A method and system for optimizing the design of automobile seat comfort

Abstract

The present application provides a method and system for optimizing the design of automobile seat comfort, which relates to the field of intelligent manufacturing technology, including: integrating high-frequency user feature sets and high-frequency homologous user feature sets to determine multiple standard user features; configuring a predetermined test plan according to the proportion of road surface features; based on digital twins, taking the seat parameter adjustment threshold as the optimization space, with the purpose of improving seat comfort, according to multiple standard user features and predetermined test plans, using a comfort prediction model to simulate and optimize the seat parameters, output the first optimal seat parameters to perform seat optimization design. This application can solve the technical problems in existing methods that the target user group of automobile seats and the actual use of automobiles cannot be analyzed in a detailed, comprehensive and accurate manner, and the seat comfort cannot be accurately quantitatively evaluated, resulting in the inability of seat parameters to accurately adapt to the needs of the main users, and can achieve the effect of improving product competitiveness and customer satisfaction.

Description

Technical Field

[0001] The present application relates to the field of intelligent manufacturing technology, and in particular to a method and system for optimizing the design of automobile seat comfort. Background Art

[0002] Car seat design is not only related to the driver's comfort, but also to the overall performance and safety of the vehicle. On the one hand, the existing seat design has a low accuracy in user demand analysis, resulting in a low match between seat parameters and user needs, which not only affects the comfort of the seat, but may also lead to long-term health problems such as back pain and neck pain. On the other hand, traditional seat comfort evaluation methods mainly rely on subjective evaluation, such as user feedback and comfort scores, and lack objective quantitative indicators, which makes it difficult to accurately evaluate seat comfort.

[0003] At present, there are technical problems in the existing car seat design methods, such as the inability to conduct a detailed, comprehensive and accurate analysis of the target user groups of car seats and the actual uses of the car, and the inability to accurately quantify the seat comfort. As a result, the seat parameters cannot be accurately adapted to the needs of the main users, resulting in a poor riding experience for users. Summary of the invention

[0004] The purpose of this application is to provide a method and system for optimizing the design of automobile seat comfort, so as to solve the technical problems in existing automobile seat design methods, that is, the inability to conduct a detailed, comprehensive and accurate analysis of the target user group of automobile seats and the actual use of automobiles, and the inability to conduct an accurate quantitative evaluation of seat comfort, resulting in the inability of seat parameters to accurately adapt to the needs of the main users, causing a poor riding experience for users.

[0005] In view of the above problems, the present application provides a method and system for optimizing the design of automobile seat comfort.

[0006] In a first aspect, the present application provides a method for optimizing the comfort design of automobile seats, which is implemented through a system for optimizing the comfort design of automobile seats, including: taking the first seat of a target vehicle as a constraint, performing positive sample retrieval based on the automobile Internet, and obtaining a high-frequency user feature set and a sample driving factor set, wherein the first seat is any one of multiple seats of the target vehicle; performing cluster analysis based on the sample driving factor set to determine key driving factors, performing positive sample retrieval based on the key driving factors, determining a high-frequency homologous user feature set, fusing the high-frequency user feature set and the high-frequency homologous user feature set, and determining multiple standard user features; retrieving the driving road surface characteristics of the target vehicle to obtain a road surface feature set, performing a comprehensive analysis based on the road surface feature set to obtain a road surface feature ratio, and configuring a predetermined test scheme based on the road surface feature ratio, wherein the road surface characteristics include road surface type and distance ratio; based on digital twins, taking the seat parameter adjustment threshold as the optimization space, and taking improving seat comfort as the optimization purpose, according to the multiple standard user features and the predetermined test scheme, using the comfort prediction model to simulate and optimize the seat parameters, outputting the first optimal seat parameters, and performing the seat optimization design of the target vehicle based on the first optimal seat parameters.

[0007] In a second aspect, the present application also provides a car seat comfort optimization design system for executing a car seat comfort optimization design method as described in the first aspect, including: a positive sample retrieval module for performing positive sample retrieval based on the automobile Internet with the first seat of the target vehicle as a constraint, and obtaining a high-frequency user feature set and a sample driving factor set, wherein the first seat is any one of multiple seats of the target vehicle; a feature determination module for performing cluster analysis based on the sample driving factor set, determining key driving factors, performing positive sample retrieval based on the key driving factors, determining a high-frequency homologous user feature set, fusing the high-frequency user feature set and the high-frequency homologous user feature set, and determining Multiple standard user features; a test scheme configuration module, used to retrieve the driving road surface features of the target vehicle, obtain a road surface feature set, perform a comprehensive analysis based on the road surface feature set, obtain the road surface feature ratio, and configure a predetermined test scheme according to the road surface feature ratio, wherein the road surface features include road surface type and distance ratio; a seat parameter optimization module, used to simulate and optimize the seat parameters based on the digital twin, with the seat parameter adjustment threshold as the optimization space, and improving the seat comfort as the optimization purpose, according to the multiple standard user features and the predetermined test scheme, using the comfort prediction model to output the first optimal seat parameters, and perform seat optimization design of the target vehicle according to the first optimal seat parameters.

[0008] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0009] By taking the first seat of the target vehicle as a constraint, positive sample retrieval is performed based on the automotive Internet to obtain a high-frequency user feature set and a sample driving factor set; then cluster analysis is performed based on the sample driving factor set to determine the key driving factors, and positive sample retrieval is performed based on the key driving factors to determine a high-frequency homologous user feature set; the high-frequency user feature set and the high-frequency homologous user feature set are further integrated to determine multiple standard user features, which can improve the accuracy and comprehensiveness of the standard user feature setting; on the other hand, the driving road surface characteristics of the target vehicle are retrieved to obtain a road surface feature set, where the road surface characteristics include road surface type and distance ratio; then a comprehensive analysis is performed based on the road surface feature set to obtain the road surface feature ratio, and a predetermined test plan is configured according to the road surface feature ratio, which can improve the matching degree between the predetermined test plan and the actual use of the car, making the simulation test more realistic and effective. ; Then, based on the digital twin, multiple standard user characteristics are simulated to obtain multiple user vital signs models. On the other hand, the target vehicle and seat parameters are simulated to construct a seat twin model; then, based on multiple user vital signs models, seat twin models and comfort prediction models, the seat parameter adjustment threshold is used as the optimization space, and improving seat comfort is used as the optimization purpose. The seat parameters are simulated and optimized using a predetermined test plan, and the first optimal seat parameters are output according to the optimization results. Finally, the seat optimization design of the target vehicle is performed according to the first optimal seat parameters; the intelligence, accuracy and comprehensiveness of the analysis of the target user group and the actual use of the car can be improved, and the accuracy and scientificity of the quantitative evaluation of seat comfort can be improved, so that car seats that better meet user needs can be designed, achieving the technical effect of improving product competitiveness and customer satisfaction.

