the data interpretation capabilities of the first three principal components.

[0130] Cumulative variance percentage: The cumulative variance percentages of the first three components are 36.925%, 68.857%, and 89.403%, respectively, indicating that the three principal components extracted can gradually and cumulatively explain most of the data variation, especially the first two principal components, which together explain more than two-thirds of the data variation.

[0131] 3. Rotational load sum of squares analysis

[0132] The sum of squares of the loadings after rotation shows the changes in the explanatory power of each component after rotation.

[0133] After rotation, the components: the sum of squares of the rotated loadings of the first component is 1.785, and the variance percentage is 35.690%; the sum of squares of the rotated loadings of the second component is 1.592, and the variance percentage is 31.843%; the sum of squares of the rotated loadings of the third component is 1.093, and the variance percentage is 21.869%. The explanatory power of the rotated components is slightly adjusted, but the overall cumulative variance percentage remains unchanged at 89.403%.

[0134] In summary, three principal components were extracted through principal component analysis, which together explained most of the variation in the data. The rotated components were more explicit in their explanatory power and provided a clearer perspective on data interpretation. The above principal components not only retained most of the information of the original data, but also made data analysis and interpretation more concise and efficient through dimensionality reduction. We further learned that the three components extracted by principal component analysis can better summarize the information of the data and explain most of the variation in the data.

[0135] Furthermore, a scree plot is drawn to show the proportion of explained variance of each principal component, which helps to intuitively understand the number and importance of principal components. For details, see Figure 3 , where the component number represents the serial number of each principal component extracted from the PCA analysis, and the numbering starts from 1 in the embodiment. The scree plot is usually presented as a line graph, with the horizontal axis representing the principal component number, i.e., the component number, and the vertical axis representing the variance proportion or eigenvalue explained by each principal component. As the principal component number increases, the explained variance proportion usually decreases gradually, forming a scree-like graph, which is shown by Figure 3 The distribution of the principal components' ability to explain data variation can be intuitively understood.

[0136] Eigenvalue decreasing trend: The scree plot clearly shows that the eigenvalues ​​of the principal components tend to decrease rapidly at the beginning, especially the first three principal components, whose eigenvalues ​​are relatively large, significantly different from the subsequent principal components.

[0137] Importance of the first three principal components: The first three principal components occupy the main part of the eigenvalues, which means that the three principal components can explain most of the variation in the data. From the perspective of data interpretation, the three principal components contain key information in the original data.

[0138] Characteristics of subsequent principal components: Starting from the fourth principal component, the decrease in eigenvalues ​​becomes gentle and gradually approaches 0, indicating that the subsequent principal components have relatively weak explanatory power for data variation, and the corresponding information they carry is more redundant or secondary.

[0139] Then, the basis for selecting the principal components can be obtained. Based on the above analysis results, in principal component analysis, in order to simplify the data and retain its main features, the first three principal components can be selected and extracted. This strategy can not only effectively reduce the dimension of the data, but also ensure that the main information of the tire test data is retained.

[0140] In terms of data interpretation efficiency, selecting the first three principal components can summarize and explain the main features of the original data in a more concise way, which not only improves the efficiency of data interpretation but also helps to understand the inherent structure of the data more deeply.

[0141] In summary, the scree plot provides an intuitive method to evaluate the explanatory power of each principal component on data variation. By observing the downward trend of the eigenvalues, the first three principal components can be reasonably selected to represent the information of the original data, thereby achieving data dimensionality reduction and retaining the main features. The above strategy has important application value in principal component analysis.

[0142] Furthermore, the component coefficients of the tire characterization parameters are analyzed according to the tire characterization index analysis mechanism.

[0143] Based on the above analysis results, the composition of key characterization indicators is further analyzed, please see Table 5 for details.

[0144] Table 5 Component matrix information table representing indicators

[0145]

[0146] The information in Table 5 can be used to conduct an in-depth analysis of the characteristics represented by each principal component.

[0147] First ingredient characteristics:

[0148] In the first component, hardness presents a significant negative load, while tear strength and elasticity have certain positive loads, while the loads of abrasion and tensile strength are relatively light. The above characteristics show that the first component largely reflects the properties closely related to hardness, while also incorporating some information about tear strength and elasticity. Therefore, it can be considered that the first component mainly depicts the balance between hardness and toughness of the tire, which is crucial to the overall performance of the tire.

[0149] Second ingredient characteristics:

[0150] In the second component, tensile strength and elasticity show significant positive loads, while tear strength has a large negative load, and the loads of other variables are relatively small. The above characteristic pattern suggests that the second component is mainly related to the specific combination of tensile strength and elasticity. Specifically, it further reflects the positive correlation between the tensile performance and elasticity of the tire, that is, when the tensile performance of the tire is strong, its tear strength will be relatively weak, which helps to understand the mechanical properties of the tire more deeply.