[0010] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented according to the contents of the specification, and in order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are specifically cited below. It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present application, nor is it intended to limit the scope of the present application. Other features of the present application will become easy to understand through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] In order to more clearly illustrate the technical solutions in the present application or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only exemplary, and for ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0012] Figure 1 A schematic diagram of a flow chart of a method for optimizing the comfort of a car seat for this application;

[0013] Figure 2 A schematic diagram of a process for obtaining a high-frequency user feature set and a sample driving factor set in a method for optimizing the comfort of a car seat in this application;

[0014] Figure 3 The present invention is a structural schematic diagram of a system for optimizing the comfort of automobile seats according to the present invention.

[0015] Description of reference numerals:

[0016] 11 positive sample retrieval module, 12 feature determination module, 13 test plan configuration module, 14 seat parameter optimization module. DETAILED DESCRIPTION

[0017] This application provides a method and system for optimizing the design of car seat comfort, which solves the technical problem that the existing car seat design methods cannot accurately analyze the target user group of the car seat and the actual use of the car, and cannot accurately quantify the seat comfort, resulting in the inability to accurately adapt the seat parameters to the needs of the main users, causing a poor riding experience for the user. The intelligent, accurate and comprehensive analysis of the target user group and the actual use of the car can be improved, and the accuracy and scientificity of the quantitative evaluation of seat comfort can be improved, so that car seats that better meet user needs can be designed, achieving the technical effect of improving product competitiveness and customer satisfaction.

[0018] Below, the technical solutions in the present application will be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments of the present application. It should be understood that the present application is not limited to the example embodiments described herein. Based on the embodiments of the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present application. It should also be noted that, for the convenience of description, only the parts related to the present application are shown in the accompanying drawings, rather than all of them.

[0019] For example, please refer to the attached Figure 1 The present application provides a method for optimizing the comfort of a car seat, which is applied to a system for optimizing the comfort of a car seat, and specifically includes the following steps:

[0020] Step 1: Taking the first seat of the target vehicle as a constraint, perform positive sample retrieval based on the automotive Internet to obtain a high-frequency user feature set and a sample driving factor set, where the first seat is any one of the multiple seats of the target vehicle.

[0021] Specifically, first, determine the target vehicle, which refers to the vehicle whose car seat comfort is to be optimized, and can be selected according to actual conditions. The target vehicle includes multiple seats, such as a driver's seat, a passenger seat, and rear seats. Then, randomly select any one of the multiple seats in the target vehicle and set it as the first seat. For example, the first seat is the driver's seat. Since each seat is aimed at a different user group, for example, in an ordinary family car, the driver's seat is usually for the male owner of the family, and the passenger seat is usually for the female owner of the family, etc., separate analysis of the seats can improve the refinement of the seat optimization design.

[0022] Taking the vehicle model of the target vehicle and the first seat of the target vehicle as the retrieval constraints, a positive sample retrieval is performed based on the automotive Internet. Positive sample retrieval is an information retrieval technology used to retrieve samples that meet specific conditions from a large amount of data to obtain a sample user feature set and a sample driving factor set, where user features refer to the physical features of the user who purchases the target vehicle and is usually the first seat user, including gender, age, height, weight, etc.; driving factors refer to the driving factors for users to purchase the vehicle, such as price indicators, vehicle size and functional configuration, etc., where driving factors include factor order. The higher the factor order, the higher the user's intention to the driving factor. Then, a high-frequency feature analysis is performed on the sample user feature set, that is, the most common or most important features are extracted from the sample user feature set. High-frequency features often reflect the common needs and preferences of most users, and a high-frequency user feature set is obtained. By screening high-frequency features and obtaining a high-frequency user feature set, the amount of data processing can be reduced and the accuracy and efficiency of user feature analysis can be improved.

[0023] Step 2: Perform cluster analysis based on the sample driving factor set to determine the key driving factors, perform positive sample retrieval based on the key driving factors, determine the high-frequency homologous user feature set, fuse the high-frequency user feature set and the high-frequency homologous user feature set, and determine multiple standard user features.

[0024] Specifically, a cluster analysis is then performed on multiple sample driving factors in the sample driving factor set, that is, the importance of the same sample driving factors is analyzed, and sample driving factors with higher importance are selected as key driving factors to determine the key driving factors; further, with the key driving factors as constraints, positive sample retrieval is performed based on the automotive Internet to obtain a sample homologous user feature set, where homologous users refer to the user group of the first seat who has similar characteristics to the purchase of the target vehicle; by retrieving homologous user features, the depth and comprehensiveness of the target user feature analysis can be improved, thereby more accurately obtaining the main user needs.

[0025] Then, a high-frequency feature analysis is performed on the sample homologous user feature set, wherein the high-frequency feature analysis method is the same as the analysis method of the sample user feature set mentioned above, which is not elaborated in this application. The high-frequency feature analysis method of the sample user feature set can be referred to to obtain a high-frequency homologous user feature set. The high-frequency user feature set and the high-frequency homologous user feature set are further fused to obtain an overall user feature set. For example, the weight ratio can be set according to the actual situation for weighted fusion, wherein the user feature weight ratio is greater than the homologous user feature weight ratio; then, representative feature extraction is performed on the overall user feature set to obtain multiple standard user features, wherein the standard user feature is a key feature that can represent the main user feature. By obtaining multiple standard user features, support is provided for the subsequent seat demand analysis of the main user.

[0026] Step 3: Retrieve the driving road characteristics of the target vehicle, obtain a road feature set, perform a comprehensive analysis based on the road feature set, obtain the road feature proportion, and configure a predetermined test plan based on the road feature proportion, wherein the road feature includes road type and distance proportion.

[0027] Specifically, with the target vehicle model as a constraint, a positive sample retrieval of road surface features is performed based on the industrial Internet, where road surface features refer to vehicle driving road surface features, including road surface type and distance ratio. Road surface types include highways, urban roads, dirt roads, off-road conditions, etc., and distance ratio refers to the driving distance ratio corresponding to the road surface type, to obtain a road surface feature set. Further, a comprehensive analysis is performed on multiple road surface features in the road surface feature set, that is, the distance ratio of the same road surface type is averaged to obtain multiple distance ratio averages of multiple road surface types, and the averages of multiple road surface types and multiple distance ratios are used as road surface feature ratios.

[0028] Finally, a predetermined test plan is configured according to the proportion of the road surface characteristics, that is, the simulated test road conditions are set according to the road surface type, and the number of simulated tests is set according to the average distance proportion, wherein the number of simulated tests is proportional to the average distance proportion, that is, the larger the average distance proportion, the larger the corresponding number of simulated tests. For example, if the overall number of tests in the test plan is 100 times and the proportion of urban roads is 60%, then the number of simulated tests on urban roads is 60 times. By configuring a predetermined test plan according to the proportion of road surface characteristics, a targeted test plan can be set in combination with the actual use of the car, thereby improving the accuracy and rationality of the predetermined test plan setting, and thereby improving the accuracy and scientificity of the seat comfort simulation test.