[0151] Characteristics of the third component:

[0152] In the third component, wear shows a significant positive load, while the loads of other variables are relatively small. The above characteristics indicate that the third component mainly focuses on the characteristic of wear, which serves as an independent dimension to represent the wear performance of the tire, thereby providing a new perspective to examine the durability and service life of the tire.

[0153] In summary, the analysis of the component matrix in Table 5 further reveals the characteristics represented by each principal component and its relationship with tire performance. The first component mainly reflects the balance between hardness and toughness of the tire; the second component is closely related to the positive relationship between the tensile properties and elasticity of the tire; and the third component mainly represents the wear performance of the tire. The above findings can provide a deeper understanding of the mechanical properties of the tire.

[0154] Furthermore, the above component matrix information is rotated and analyzed to better analyze the relevant information. Please see Table 6 for details.

[0155] Table 6 Rotated component matrix information table

[0156]

[0157] The extraction method is principal component analysis; the rotation method is Kaiser normalized maximum variance method; the rotation has converged after 6 iterations. The Kaiser normalized maximum variance rotation method is used in the embodiment to conduct a more in-depth principal component analysis on the tire mechanical performance indicators in Table 5. Table 6 provides a clearer perspective, effectively separating different performance dimensions, making each principal component more physically meaningful and explanatory.

[0158] Principal component 1 is a trade-off between hardness and toughness. Principal component 1 reflects the opposing relationship between tire hardness and tear strength. In the embodiment, hardness and tear strength are negatively correlated in principal component 1, that is, the higher the hardness, the lower the tear strength. The above relationship can be explained as a trade-off between the hardness and toughness of the tire material. Among them, hardness (0.867) occupies a dominant position in principal component 1, and its positive correlation indicates that hardness is a positive contributing factor to this principal component; while the negative correlation of tear strength (-0.938) reveals its negative impact in this principal component, that is, an increase in tear strength will lead to a decrease in the performance of principal component 1. Therefore, principal component 1 can be regarded as a performance dimension that comprehensively affects hardness and tear strength, in which the hardness index occupies a core position.

[0159] Principal component 2 is a comprehensive reflection of elasticity and flexibility. Principal component 2 mainly reflects the performance characteristics of tire elasticity and tensile strength, and is inversely proportional to hardness, which means that tire materials with high elasticity and tensile strength have lower hardness. Therefore, principal component 2 can be regarded as a comprehensive reflection of tire elasticity and flexibility. The above principal components provide important information about the flexibility and elasticity of tire materials, which helps to understand the deformation and recovery ability of tires under stress.

[0160] Principal component 3 is independent wear performance. Principal component 3 focuses on the wear resistance of the tire and has relatively little correlation with other mechanical properties. This principal component can be interpreted as an independent dimension specifically for tire wear performance. Principal component 3 can be used to more accurately evaluate the wear resistance of the tire, thereby providing a basis for improving the durability and service life of the tire in actual applications.

[0161] In summary, the principal component analysis of the tire mechanical performance indicators was carried out using the Kaiser normalization maximum variance rotation method. The rotated component matrix can separate the different indicator dimensions more clearly and make the indicators in each component more physically meaningful and explanatory, which can better understand the relationship between the various mechanical properties of the tire and the contribution of different indicators to the tire performance.

[0162] In this embodiment, a component diagram in the rotated space is drawn based on the information in Table 6. For details, see Figure 4 .based on Figure 4 The unique distribution patterns of the tire's mechanical performance indicators on the three components can be clearly observed. The above distribution not only reveals the intrinsic connection between the indicators, but also provides an intuitive basis for optimizing tire performance.

[0163] There is a significant correlation between hardness and tensile strength in component 1. In the dimension of component 1, the hardness index shows an obvious positive distribution characteristic, which indicates that hardness occupies an important position in component 1. At the same time, tensile strength also shows a certain positive distribution in this component. Although its influence is slightly less than that of hardness, it cannot be ignored. The above findings reveal the close connection between hardness and tensile strength in tire performance, and indicate that the two have a synergistic effect in material composition or structural design.

[0164] The comprehensive manifestation of elasticity and tensile strength in component 2. Based on component 2, it can be observed that both elasticity and tensile strength show a significant positive distribution. Among them, the positive distribution of elasticity is particularly prominent, indicating that elasticity occupies a dominant position in component 2. At the same time, although the distribution range of tensile strength is small, it still maintains a positive trend, further strengthening the correlation between elasticity and tensile strength. On the other hand, hardness shows a negative distribution on component 2, which further indicates that there is an opposing or complementary relationship between the elasticity-tensile strength characteristics.