[0029] Step 4: Based on the digital twin, with the seat parameter adjustment threshold as the optimization space and improving seat comfort as the optimization goal, the seat parameters are simulated and optimized using the comfort prediction model according to the multiple standard user characteristics and the predetermined test plan, and the first optimal seat parameters are output. The seat optimization design of the target vehicle is performed based on the first optimal seat parameters.

[0030] Specifically, digital twin is a digital technology that creates a virtual model of a physical entity to achieve real-time monitoring, predictive analysis and optimal decision-making of the physical entity. It has many advantages such as interoperability and fidelity, and can improve the authenticity, accuracy and reliability of simulation. First, based on digital twin, multiple standard user characteristics are simulated to obtain multiple user sign models. On the other hand, the target vehicle and seat parameters are simulated to build a seat twin model.

[0031] Then, the seat parameters are obtained, where the seat parameters refer to the adjustable features in the seat design process, including material type, filling type, cushion length, backrest height, lumbar support height, headrest shape, headrest material and other features, and the seat parameter adjustment threshold is obtained; on the other hand, a comfort prediction model is constructed based on the BP neural network. The comfort prediction model is a feedforward neural network model that can be iteratively optimized in machine learning, and is used to predict seat comfort based on user test data in the simulation test process, where the input data of the comfort prediction model is pressure distribution, vibration influence and noise influence, and the output data is the seat comfort coefficient.

[0032] Based on multiple user vital sign models and seat twin models, simulation tests are carried out according to a predetermined test plan, and a simulation test data set is output. The simulation test data set is predicted by the comfort prediction model to output the seat comfort coefficient. Then, the seat parameter adjustment threshold is used as the optimization space, and the maximization of the comfort coefficient of the seat parameters is used as the optimization purpose. The seat parameters are simulated and optimized using the optimization algorithm, and the seat parameters corresponding to the maximum comfort coefficient in the optimization process are set as the first optimal seat parameters.

[0033] Using the same method as that for obtaining the first optimal seat parameters, the other seats of the target vehicle are optimized to obtain multiple optimal seat parameters; finally, the seat optimization design of the target vehicle is performed based on the multiple optimal seat parameters, thereby designing a car seat that better meets the needs of the main users.

[0034] The method for optimizing the comfort of automobile seats is applied to an optimization design system for the comfort of automobile seats, which can solve the technical problems in the existing automobile seat design methods that the target user group of automobile seats and the actual use of automobiles cannot be analyzed in a detailed, comprehensive and accurate manner, and the seat comfort cannot be accurately and quantitatively evaluated, resulting in the inability of seat parameters to accurately adapt to the needs of the main users, causing a poor user riding experience. By taking the first seat of the target vehicle as a constraint, a positive sample search is performed based on the automobile Internet to obtain a high-frequency user feature set and a sample driving factor set; then a cluster analysis is performed based on the sample driving factor set to determine the key driving factors, and a positive sample search is performed based on the key driving factors to determine a high-frequency homologous user feature set; the high-frequency user feature set and the high-frequency homologous user feature set are further integrated to determine multiple standard user features, which can improve the accuracy and comprehensiveness of the standard user feature setting; on the other hand, the driving road surface characteristics of the target vehicle are retrieved to obtain a road surface feature set, wherein the road surface characteristics include road surface type and distance ratio, and then a comprehensive analysis is performed based on the road surface feature set to obtain the road surface feature ratio, and a predetermined test plan is configured according to the road surface feature ratio, which can improve the matching degree between the predetermined test plan and the actual use of the automobile, making the simulation test more realistic and effective. ; Then, based on the digital twin, multiple standard user characteristics are simulated to obtain multiple user vital signs models. On the other hand, the target vehicle and seat parameters are simulated to construct a seat twin model; then, based on multiple user vital signs models, seat twin models and comfort prediction models, the seat parameter adjustment threshold is used as the optimization space, and improving seat comfort is used as the optimization purpose. The seat parameters are simulated and optimized using a predetermined test plan, and the first optimal seat parameters are output according to the optimization results. Finally, the seat optimization design of the target vehicle is performed according to the first optimal seat parameters; the intelligence, accuracy and comprehensiveness of the analysis of the target user group and the actual use of the car can be improved, and the accuracy and scientificity of the quantitative evaluation of seat comfort can be improved, so that car seats that better meet user needs can be designed, achieving the technical effect of improving product competitiveness and customer satisfaction.

[0035] Further, obtain the high-frequency user feature set and sample driving factor set, as shown in the attached Figure 2 As shown, step one of this application includes:

[0036] Taking the first seat of the target vehicle as a constraint, a positive sample search is performed based on the automotive Internet to obtain a sample user feature set and a sample driving factor set, wherein the user feature includes at least height, weight and age, the driving factor includes at least price index, vehicle size and functional configuration, and the driving factor is identified with a factor order; a first sample user feature is randomly selected from the sample user feature set, wherein the first sample user feature includes a first sample height, a first sample weight and a first sample age; a deviation is calculated between the first sample height and other sample heights in the sample user feature set, and the number of height deviations less than a predetermined height deviation threshold is counted and set as the first key degree; if the first criticality is greater than a predetermined threshold, a deviation calculation is performed on the first sample weight and other sample weights in the sample user feature set, and the number of weight deviations less than a predetermined weight deviation threshold is counted, which is set as the second criticality; if the second criticality is greater than a predetermined threshold, a deviation calculation is performed on the first sample age and other sample ages in the sample user feature set, and the number of age deviations less than a predetermined age deviation threshold is counted, which is set as the third criticality; if the third criticality is greater than a predetermined threshold, the first sample user feature is added to the high-frequency user feature set until the sample user features are traversed and the high-frequency user feature set is obtained.

[0037] Specifically, firstly, with the target vehicle model and the first seat of the target vehicle as constraints, positive sample retrieval is performed based on the automotive Internet to obtain a sample user feature set and a sample driving factor set, wherein user features include at least height, weight and age, and technical personnel in this field may also add additional features according to actual needs, such as gender, waist circumference, hip circumference, etc.; driving factors include at least price indicators, vehicle size and functional configuration, and indicators such as power performance may also be added, wherein the driving factors are marked with a factor order, and the factor order represents the degree of importance the user places on the driving factors, and the higher the factor order, the higher the degree of importance.