[0165] Independent dimension of wear performance in component 3. The wear index in component 3 is the core of this component with its extremely positive distribution. This feature clearly shows that component 3 focuses on the wear performance of the tire, and wear has a higher independence and importance in component 3 compared to other mechanical performance indicators. At the same time, the distribution of tear strength in component 3 is negligible, further emphasizing the uniqueness of wear as a characterization indicator.

[0166] In summary, through analysis Figure 4 The distribution characteristics of the mechanical performance indicators of the tire in the three components and their mutual relationships can be clearly grasped. Hardness, elasticity and tensile strength, and wear are dominant in components 1, 2, and 3 respectively. The above findings not only help to understand tire performance, but also provide a basis for the optimal design of tires.

[0167] Furthermore, in the present embodiment, the method for obtaining and analyzing tire characterization parameters is only an optional condition of the present embodiment. In one or some other embodiments, the method for analyzing tire characterization parameters may be adjusted according to the actual situation of the tire and the tire performance evaluation target. The method for obtaining and analyzing tire characterization parameters in the embodiment is an optional condition, which makes the method of the present invention have better flexibility and adaptability. Adjusting the analysis method according to the actual situation of the tire and the performance evaluation target can improve the accuracy and efficiency of the analysis.

[0168] S4, obtaining characteristic parameter feature information of tire characteristic parameters based on target test data, the specific implementation steps and contents are as follows:

[0169] In this embodiment, a feature information extraction model is provided; the feature information extraction model obtains the characterization parameter feature information of the tire characterization parameter based on the target test data.

[0170] The core function of the feature information extraction model is to extract the feature information of tire characterization parameters from the target test data. The feature information extraction process is guided by the model. The feature information extraction model satisfies the following relationship:

[0171]

[0172] in, Indicates The feature matrix of layer data output, Indicates The feature map matrix of the ResNet path in the layer data, represents the addition operation, It indicates the function mechanism of entering the ResNet path after feature extraction through the Bottlenec structure during the information extraction process. Indicates The feature matrix of the layer data output, Represents a merge operation. Indicates The eigenvalue matrix of the DenseNe path in the layer data, It represents the function mechanism of entering the DenseNet path after feature extraction through the Bottleneck structure during the information extraction process.

[0173] The feature matrix output by different layers of data can be used to describe or represent the feature set of input data. The above features are obtained through layer-by-layer transmission, transformation and extraction. Each layer performs different forms of nonlinear transformation or feature extraction on the input data, thereby gradually revealing the inherent structure and pattern of the data. In the embodiment, different layers of data can be used to extract data for different characterization indicators. The above feature matrix not only contains the key information extracted from the input data, but also contains other feature information obtained through model learning and useful for tire performance prediction.

[0174] The feature mapping matrix of the RESNET path in different layers of data in deep learning refers to the feature representation matrix processed by the residual block in different levels of the ResNet model. The core idea of ​​ResNet is to introduce a residual structure, that is, each residual block contains a direct connection, an identity mapping and a residual path. The above structure allows the model to learn the residual between input and output during training, which helps to solve problems such as gradient disappearance and gradient explosion in deep networks.

[0175] In the context of mathematics, computer science, and deep learning, the notation Represents the addition operation, symbol Represents a merge operation.

[0176] In the information extraction process, after feature extraction is performed through the BOTTLENEC structure, the function mechanism entering the RESNET path refers to that after feature extraction is performed through the Bottleneck structure in the above process, entering the ResNet path is a composite function. This function first reduces the dimension, extracts features, and increases the dimension of the data through the Bottleneck structure, and then passes the result as input to the residual block in the ResNet path. The function mechanism in the residual block further processes the above data to generate a more accurate and rich feature representation, which helps the model learn more complex and abstract feature information from the input data.

[0177] After feature extraction through the BOTTLENECK structure in the information extraction process, the function mechanism of entering the DENSENET path means that the input data is first reduced in dimension and feature extracted through the Bottleneck structure to obtain an output with a small number of channels but rich in feature information; the output of the Bottleneck structure is used as the input of the next layer in the Dense Block, and the features are fused with the outputs of all previous layers. The above process is closely connected so that each layer can receive the feature information of all previous layers; as the number of layers in the Dense Block increases, the feature information is continuously transmitted and updated to form a more complex and abstract feature representation, which will be used for subsequent classification, regression or other tasks. The above process helps the model learn more complex and abstract feature information from the input data, thereby improving the actual performance of the model.

[0178] Based on the feature information extraction model, it can be seen that the feature matrix output by different layers of data is merged from two parts. One part is the feature mapping matrix of the RESNET path in the previous layer of data, which is obtained through a merging operation and a specific function mechanism. The above function mechanism is also in the process of information extraction. After feature extraction through the BOTTLENECK structure, the extracted features are sent to the DENSENET path.