[0038] Then, a sample user feature is randomly selected from the sample user feature set and set as the first sample user feature, wherein the first sample user feature includes a first sample height, a first sample weight and a first sample age; then, the deviation between the first sample height and other sample heights in the sample user feature set is calculated to obtain multiple height deviations, and the number of height deviations less than a predetermined height deviation threshold is counted and set as a first criticality, wherein the first criticality represents the criticality of the first sample height; if the first criticality is greater than a predetermined threshold, and the predetermined threshold can be set according to actual conditions, then the deviation between the first sample weight and other sample weights in the sample user feature set is calculated, and the number of weight deviations less than a predetermined weight deviation threshold is counted and set as a second criticality; if the second criticality is greater than a predetermined threshold, then the deviation between the first sample age and other sample ages in the sample user feature set is calculated, and the number of age deviations less than a predetermined age deviation threshold is counted and set as a third criticality; if the third criticality is greater than a predetermined threshold, then the first sample user feature is added to the high-frequency user feature set until the sample user feature traversal is completed to obtain the high-frequency user feature set.

[0039] If the first criticality is less than or equal to a predetermined threshold, or the second criticality is less than or equal to a predetermined threshold, or the third criticality is less than or equal to a predetermined threshold, the first sample user feature is discarded.

[0040] By screening the sample user feature set through the above steps to obtain a high-frequency user feature set, it is possible to reduce the interference of low-frequency features while retaining the main features of the user, thereby improving the accuracy and efficiency of user feature analysis.

[0041] Further, to determine the key driving factors, step 2 of this application includes:

[0042] Based on the coefficient of variation method, the weight ratio is configured according to the factor order, wherein the factor order and the weight ratio are negatively correlated; the sample driving factor set is weightedly calculated according to the weight ratio to obtain multiple correlation frequencies of multiple sample driving factors, and the sample driving factor corresponding to the correlation frequency greater than the predetermined frequency threshold is selected and set as the key driving factor.

[0043] Specifically, the coefficient of variation method is a method that calculates the degree of change of each indicator of the system based on statistical methods, and assigns weights to each indicator according to the degree of variation between the current value and the target value of each evaluation indicator. If the numerical difference of an indicator is large, it can clearly distinguish the evaluated objects, indicating that the indicator has rich distinguishing information, and thus the indicator should be given a larger weight; conversely, if the numerical difference of each evaluated object on a certain indicator is small, then the ability of this indicator to distinguish the evaluation objects is weak, and thus the indicator should be given a smaller weight.

[0044] Based on the coefficient of variation method, the weight proportion is configured according to the order of factors, where the order of factors and the weight proportion are negatively correlated, that is, the smaller the order of factors, the larger the weight proportion. For example, the first driving factor represents the first intention factor of users to purchase the target vehicle, and the weight proportion is the largest; first, use appropriate statistical methods (such as principal component analysis) to calculate the scores of each factor, which reflects the importance of the factor; then for each factor, calculate its coefficient of variation, which is the ratio of the standard deviation to the mean, reflecting the degree of dispersion of the data; further, according to the size of the coefficient of variation, determine the order of factors, and the factors with larger coefficients of variation are more advanced in the order; since the order of factors and the weight proportion are negatively correlated, that is, the factors with higher order have larger weight proportions, finally set a reasonable proportional coefficient to convert the order of factors into weight proportions. By configuring weight proportions based on the coefficient of variation method, the accuracy and scientificity of weight setting can be improved.

[0045] Then, the sample driving factor set is weightedly calculated according to the weight ratio to obtain multiple correlation frequencies of multiple sample driving factors; finally, the sample driving factor corresponding to the correlation frequency greater than the predetermined frequency threshold is selected and set as the key driving factor to obtain the key driving factor.

[0046] Further, a plurality of standard user characteristics are determined, and step 2 of this application includes:

[0047] The high-frequency user feature set and the high-frequency homologous user feature set are integrated to obtain a user feature set; the user feature set is clustered according to a predetermined step length to determine a plurality of user feature intervals, wherein the user feature intervals are marked with a degree of aggregation; user feature intervals with a degree of aggregation less than a predetermined aggregation threshold are eliminated to obtain a plurality of standard user feature intervals, and intermediate value operations are performed on the plurality of standard user feature intervals to obtain the plurality of standard user features.

[0048] Specifically, first, the high-frequency user feature set and the high-frequency homologous user feature set are integrated. For example, weights can be set according to actual conditions for weighted fusion, where the user feature weight is greater than the homologous user feature weight, for example, the user feature weight is 0.7, and the homologous user feature weight is 0.3, to obtain a user feature set. Then, the user feature set is clustered according to a predetermined step length, where the predetermined step length includes a height step length, a weight step length, and an age step length. For example, the height step length can be set to 1 cm, the weight step length can be set to 1 kg, and the age step length can be set to 1 year old, that is, the user features with a height deviation within 1 cm, a weight deviation within 1 kg, and an age deviation within 1 year old are clustered into one category to obtain multiple user feature intervals; then, the number of features in each user feature interval is counted and set as the interval aggregation degree, where the more features there are, the greater the aggregation degree is, and the higher the centrality of the user feature is characterized. Then, based on a predetermined aggregation threshold, multiple user feature intervals are screened, and user feature intervals with aggregation degrees less than the predetermined aggregation threshold are eliminated to obtain multiple standard user feature intervals; further, the middle values ​​of the multiple standard user feature intervals are selected as standard user features to obtain multiple standard user features.

[0049] Further, the first optimal seat parameters are outputted, and step 4 of the present application includes:

[0050] Acquire seat parameter adjustment thresholds, wherein the seat parameters include at least a material type threshold, a filler type threshold, a cushion length threshold, a backrest height threshold, and a lumbar support height threshold; randomly select a parameter from the material type threshold, the filler type threshold, the cushion length threshold, the backrest height threshold, and the lumbar support height threshold for combination to generate a plurality of initial seat parameters; based on digital twins, simulate the plurality of standard user features to obtain a plurality of user vital signs models, and simulate according to the target vehicle attribute features and the plurality of initial seat parameters to obtain a plurality of seat twin models; with the purpose of optimizing to improve seat comfort, based on the plurality of user vital signs models and the plurality of seat twin models, according to the predetermined test scheme, use a comfort prediction model to simulate and optimize the seat parameters, and output the first optimal seat parameters.

[0051] Specifically, a seat parameter adjustment threshold is obtained, wherein the seat parameters include at least a material type threshold, a filling type threshold, a cushion length threshold, a backrest height threshold and a lumbar support height threshold; then a parameter is randomly selected from the material type threshold, the filling type threshold, the cushion length threshold, the backrest height threshold and the lumbar support height threshold for combination to obtain a first initial seat parameter, and the same method is used for iterative selection to obtain multiple initial seat parameters.

[0052] Then, based on the digital twin technology, the multiple standard user characteristics are simulated in a three-dimensional simulation platform to obtain multiple user physical sign models. These models have different heights, weights, body shapes and other characteristics, which can simulate the physical characteristics and behaviors of different users when using seats. On the other hand, in the three-dimensional simulation platform, simulation is performed according to the target vehicle attribute characteristics and the multiple initial seat parameters, that is, the position structure of the seat in the target vehicle is simulated to obtain multiple seat twin models. By obtaining the seat twin model, the layout of the seat in the vehicle can be simulated, the support and comfort of the seat can be evaluated, and the vibration and impact of the seat during vehicle driving can also be simulated.

[0053] Finally, with the purpose of optimizing the seat comfort, based on the multiple user physical sign models and the multiple seat twin models, and according to the predetermined test plan, the seat parameters are simulated and optimized using the comfort prediction model to output the first optimal seat parameters.

[0054] Further, to construct a comfort prediction model, the present application also includes the following steps:

[0055] Retrieve and obtain a sample comfort data set, wherein the sample comfort data includes sample pressure distribution, sample vibration influence, sample noise influence and sample comfort coefficient, and the sample pressure distribution includes sample back average pressure, sample back peak pressure, sample back average pressure, sample back peak pressure, sample back contact area, sample buttocks average pressure, sample buttocks peak pressure, sample buttocks average pressure, sample buttocks peak pressure, and sample buttocks contact area; divide the sample comfort data set into Q equal parts, and select Q times with replacement to construct a first training set, iterate and select Q times to generate Q data sets; use the Q data sets to perform supervised training on the BP neural network to obtain Q convergence prediction units, and construct the comfort prediction model based on the Q convergence prediction units, wherein the output of the comfort prediction model is the mean of the output results of the Q convergence prediction units.

[0056] Specifically, firstly, a sample comfort data set is obtained based on automobile Internet search, wherein the sample comfort data includes sample pressure distribution, sample vibration influence, sample noise influence and sample comfort coefficient. The sample comfort coefficient represents the specific comfort of the seat. The larger the comfort coefficient, the higher the seat comfort. It can be comprehensively set through historical user's subjective rating and system evaluation. The sample pressure distribution includes sample back average pressure, sample back peak pressure, sample back average pressure, sample back peak pressure, sample back contact area, sample buttocks average pressure, sample buttocks peak pressure, sample buttocks average pressure, sample buttocks peak pressure, and sample buttocks contact area. Then the sample comfort data set is divided into Q parts, and selected Q times with replacement to construct the first training set. The same method is used to iteratively select Q times to generate Q data sets.

[0057] Then, taking the sample pressure distribution, sample vibration influence, and sample noise influence as input and the sample comfort coefficient as supervision, the BP neural network is supervised and trained using Q data sets. It can be trained through the back propagation algorithm and the loss function to obtain Q convergence prediction units that meet the expected conditions; then, the comfort prediction model is constructed based on the Q convergence prediction units, wherein the output of the comfort prediction model is the mean of the output results of the Q convergence prediction units.

[0058] By constructing a comfort prediction model based on the BP neural network, accurate quantitative evaluation of seat comfort can be achieved, which provides a basis for the subsequent simulation and optimization of seat parameters. In addition, since the comfort prediction model is constructed in parallel based on Q convergent prediction units, the prediction error of a single model can be reduced, thereby improving the accuracy and reliability of seat comfort prediction.

[0059] Further, to output the first optimal seat parameters, the present application further comprises the following steps:

[0060] A first seat twin model is randomly selected from the multiple seat twin models; through the first seat twin model, according to the predetermined test plan, the simulation test of the multiple user vital sign models is performed to obtain multiple simulation test data sets; the multiple simulation test data sets are input into the comfort prediction model, multiple predicted comfort sets are output, and the mean is calculated to obtain a first comfort; multiple comforts of multiple seat twin models are obtained in turn, and the seat parameters are simulated and optimized based on the multiple comforts for the purpose of optimizing the seat comfort, so as to obtain the first optimal seat parameters.

[0061] Specifically, first, randomly select any one of the multiple seat twin models and set it as the first seat twin model; then, through the first seat twin model, according to the predetermined test plan, perform simulation tests on the multiple user physical sign models to obtain multiple simulation test data sets; further input the multiple simulation test data sets into the comfort prediction model, output multiple predicted comfort sets, that is, multiple comfort coefficient sets, and then fuse the multiple comfort coefficient sets to obtain the overall comfort coefficient set and perform mean calculation, and set the mean calculation result as the first comfort. Use the same method to obtain multiple comforts of multiple seat twin models in turn, and then, with the purpose of improving seat comfort, simulate and optimize the seat parameters based on the multiple comforts to obtain the first optimal seat parameters.

[0062] Furthermore, based on the multiple comfort levels, seat parameters are simulated and optimized, and the present application further includes the following steps:

[0063] The initial seat parameters are regarded as the initial solutions, and multiple initial solutions are arranged from large to small according to the comfort level to obtain an initial solution sequence; the first P initial solutions of the initial solution sequence are selected as optimal solutions, and the last K initial solutions are set as inferior solutions, wherein the sum of P and K is the number of initial solutions, and K is an integer multiple of P; with the P optimal solutions as the center, the K inferior solutions are randomly clustered equally to obtain P domains, wherein the number of inferior solutions in each domain is the same; within the P domains, the optimal solution is used as the optimization direction, and the inferior solutions in the same domain are mutated once according to the predetermined variation step length to obtain P update domains, wherein if the mutated inferior solution does not meet the seat parameter adjustment threshold, this mutation is not performed; P update domains are identified, and if the comfort level of the inferior solution is greater than the comfort level of the optimal solution in the same update domain, the superior solution is replaced by the inferior solution; iterative optimization is performed until the predetermined number of optimization searches is met, P current domains are output, and the current domain with the largest sum of comfort levels is set as the optimal domain, and the superior solution of the optimal domain is selected and set as the first optimal seat parameter.

[0064] Specifically, first, the initial seat parameters are regarded as the initial solutions, and multiple initial solutions are arranged in descending order of comfort to obtain an initial solution sequence; then the first P initial solutions of the initial solution sequence are selected as optimal solutions, and the last K initial solutions are set as inferior solutions, where the sum of P and K is the number of initial solutions, and K is much larger than P, and K is an integer multiple of P; further, with the P optimal solutions as the center, the K inferior solutions are randomly clustered equally to obtain P domains, where the number of inferior solutions in each domain is the same; then, within the P domains, the optimal solutions are used as the center. In the optimization direction, the inferior solution in the same domain is mutated once according to the predetermined variation step length, wherein the predetermined variation step length includes the adjustment step lengths of multiple seat parameters, such as the cushion length step length, the backrest height step length, and the lumbar support height step length, etc.; in a variation process, if the mutated inferior solution does not meet the seat parameter adjustment threshold, the inferior solution is not mutated this time, and P update domains are obtained; further identify the P update domains, if the comfort of the inferior solution is greater than the comfort of the optimal solution in the same update domain, the optimal solution is replaced by the inferior solution. Iterate the optimization until the predetermined number of optimizations is met. The predetermined number of optimizations can be set according to the optimization requirements, output P current domains, and set the current domain with the largest sum of comfort as the optimal domain, and select the optimal solution in the optimal domain as the first optimal seat parameter.