[0179] The feature information extraction model can fully utilize the information in the target test data, process the RESNET and DENSENET paths in parallel, extract features through the BOTTLENECK structure, and generate more comprehensive and accurate tire characterization parameter feature information.

[0180] In this embodiment, the feature enhancement advantages of ResNet and DenseNet are cleverly combined to construct a feature information extraction model, which not only effectively avoids problems such as gradient disappearance, but also reduces the total number of network parameters, significantly reduces computational complexity and memory usage, and is conducive to real-time processing of large-scale data.

[0181] The above model divides the output into two parallel streams: one enters the ResNet path, and the other flows into the DenseNet path. The two paths generate feature representations respectively and merge them before entering the next unit structure, which promotes the cross-fusion of feature information and achieves more efficient and comprehensive feature extraction. At the same time, the feature mapping of each unit structure depends on the output of the previous unit, forming a feature accumulation and optimization cycle, further improving the accuracy and robustness of feature extraction.

[0182] Furthermore, the method for obtaining characteristic parameter information in the present embodiment is only an optional condition of the present embodiment. In one or some other embodiments, the method for obtaining parameter characteristic information may be adjusted according to the actual needs of characteristic information and the information standard for performance evaluation. By setting the method for obtaining characteristic information as an optional condition, it can be ensured that the method of the present invention can flexibly adapt to different application scenarios and needs. When faced with new data sets, task requirements or performance evaluation standards, the feature extraction strategy can be conveniently adjusted to better meet actual needs.

[0183] S5. Obtain the lateral force friction coefficient according to the characteristic parameter information, and analyze and evaluate the tire performance of the test tire according to the lateral force friction coefficient. The specific implementation steps and contents are as follows:

[0184] The component coefficients of the characterization parameter indicators are analyzed based on the characteristic information of the characterization parameters. Please see Table 7 for details.

[0185] Table 7 Score coefficient matrix

[0186]

[0187] The scores of key components (main components) in the tire samples can be calculated based on the information in Table 7. The relevant score results not only intuitively reflect the performance of the tire in terms of tensile strength, hardness, abrasion, tear strength and elasticity, but also provide a comprehensive tire performance evaluation framework, through which the tire can be analyzed and evaluated in a more detailed and in-depth manner.

[0188] Principal component analysis (PCA) is a powerful data analysis tool used in this embodiment to reveal the intrinsic relationship between different tire performance indicators. Using the principal component analysis method, three main principal components were found, representing the opposing relationship between tire hardness and tear strength, performance characteristics related to elasticity and tensile strength, and characteristics closely related to wear resistance. The extraction of the above principal components can understand and analyze the complex composition of tire performance in a more concise and intuitive way.

[0189] Among the performance indicators of tire performance, hardness, elasticity and wear can better characterize tire performance. Among them, hardness not only directly affects the wear resistance and driving stability of the tire, but is also closely related to properties such as tear strength; elasticity determines the deformation ability and grip of the tire when subjected to force, and is an important manifestation of the dynamic performance of the tire; and wear is directly related to the service life and replacement cycle of the tire. Therefore, the selection of the three parameters of hardness, elasticity and wear is of great significance to the design, production and performance evaluation of tires.

[0190] Finally, quantitative evaluation and comprehensive comparison of tire performance based on the scoring results can provide guidance for tire design and production. In an optional embodiment, key parameters such as hardness, elasticity and wear of the tire can be adjusted according to the target usage scenario and performance requirements to achieve the best performance balance, which helps to improve the overall performance of the tire and user satisfaction.

[0191] Hardness, elasticity and wear are key parameters for characterizing tire performance, and play a vital role in tire design, production and performance evaluation. In the embodiment, a feature information extraction model is set to mine and extract features from the data information of the three parameters, further revealing the intrinsic connection between the key parameters and tire performance, which not only helps to have a deeper understanding of the performance characteristics of tires, but also provides a scientific basis for the optimal design and performance improvement of tires.

[0192] The relationship between the tire characterization parameters and the lateral force friction coefficient is analyzed according to the characteristic information of the characterization parameters, and then a lateral force friction coefficient calculation model is established based on the relationship and the tire characterization parameters.

[0193] In an optional embodiment, the lateral force model coefficient actually measured by the test tire on a specific test section is used as the core data, and the linear regression analysis module in the SPSS statistical software is used for in-depth analysis. A series of key conclusions are obtained through SPSS linear regression analysis, which reveal the intrinsic relationship between the lateral force model coefficient and tire performance.

[0194] During the analysis, it was found that there is a linear relationship between the lateral force model coefficient and key performance indicators such as tire grip and handling stability. That is, as the lateral force model coefficient increases, the tire grip shows an increasing trend, and the handling stability is also improved accordingly. Based on this, the assumption of this embodiment is verified and a reference is provided for tire design and performance optimization.

[0195] See Table 8 for tire test data of the tires tested on the test section.

[0196] Table 8 Test section data statistics table

[0197]

[0198] Based on the information in Table 8, we know that this implementation sets up 3120 samples, which involve four key variables: dependent variable (lateral force friction coefficient) and three independent variables (hardness, wear, and elasticity). Statistical analysis of the above variables yields the following key data characteristics:

[0199] First, the average value of the lateral force friction coefficient is 54.4053, and the standard deviation is 5.2272. This data shows that the distribution of the lateral force friction coefficient in the test tire samples is relatively uniform, but there is a certain degree of discreteness; the standard deviation reflects the fluctuation range of the data, that is, the lateral force friction coefficient values ​​of most test tires fluctuate around the average value, but there are also some test tires with large deviations.

[0200] Secondly, the average value of hardness is 58.4615 and the standard deviation is 4.1631. Compared with the lateral force friction coefficient, the standard deviation of hardness is smaller, which means that the distribution of hardness data in the sample is more concentrated, that is, the hardness values ​​of most tested tires are closer to the average value, and the fluctuation range is relatively small.

[0201] The average value of the wear data is 0.1901, and the standard deviation is 0.1346. Since the average value of wear is small, its standard deviation is also relatively small, which further shows that the fluctuation range of wear data in the sample is relatively narrow, that is, the wear values ​​of most tested tires are close to the average value, and the difference is not large.

[0202] Finally, the mean value of elasticity is 35.3846 and the standard deviation is 7.8020. Compared with other variables, the standard deviation of elasticity is larger, indicating that the distribution of elasticity data in the sample is more dispersed, which further indicates that the elasticity differences between different observations are relatively large and there is greater volatility.

[0203] Furthermore, the correlation of the test section data is analyzed, as shown in Table 9.

[0204] Table 9 Correlation analysis table of test section data

[0205]

[0206] Table 9 reveals four key variables: lateral force friction coefficient, hardness, wear and elasticity, as well as the correlations between the variables and their statistical significance.

[0207] First, the Pearson correlation coefficient between the lateral force friction coefficient and hardness is as high as 0.820, showing a positive correlation trend, and the significance level reaches 0.000 in the one-tail test, indicating that the positive correlation between the two is extremely significant. That is, as the hardness increases, the lateral force friction coefficient also shows a corresponding upward trend, indicating that hardness is an important factor affecting the lateral force friction coefficient.

[0208] Secondly, the correlation between the lateral force friction coefficient and wear shows a negative trend, with a correlation coefficient of -0.152 and a significance level of 0.000 in the one-tailed test, indicating that the correlation between the two is significant, which means that with the increase of wear, the lateral force friction coefficient shows a downward trend, which is related to the changes in surface characteristics after tire wear.

[0209] Finally, the correlation between the lateral force friction coefficient and elasticity is also significant, but in a negative direction, with a correlation coefficient of -0.675. Its significance level is also 0.000 in the one-tailed test, indicating that the correlation between the two is very strong, that is, with the increase of elasticity, the lateral force friction coefficient shows a significant downward trend, which is related to the deformation characteristics of the elastic material when subjected to force, and thus affects the contact and friction between the tire and the ground.

[0210] The data in Table 9 not only reveals the direction and strength of the correlation between the four variables, but also confirms the statistical reliability of the correlation through the significance level test, providing important information support and basis for tire performance and its influencing factors.

[0211] At the same time, the prediction of the test section data is analyzed, please see Table 10 for details.

[0212] Table 10 Prediction analysis table of test section data

[0213]

[0214] Model 1 refers to the lateral force friction coefficient calculation model.

[0215] According to the data in Table 10, it can be clearly seen that the linear relationship between the independent variables elasticity, wear, hardness and the dependent variable lateral force friction coefficient is very close, and the R value is 0.853. This further shows that there is a significant linear correlation between the independent variables and the dependent variables, which provides a basis for the establishment of the coefficient calculation model.

[0216] Furthermore, the coefficient of determination is 0.727, which means that the independent variable can explain 72.7% of the variability of the dependent variable, indicating that the lateral force friction coefficient calculation model has a high degree of fit to the data and can well analyze the impact of the independent variable on the dependent variable. At the same time, the adjusted R also remains at a level of 0.727, further verifying the feasibility of the model.

[0217] The standard estimated error is 2.73297. The above value can measure the average difference between the model prediction value and the actual observation value. If the error is small, it means that the model has strong predictive ability and can predict the value of the dependent variable more accurately.