[0065] By using the above algorithm to optimize the seat parameters, due to the strong global convergence ability of the algorithm, it is possible to avoid falling into the local optimum, thereby improving the accuracy and rationality of the optimal seat parameters.

[0066] In summary, the vehicle seat comfort optimization design method provided by this application has the following technical effects:

[0067] 1. Taking the first seat of the target vehicle as a constraint, positive sample retrieval is performed based on the automotive Internet to obtain a high-frequency user feature set and a sample driving factor set; then cluster analysis is performed based on the sample driving factor set to determine the key driving factors, and positive sample retrieval is performed based on the key driving factors to determine a high-frequency homologous user feature set; the high-frequency user feature set and the high-frequency homologous user feature set are further integrated to determine multiple standard user features, which can improve the accuracy and comprehensiveness of the standard user feature setting; on the other hand, the driving road surface characteristics of the target vehicle are retrieved to obtain a road surface feature set, where the road surface characteristics include road surface type and distance ratio, and then a comprehensive analysis is performed based on the road surface feature set to obtain the road surface feature ratio, and a predetermined test plan is configured according to the road surface feature ratio, which can improve the matching degree between the predetermined test plan and the actual use of the car, making the simulation test more realistic and effective. effect; then, based on the digital twin, multiple standard user characteristics are simulated to obtain multiple user vital signs models. On the other hand, the target vehicle and seat parameters are simulated to construct a seat twin model; then, based on multiple user vital signs models, seat twin models and comfort prediction models, the seat parameter adjustment threshold is used as the optimization space, and the improvement of seat comfort is used as the optimization purpose. The seat parameters are simulated and optimized using a predetermined test plan, and the first optimal seat parameters are output according to the optimization results. Finally, the seat optimization design of the target vehicle is performed according to the first optimal seat parameters; it can improve the intelligence, accuracy and comprehensiveness of the analysis of the target user group and the actual use of the car, and at the same time improve the accuracy and scientificity of the quantitative evaluation of seat comfort, so that car seats that better meet user needs can be designed, achieving the technical effect of improving product competitiveness and customer satisfaction.

[0068] 2. By constructing a comfort prediction model based on the BP neural network, accurate quantitative evaluation of seat comfort can be achieved, which provides a basis for the subsequent simulation and optimization of seat parameters. In addition, since the comfort prediction model is constructed in parallel based on Q convergent prediction units, the prediction error of a single model can be reduced, thereby improving the accuracy and reliability of seat comfort prediction.

[0069] Embodiment 2: Based on the same invention concept as the method for optimizing the comfort of a car seat in the above embodiment, the present application also provides a system for optimizing the comfort of a car seat. Figure 3 ,include:

[0070] A positive sample retrieval module 11 is used to perform positive sample retrieval based on the automobile Internet with the first seat of the target vehicle as a constraint, and obtain a high-frequency user feature set and a sample driving factor set, wherein the first seat is any one of the multiple seats of the target vehicle; a feature determination module 12 is used to perform cluster analysis based on the sample driving factor set, determine the key driving factor, perform positive sample retrieval based on the key driving factor, determine a high-frequency homologous user feature set, fuse the high-frequency user feature set and the high-frequency homologous user feature set, and determine multiple standard user features; a test scheme configuration module 13 is used to retrieve the driving road surface characteristics of the target vehicle, obtain a road surface feature set, perform a comprehensive analysis based on the road surface feature set, obtain a road surface feature ratio, and configure a predetermined test scheme based on the road surface feature ratio, wherein the road surface characteristics include road surface type and distance ratio; a seat parameter optimization module 14 is used to simulate and optimize the seat parameters based on the digital twin, with the seat parameter adjustment threshold as the optimization space, and with improving seat comfort as the optimization purpose, according to the multiple standard user features and the predetermined test scheme, using the comfort prediction model to output the first optimal seat parameters, and perform seat optimization design of the target vehicle according to the first optimal seat parameters.

[0071] Furthermore, the automobile seat comfort optimization design system is also used for:

[0072] Taking the first seat of the target vehicle as a constraint, a positive sample search is performed based on the automotive Internet to obtain a sample user feature set and a sample driving factor set, wherein the user feature includes at least height, weight and age, the driving factor includes at least price index, vehicle size and functional configuration, and the driving factor is identified with a factor order; a first sample user feature is randomly selected from the sample user feature set, wherein the first sample user feature includes a first sample height, a first sample weight and a first sample age; a deviation is calculated between the first sample height and other sample heights in the sample user feature set, and the number of height deviations less than a predetermined height deviation threshold is counted and set as the first key degree; if the first criticality is greater than a predetermined threshold, a deviation calculation is performed on the first sample weight and other sample weights in the sample user feature set, and the number of weight deviations less than a predetermined weight deviation threshold is counted, which is set as the second criticality; if the second criticality is greater than a predetermined threshold, a deviation calculation is performed on the first sample age and other sample ages in the sample user feature set, and the number of age deviations less than a predetermined age deviation threshold is counted, which is set as the third criticality; if the third criticality is greater than a predetermined threshold, the first sample user feature is added to the high-frequency user feature set until the sample user features are traversed and the high-frequency user feature set is obtained.

[0073] Furthermore, the automobile seat comfort optimization design system is also used for:

[0074] Based on the coefficient of variation method, the weight ratio is configured according to the factor order, wherein the factor order and the weight ratio are negatively correlated; the sample driving factor set is weightedly calculated according to the weight ratio to obtain multiple correlation frequencies of multiple sample driving factors, and the sample driving factor corresponding to the correlation frequency greater than the predetermined frequency threshold is selected and set as the key driving factor.

[0075] Furthermore, the automobile seat comfort optimization design system is also used for:

[0076] The high-frequency user feature set and the high-frequency homologous user feature set are integrated to obtain a user feature set; the user feature set is clustered according to a predetermined step length to determine a plurality of user feature intervals, wherein the user feature intervals are marked with a degree of aggregation; user feature intervals with a degree of aggregation less than a predetermined aggregation threshold are eliminated to obtain a plurality of standard user feature intervals, and intermediate value operations are performed on the plurality of standard user feature intervals to obtain the plurality of standard user features.

[0077] Furthermore, the automobile seat comfort optimization design system is also used for:

[0078] Acquire seat parameter adjustment thresholds, wherein the seat parameters include at least a material type threshold, a filler type threshold, a cushion length threshold, a backrest height threshold, and a lumbar support height threshold; randomly select a parameter from the material type threshold, the filler type threshold, the cushion length threshold, the backrest height threshold, and the lumbar support height threshold for combination to generate a plurality of initial seat parameters; based on digital twins, simulate the plurality of standard user features to obtain a plurality of user vital signs models, and simulate according to the target vehicle attribute features and the plurality of initial seat parameters to obtain a plurality of seat twin models; with the purpose of optimizing to improve seat comfort, based on the plurality of user vital signs models and the plurality of seat twin models, according to the predetermined test scheme, use a comfort prediction model to simulate and optimize the seat parameters, and output the first optimal seat parameters.