[0218] In terms of degrees of freedom, degree of freedom 1 is 3, which corresponds to the number of independent variables in the model, three independent variables of elasticity, wear, and hardness, plus a constant term, but a constant term needs to be subtracted when calculating the degrees of freedom, so it is 3. Degree of freedom 2 is 3116, which is obtained by subtracting the number of independent variables 3 from the sample size 3120 and then subtracting 1, indicating that there is enough sample size in the embodiment to support regression analysis.

[0219] Finally, the significant F change is 0.000. The above value further confirms that the model is highly significant as a whole, that is, the independent variable has a significant impact on the dependent variable.

[0220] The coefficients of the test section are further analyzed, see Table 11 for details.

[0221] Table 11 Coefficient analysis table of test section data

[0222]

[0223] Model 1 refers to the lateral force friction coefficient calculation model.

[0224] According to the information in Table 11, the estimated value of the constant term is 12.766, which is statistically significant because it represents the expected level of the dependent variable lateral force friction coefficient when all independent variables hardness, wear, and elasticity are taken as 0, and thus it can be expected that the value of the lateral force friction coefficient is 12.766.

[0225] Furthermore, the standard error of the constant term is 1.255. The above numerical value measures the accuracy of the constant term estimate. The smaller the standard error, the more reliable the constant term estimate. The standard error in the embodiment is relatively small, indicating that the constant term estimate is relatively accurate.

[0226] In addition, the t value is 10.170, which is a statistic that measures the significance level of the constant term in the model. The larger the t value, the more significant the contribution of the constant term in the model. The t value in the embodiment is high, indicating that the constant term plays a very significant role in the model.

[0227] Finally, the significance level is 0.000, which is an important indicator for measuring whether the constant term is statistically significant. The smaller the significance level, the higher the significance of the constant term in the model. In the embodiment, the significance level is 0.000, which is a very small value indicating that the constant term is highly significant statistically.

[0228] In order to more comprehensively explain the coefficients, statistical significance, correlations, and collinearity indicators of each independent variable, the specific analysis is as follows:

[0229] 1. Independent variable coefficient and statistical significance. Hardness: The unstandardized coefficient is 0.830. This means that when other independent variables remain unchanged, for every unit increase in hardness, the lateral force friction coefficient will increase by 0.830 units accordingly; the standardized coefficient is 0.661, which can be used to compare the relative size of the influence of different independent variables on the dependent variable. Hardness has a higher influence among the independent variables; the statistically significant t value is 48.992, and the significance level is 0.000, indicating that the influence of hardness on the lateral force friction coefficient is statistically extremely significant and is a key predictor variable.

[0230] Wear: The unstandardized coefficient is -7.690, that is, for every unit increase in wear, the lateral force friction coefficient will decrease by 7.690 units; the standardized coefficient is -0.198. Although the absolute value is relatively small, it still shows that wear has a significant effect on the dependent variable; the statistically significant t value is -20.840, and the significance level is 0.000, indicating that the effect of wear on the lateral force friction coefficient is also very significant.

[0231] Elasticity: The unstandardized coefficient is -0.153, indicating that for every unit increase in elasticity, the lateral force friction coefficient will decrease by 0.153 units; the standardized coefficient is -0.229. Among the independent variables, the impact of elasticity on the dependent variable is also significant; the statistically significant t value is -16.800, and the significance level is 0.000, which further confirms the high statistical significance of elasticity on the lateral force friction coefficient.

[0232] 2. Correlation analysis. Zero-order correlation: reflects the original correlation between the independent variable and the dependent variable, among which the zero-order correlation between hardness and lateral force friction coefficient is 0.820.

[0233] Partial correlation: The correlation between the independent variable and the dependent variable when other independent variables are controlled. For example, the partial correlation of hardness is 0.660, indicating that after excluding the influence of other variables, there is still a strong correlation between hardness and lateral force friction coefficient.

[0234] Partial correlation: The degree to which an independent variable contributes to the dependent variable when considering the impact of all independent variables on the dependent variable, such as the partial correlation of hardness is 0.459.

[0235] 3. Collinearity analysis. Tolerance can be used to measure the degree of collinearity between independent variables. The tolerance of hardness is 0.459, indicating that the degree of collinearity between hardness and other independent variables is relatively low, and the tolerances of all independent variables in the embodiment are within a reasonable range.

[0236] Furthermore, based on the above analysis results, a mathematical model of the relationship between the lateral force friction coefficient, hardness, wear and elasticity is established, that is, the lateral force friction coefficient calculation model of the embodiment, and the following relationship is satisfied:

[0237]

[0238] in, represents the lateral force friction coefficient, Indicates tire hardness. Indicates tire wear. Indicates the elasticity of the tire.

[0239] The lateral force friction coefficient is obtained through the lateral force friction coefficient calculation model; and the tire performance of the test tire is analyzed and evaluated based on the lateral force friction coefficient. The above lateral force friction coefficient calculation model is constructed based on the analysis results and can be used to describe the mathematical relationship between the lateral force friction coefficient dependent variable and the hardness, wear and elasticity independent variables.

[0240] The constant term 12.766 has a specific statistical significance, representing the expected value of the lateral force friction coefficient when the three independent variables of hardness, wear and elasticity are all zero. The constant term is statistically significant.

[0241] The coefficient of hardness is 0.830, which means that when other independent variables remain unchanged, for every unit increase in hardness, the lateral force friction coefficient will increase by 0.830 units accordingly. The above coefficient is also statistically significant, verifying that hardness has a significant positive effect on the lateral force friction coefficient; the coefficient of wear is -7.690, indicating that when other conditions remain unchanged, for every unit increase in wear, the lateral force friction coefficient will decrease by 7.690 units. The above negative coefficient reveals the negative impact of wear on the lateral force friction coefficient; the coefficient of elasticity is -0.153, which means that when other independent variables are fixed, for every unit increase in elasticity, the lateral force friction coefficient will decrease by 0.153 units. Although the absolute value of the coefficient is relatively small, it is still statistically significant, indicating that elasticity also has a certain influence on the lateral force friction coefficient.

[0242] In addition, the calculation model also shows good statistical characteristics. The coefficients of each independent variable in the model are not only statistically significant, but also the collinearity problem between the independent variables has little effect on the model, which further ensures the stability and reliability of the model. Therefore, the above calculation model not only has a high explanatory power and can clearly reveal the influence mechanism of the independent variables on the dependent variables, but also has good predictive ability. It can be used to predict the possible value of the lateral force friction coefficient under the condition of given independent variable values, which is conducive to understanding and predicting the relationship between the lateral force friction coefficient and hardness, wear and elasticity.

[0243] Existing characterization and analysis methods cover a variety of indicators of the physical and mechanical properties of tires, such as tensile strength, hardness, abrasion, tear strength, elasticity, and the rate of decrease in tensile strength after hot air aging. These performance parameters together constitute a complex system for tire performance evaluation.

[0244] In order to more effectively and quickly analyze and evaluate the performance of the tire, the principal component analysis method is used in the embodiment to screen the key characterization parameters, among which hardness, wear and elasticity are the key characterization parameters of the tire performance. The three key factors are screened out from many performance parameters through principal component analysis and other related statistical means, and the feature information extraction model is used to obtain relevant feature information, thereby avoiding the interference and influence of complex and numerous data, thereby improving the pertinence and accuracy of the tire characterization method.

[0245] Selecting hardness, wear and elasticity as key parameters lays the foundation for accurately grasping the performance of tires in actual use. The above parameters are not only representative, but also can fully reflect the physical and mechanical performance characteristics of tires. Therefore, in the subsequent tire performance analysis and evaluation, relevant key information can be used more efficiently to provide strong support for tire research and development, production and optimization.

[0246] In the embodiment, the least squares method principle is adopted to construct a lateral force friction coefficient calculation model, and based on the hardness, wear and elasticity data of the tire, the lateral force friction coefficient of the tire can be quickly and accurately calculated. The above model has a high degree of accuracy and quantitative ability, and can clearly reveal the influence degree and action relationship of the three key parameters of hardness, wear and elasticity on the lateral force friction coefficient, and can more deeply understand the intrinsic relationship between the physical and mechanical properties of the tire.

[0247] In practical applications, the lateral force friction coefficient calculation model only requires the hardness, wear and elasticity data of the tire to quickly determine the lateral force friction coefficient of the tire, thereby greatly improving the efficiency of tire performance evaluation. It is not only suitable for the research and development stage of tires, helping engineers to quickly determine the key performance indicators of tires and optimize tire design; it is also widely used in the production quality control of tires to ensure that each batch of tires meets the established performance standards; at the same time, in the safety evaluation of tire use, the lateral force friction coefficient calculation model also plays an important role, which helps to promptly discover and resolve safety hazards that may be caused by lateral force friction coefficient problems.

[0248] The lateral force friction coefficient calculation model of this embodiment not only provides an efficient and accurate tire performance evaluation tool, but also provides strong technical support for tire research and development, production, quality control, and safety evaluation, which helps to improve the overall performance of the tire, ensure safety in use, and reduce production costs and safety risks.

[0249] See also Figure 5 In an optional embodiment, in order to efficiently execute the tire characterization method based on the lateral force coefficient provided by the present invention, the present invention also provides a tire characterization system based on the lateral force coefficient, in which the input device, processor, output device and memory are interconnected, wherein the memory is used to store a computer program, the computer program includes program instructions, and the processor is configured to call the program instructions to execute the specific steps of the embodiment of the tire characterization method based on the lateral force coefficient provided by the present invention. The tire characterization system based on the lateral force coefficient of the present invention has a complete structure, objective stability, and can efficiently execute the tire characterization method based on the lateral force coefficient of the present invention, thereby improving the overall applicability and practical application ability of the present invention.