[0079] Furthermore, the automobile seat comfort optimization design system is also used for:

[0080] Retrieve and obtain a sample comfort data set, wherein the sample comfort data includes sample pressure distribution, sample vibration influence, sample noise influence and sample comfort coefficient, and the sample pressure distribution includes sample back average pressure, sample back peak pressure, sample back average pressure, sample back peak pressure, sample back contact area, sample buttocks average pressure, sample buttocks peak pressure, sample buttocks average pressure, sample buttocks peak pressure, and sample buttocks contact area; divide the sample comfort data set into Q equal parts, and select Q times with replacement to construct a first training set, iterate and select Q times to generate Q data sets; use the Q data sets to perform supervised training on the BP neural network to obtain Q convergence prediction units, and construct the comfort prediction model based on the Q convergence prediction units, wherein the output of the comfort prediction model is the mean of the output results of the Q convergence prediction units.

[0081] Furthermore, the automobile seat comfort optimization design system is also used for:

[0082] A first seat twin model is randomly selected from the multiple seat twin models; through the first seat twin model, according to the predetermined test plan, the simulation test of the multiple user vital sign models is performed to obtain multiple simulation test data sets; the multiple simulation test data sets are input into the comfort prediction model, multiple predicted comfort sets are output, and the mean is calculated to obtain a first comfort; multiple comforts of multiple seat twin models are obtained in turn, and the seat parameters are simulated and optimized based on the multiple comforts for the purpose of optimizing the seat comfort, so as to obtain the first optimal seat parameters.

[0083] Furthermore, the automobile seat comfort optimization design system is also used for:

[0084] The initial seat parameters are regarded as the initial solutions, and multiple initial solutions are arranged from large to small according to the comfort level to obtain an initial solution sequence; the first P initial solutions of the initial solution sequence are selected as optimal solutions, and the last K initial solutions are set as inferior solutions, wherein the sum of P and K is the number of initial solutions, and K is an integer multiple of P; with the P optimal solutions as the center, the K inferior solutions are randomly clustered equally to obtain P domains, wherein the number of inferior solutions in each domain is the same; within the P domains, the optimal solution is used as the optimization direction, and the inferior solutions in the same domain are mutated once according to the predetermined variation step length to obtain P update domains, wherein if the mutated inferior solution does not meet the seat parameter adjustment threshold, this mutation is not performed; P update domains are identified, and if the comfort level of the inferior solution is greater than the comfort level of the optimal solution in the same update domain, the superior solution is replaced by the inferior solution; iterative optimization is performed until the predetermined number of optimization searches is met, P current domains are output, and the current domain with the largest sum of comfort levels is set as the optimal domain, and the superior solution of the optimal domain is selected and set as the first optimal seat parameter.

[0085] Each embodiment in this specification is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The above-mentioned car seat comfort optimization design method and specific examples in the first embodiment are also applicable to the car seat comfort optimization design system in this embodiment. Through the above-mentioned detailed description of the car seat comfort optimization design method, those skilled in the art can clearly know the car seat comfort optimization design system in this embodiment, so for the sake of brevity of the specification, it will not be described in detail here. For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part description.

[0086] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

[0087] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application belong to the scope of the present application and its equivalent technology, the present application is also intended to include these modifications and variations.

Claims

1. A method for optimizing the comfort of a car seat, characterized in that: include: Taking the first seat of the target vehicle as a constraint, positive sample retrieval is performed based on the Internet of Vehicles to obtain a high-frequency user feature set and a sample driving factor set, where the first seat is any one of the multiple seats of the target vehicle; Performing cluster analysis based on the sample driving factor set to determine key driving factors, performing positive sample retrieval based on the key driving factors to determine a high-frequency homologous user feature set, fusing the high-frequency user feature set and the high-frequency homologous user feature set to determine a plurality of standard user features; Retrieving the driving road characteristics of the target vehicle, obtaining a road characteristic set, performing a comprehensive analysis based on the road characteristic set to obtain a road characteristic ratio, and configuring a predetermined test plan according to the road characteristic ratio, wherein the road characteristic includes a road type and a distance ratio; Based on the digital twin, taking the seat parameter adjustment threshold as the optimization space and improving the seat comfort as the optimization purpose, according to the multiple standard user characteristics and the predetermined test plan, the seat parameters are simulated and optimized using the comfort prediction model, and the first optimal seat parameters are output, and the seat optimization design of the target vehicle is performed according to the first optimal seat parameters; Obtain high-frequency user feature sets and sample driving factor sets, including: Taking the first seat of the target vehicle as a constraint, positive sample retrieval is performed based on the Internet of Vehicles to obtain a sample user feature set and a sample driving factor set, wherein the user features include at least height, weight and age, and the driving factors include at least price index, vehicle size and functional configuration, and the driving factors are marked with factor order; Randomly selecting a first sample user feature from the sample user feature set, wherein the first sample user feature includes a first sample height, a first sample weight, and a first sample age; Calculate the deviation between the first sample height and the heights of other samples in the sample user feature set, and count the number of height deviations less than a predetermined height deviation threshold, and set it as a first criticality; If the first criticality is greater than a predetermined threshold, a deviation calculation is performed on the first sample weight and other sample weights in the sample user feature set, and the number of weight deviations less than a predetermined weight deviation threshold is counted and set as a second criticality; If the second criticality is greater than a predetermined threshold, performing deviation calculation on the first sample age and other sample ages in the sample user feature set, and counting the number of age deviations less than a predetermined age deviation threshold, setting the number as the third criticality; If the third criticality is greater than a predetermined threshold, the first sample user feature is added to the high-frequency user feature set until the sample user features are traversed to obtain the high-frequency user feature set; Identify key drivers, including: Based on the coefficient of variation method, the weight ratio is configured according to the factor order, where the factor order and weight ratio are negatively correlated; The sample driving factor set is weightedly calculated according to the weight proportion to obtain multiple correlation frequencies of multiple sample driving factors, and the sample driving factor corresponding to the correlation frequency greater than a predetermined frequency threshold is selected and set as the key driving factor.

2. The method for optimizing the comfort of a car seat according to claim 1, characterized in that: Identify several standard user characteristics, including: Merging the high-frequency user feature set and the high-frequency homologous user feature set to obtain a user feature set; Clustering the user feature set according to a predetermined step length to determine a plurality of user feature intervals, wherein the user feature intervals are marked with a degree of aggregation; User feature intervals with aggregation degrees less than a predetermined aggregation threshold are eliminated to obtain a plurality of standard user feature intervals, and intermediate value operations are performed on the plurality of standard user feature intervals to obtain the plurality of standard user features.

3. The method for optimizing the comfort of a car seat according to claim 1, characterized in that: Output the first optimal seat parameters, including: Obtaining a seat parameter adjustment threshold, wherein the seat parameters at least include a material type threshold, a filler type threshold, a cushion length threshold, a backrest height threshold, and a lumbar support height threshold; Randomly select a parameter from the material type threshold, the filling type threshold, the cushion length threshold, the backrest height threshold, and the lumbar support height threshold to combine and generate a plurality of initial seat parameters; Based on the digital twin, the plurality of standard user characteristics are simulated to obtain a plurality of user physical sign models, and the plurality of seat twin models are simulated according to the target vehicle attribute characteristics and the plurality of initial seat parameters; With the purpose of optimizing the seat comfort, based on the multiple user physical sign models and the multiple seat twin models, according to the predetermined test plan, the seat parameters are simulated and optimized using the comfort prediction model to output the first optimal seat parameters.

4. The method for optimizing the comfort of a car seat according to claim 3, characterized in that: Construct a comfort prediction model, including: Retrieve and obtain a sample comfort data set, wherein the sample comfort data includes sample pressure distribution, sample vibration influence, sample noise influence and sample comfort coefficient, and the sample pressure distribution includes sample back average pressure, sample back peak pressure, sample back average pressure, sample back peak pressure, sample back contact area, sample buttocks average pressure, sample buttocks peak pressure, sample buttocks average pressure, sample buttocks peak pressure, and sample buttocks contact area; The sample comfort data set is divided into Q equal parts, and Q parts are selected with replacement to construct a first training set, and the selection is iteratively selected Q times to generate Q data sets; The BP neural network is supervisedly trained using Q data sets to obtain Q convergence prediction units, and the comfort prediction model is constructed based on the Q convergence prediction units, wherein the output of the comfort prediction model is the mean of the output results of the Q convergence prediction units.

5. The method for optimizing the comfort of a car seat according to claim 3, characterized in that: Outputting the first optimal seat parameter includes: randomly selecting a first seat twin model from the plurality of seat twin models; By using the first seat twin model, according to the predetermined test scheme, a simulation test of the plurality of user physical sign models is performed to obtain a plurality of simulation test data sets; Inputting the multiple simulated test data sets into the comfort prediction model, outputting multiple predicted comfort sets, and performing mean calculation to obtain a first comfort level; Multiple comfort levels of multiple seat twin models are obtained in sequence, with the purpose of optimizing by improving the seat comfort level. The seat parameters are simulated and optimized based on the multiple comfort levels to obtain the first optimal seat parameters.

6. The method for optimizing the comfort of a car seat according to claim 5, characterized in that: Simulating and optimizing the seat parameters based on the multiple comfort levels includes: The initial seat parameters are regarded as initial solutions, and multiple initial solutions are arranged from large to small according to the comfort level to obtain an initial solution sequence; The first P initial solutions of the initial solution sequence are selected as optimal solutions, and the last K initial solutions are set as inferior solutions, where the sum of P and K is the number of initial solutions, and K is an integer multiple of P; Taking P optimal solutions as the center, randomly cluster the K inferior solutions into equal groups to obtain P domains, where the number of inferior solutions in each domain is the same; In P areas, the optimal solution is used as the optimization direction, and the inferior solution in the same area is mutated once according to the predetermined mutation step length to obtain P updated areas. If the inferior solution after mutation does not meet the seat parameter adjustment threshold, this mutation is not performed; Identify P update areas. If the comfort level of the inferior solution is greater than that of the superior solution in the same update area, replace the superior solution with the inferior solution. Iterate the optimization until the predetermined number of optimizations is met, output P current areas, set the current area with the largest sum of comfort as the optimal area, and select the optimal solution of the optimal area as the first optimal seat parameter.

7. A system for optimizing the comfort of automobile seats, characterized in that: The steps for implementing the method for optimizing the comfort of a car seat as claimed in any one of claims 1 to 6 include: A positive sample retrieval module is used to perform positive sample retrieval based on the Internet of Vehicles with the first seat of the target vehicle as a constraint to obtain a high-frequency user feature set and a sample driving factor set, wherein the first seat is any one of the multiple seats of the target vehicle; A feature determination module, configured to perform cluster analysis based on the sample driving factor set to determine key driving factors, perform positive sample retrieval based on the key driving factors, determine a high-frequency homologous user feature set, fuse the high-frequency user feature set with the high-frequency homologous user feature set, and determine a plurality of standard user features; A test scheme configuration module is used to retrieve the driving road surface characteristics of the target vehicle, obtain a road surface characteristic set, perform a comprehensive analysis based on the road surface characteristic set, obtain a road surface characteristic proportion, and configure a predetermined test scheme according to the road surface characteristic proportion, wherein the road surface characteristics include road surface type and distance proportion; A seat parameter optimization module is used to simulate and optimize the seat parameters based on the digital twin, with the seat parameter adjustment threshold as the optimization space and the improvement of seat comfort as the optimization purpose, according to the multiple standard user characteristics and the predetermined test plan, using the comfort prediction model to output the first optimal seat parameters, and perform seat optimization design of the target vehicle according to the first optimal seat parameters; Obtain high-frequency user feature sets and sample driving factor sets, including: Taking the first seat of the target vehicle as a constraint, positive sample retrieval is performed based on the Internet of Vehicles to obtain a sample user feature set and a sample driving factor set, wherein the user features include at least height, weight and age, and the driving factors include at least price index, vehicle size and functional configuration, and the driving factors are marked with factor order; Randomly selecting a first sample user feature from the sample user feature set, wherein the first sample user feature includes a first sample height, a first sample weight, and a first sample age; Calculate the deviation between the first sample height and the heights of other samples in the sample user feature set, and count the number of height deviations less than a predetermined height deviation threshold, and set it as a first criticality; If the first criticality is greater than a predetermined threshold, a deviation calculation is performed on the first sample weight and other sample weights in the sample user feature set, and the number of weight deviations less than a predetermined weight deviation threshold is counted and set as a second criticality; If the second criticality is greater than a predetermined threshold, performing deviation calculation on the first sample age and other sample ages in the sample user feature set, and counting the number of age deviations less than a predetermined age deviation threshold, setting the number as the third criticality; If the third criticality is greater than a predetermined threshold, the first sample user feature is added to the high-frequency user feature set until the sample user features are traversed to obtain the high-frequency user feature set; Identify key drivers, including: Based on the coefficient of variation method, the weight ratio is configured according to the factor order, where the factor order and weight ratio are negatively correlated; The sample driving factor set is weightedly calculated according to the weight proportion to obtain multiple correlation frequencies of multiple sample driving factors, and the sample driving factor corresponding to the correlation frequency greater than a predetermined frequency threshold is selected and set as the key driving factor.

Images