[0250] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein by equivalents. These modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be included in the scope of the claims and specification of the present invention.

Claims

1. A tire characterization method based on lateral force coefficient, characterized in that: The steps include: Selecting a test tire and a test road section, and obtaining tire test data based on the test tire and the test road section; Setting a data quality analysis model, and using the data quality analysis model to perform quality inspection on tire test data to obtain target test data for the test tire; Constructing a tire characterization index analysis mechanism, and obtaining tire characterization parameters based on the tire characterization index analysis mechanism and the target test data; Obtaining characteristic parameter feature information of tire characteristic parameters based on the target test data; Obtaining a lateral force friction coefficient according to the characteristic parameter feature information, and analyzing and evaluating tire performance of the test tire according to the lateral force friction coefficient; The setting of the data quality analysis model comprises: Dividing tire test data into a plurality of sample data sets, and establishing a data quality analysis model according to the plurality of sample data sets; The data quality analysis model satisfies the following relationship: ， in, represents the loss value of tire test data, Indicates the capacity of tire test data, Indicates tire test data The predicted label of the sample, Indicates tire test data The test label of the sample; The characterization parameter characteristic information of the tire characterization parameter obtained based on the target test data includes: Set up the feature information extraction model; The feature information extraction model obtains the characterization parameter feature information of the tire characterization parameter based on the target test data; The feature information extraction model satisfies the following relationship: ， in, Indicates The feature matrix of layer data output, Indicates The feature map matrix of the ResNet path in the layer data, represents the addition operation, It indicates the function mechanism of entering the ResNet path after feature extraction through the Bottlenec structure during the information extraction process. Indicates The feature matrix of layer data output, Represents a merge operation. Indicates The feature map matrix of the DenseNe path in the layer data, It represents the function mechanism of entering the DenseNet path after feature extraction through the Bottleneck structure during the information extraction process.

2. The tire characterization method based on lateral force coefficient according to claim 1, characterized in that: The method of using the data quality analysis model to perform quality inspection on the tire test data to obtain target test data of the test tire includes: Obtaining the loss value of tire test data through the data quality analysis model; Performing quality inspection on tire test data according to the loss value and obtaining a data quality inspection result; Target test data of the test tire is obtained according to the quality inspection result.

3. The tire characterization method based on lateral force coefficient according to claim 1, characterized in that: The constructing of a tire characterization index analysis mechanism and obtaining tire characterization parameters based on the tire characterization index analysis mechanism and the target test data comprises: Construct tire characterization index analysis mechanism based on tire characterization requirements; Selecting a first tire characteristic index according to the tire characteristic index analysis mechanism; Analyzing the basic attributes, correlation, common factor variance and total variance of the first tire characterization index according to the tire characterization index analysis mechanism; The tire characterization parameters are obtained by combining the basic attributes, the correlation, the common factor variance and the total variance.

4. The tire characterization method based on lateral force coefficient according to claim 3, characterized in that: The tire characterization method based on the lateral force coefficient also includes: The component coefficients of the tire characterization parameters are analyzed according to the tire characterization index analysis mechanism.

5. The tire characterization method based on lateral force coefficient according to claim 1, characterized in that: The obtaining of the lateral force friction coefficient according to the characteristic parameter feature information comprises: Analyzing the relationship between the tire characteristic parameter and the lateral force friction coefficient according to the characteristic parameter feature information; Establishing a lateral force friction coefficient calculation model based on the relationship and the tire characterization parameters; The lateral force friction coefficient calculation model satisfies the following relationship: ， in, represents the lateral force friction coefficient, Indicates tire hardness. Indicates tire wear. Indicates the elasticity of the tire.

6. The tire characterization method based on lateral force coefficient according to claim 5, characterized in that: The step of obtaining the lateral force friction coefficient according to the characteristic parameter feature information and analyzing and evaluating the tire performance of the test tire according to the lateral force friction coefficient includes: Obtaining the lateral force friction coefficient by using the lateral force friction coefficient calculation model; The tire performance of the test tire is analyzed and evaluated based on the lateral force friction coefficient.

7. A tire characterization system based on lateral force coefficient, characterized in that, The system includes a processor, an input device, an output device and a memory, wherein the processor, the input device, the output device and the memory are interconnected, wherein the memory is used to store a computer program, the computer program includes program instructions, and the processor is configured to call the program instructions to execute the tire characterization method based on the lateral force coefficient as described in any one of claims 1-6.

Images