Correlation analysis and foaming optimization design method based on foaming and seat comfort subjective evaluation

Abstract

The invention relates to the technical field of automobile seat riding comfort, and provides a correlation analysis and foaming optimization design method based on foaming and seat comfort subjective evaluation. Comprising the steps that human body characteristic parameters and riding comfort set scores of a target user group are input into a trained seat foaming physical property and riding comfort relevancy model, and optimal physical property parameters, capable of guiding seat foaming material type selection and design, of automobile seat foaming are obtained. According to the method, the correlation between the foaming physical property parameters and the seat riding comfort score is constructed, the problem that the nonlinear relation between the foaming design parameters and the user comfort feeling is difficult to describe is solved, and technical support is provided for efficient and accurate optimization design of seat foaming.

Description

Technical Field

[0001] This invention belongs to the field of automotive seat comfort technology, specifically relating to a correlation analysis and foam optimization design method based on the relationship between foaming and subjective evaluation of seat comfort, applicable to automotive seat comfort optimization design. Background Technology

[0002] In the field of seat comfort optimization design, one type of research focuses on calibrating and refining objective and subjective evaluation indicators of seat comfort by analyzing and establishing the correlation between objective indicators (such as pressure distribution) and subjective indicators (occupant subjective ratings). Another type of research focuses on the correlation between occupant posture parameters, seat ergonomic dimensions, and seat ergonomics evaluation, providing relatively objective evaluation parameters for the optimized design and layout of seat posture forms. However, there is currently little research on the relationship between the physical properties of seat components and seat comfort.

[0003] Among the various components of a car seat, foam is a key component, playing a role in support, vibration reduction, and cushioning, and greatly affecting occupant comfort. The various parameters of the foam directly affect the seat's elasticity, support, and durability, and are important factors determining the seat's elastic characteristics. The setting and design of these foam parameters directly determine the seating experience. Therefore, researching the relationship between foam properties and the comfort of car seats becomes crucial.

[0004] CN119830595A provides a method for optimizing vehicle seat comfort. First, it acquires seating posture test data for the test subject in both static and driving states. Then, based on the seating posture test data, it determines the seat force model. Finally, based on the seat force model, it generates seat comfort optimization strategies, including material optimization strategies, softness optimization strategies, and seat contour optimization strategies. CN120296867A provides a method for optimizing seat parameters based on driver comfort. Based on a vehicle finite element model, a dummy model, and a pre-established finite element model, it determines seat parameters including seat back angle, seat cushion angle, seat leg support angle, and seat filling stiffness. It then iteratively optimizes the model parameters based on seat comfort simulation results, thereby improving seat comfort while ensuring driver safety. CN118500530A discloses a method for subjective and objective evaluation of mechanical vibration comfort. It establishes a subjective and objective evaluation mapping model for vibration comfort under different working conditions to calibrate vibration comfort evaluation indicators. CN119089590A provides a method for optimizing the design of automotive seat comfort, including: fusing high-frequency user feature sets and high-frequency homogeneous user feature sets to determine multiple standard user features; configuring a predetermined test plan based on the proportion of road surface features; based on digital twins, using seat parameter adjustment thresholds as the optimization space, with the aim of improving seat comfort, and using a comfort prediction model to simulate and optimize seat parameters according to multiple standard user features and predetermined test plans, outputting the first optimal seat parameters to execute seat optimization design; CN120716536A provides a method for seat comfort optimization calculation based on sensors and AI modeling, which establishes a pressure distribution prediction model for each zone by collecting occupant sitting posture and pressure distribution data in real time; then, by inputting average pressure, maximum pressure, pressure center of gravity, body shape parameters, and pathological weight factors, the artificial intelligence model outputs the target support pressure value, pressure avoidance curve, and neighboring zone migration ratio for each zone, and combines feedback monitoring to achieve adaptive control of different areas of the occupant's spine.

[0005] The existing technical solutions described above have two significant drawbacks: (1) The influence of the physical properties of seat components on the subjective feeling of passenger comfort is not considered, thus limiting its guiding role in seat comfort design; it cannot solve the problem of how to accurately map passenger sitting posture and pressure distribution data to the selection and optimization of seat components such as foam. (2) Subjective evaluation indicators of comfort often reflect the comprehensive influence of design factors in multiple fields such as seat surface design, human-machine layout, physical properties, and frame structure, making it difficult to provide quantitative reference for the parameter design of seat comfort. Summary of the Invention

[0006] To address at least one of the problems existing in current technologies, this invention provides a correlation analysis and foam optimization design method based on the relationship between foaming and subjective evaluation of seat comfort. By using a subjective evaluation scoring rule for seat comfort based on the performance of the foam itself, it overcomes the deficiency of existing subjective evaluation methods in decoupling and establishing the correspondence between objective seat design parameters and user comfort perception. Simultaneously, by introducing a self-learning artificial neural network model, a correlation is constructed between foaming material properties and seat comfort scores, solving the problem of describing the nonlinear relationship between foaming design parameters and user comfort perception. This provides technical support for the efficient and precise optimization design of seat foaming.

[0007] To achieve the objectives of this invention, this invention provides a correlation analysis and foam optimization design method based on the correlation between foaming and subjective evaluation of seat comfort. The method inputs the anthropometric parameters of the target user group and their ride comfort scores into a pre-trained correlation model of seat foaming properties and ride comfort, obtaining optimal physical property parameters for automotive seat foaming that can guide material selection and design. The correlation model of seat foaming properties and ride comfort is obtained through the following steps: Collect human characteristic parameters from the selected target user group; Collect data on the physical properties of seat foam; The selected target user group rates the seat comfort, which is based on subjective evaluation scoring rules, and the subjective evaluation scoring rules are determined based on the performance of the foam body to determine the seat comfort. A training dataset was constructed based on human feature parameters, seat foam properties, and rating data. An artificial neural network model was used to construct the correlation between seat foam material properties and ride comfort. The training dataset was imported to train the artificial neural network model, resulting in a correlation model between seat foam material properties and ride comfort.

[0008] Furthermore, the human body characteristic parameters include height and weight characteristic parameters.

[0009] Furthermore, the material properties of the seat foam include the thickness, density, hardness, hysteresis loss rate, indentation ratio, and resilience of the seat cushion foam and backrest foam.

[0010] Furthermore, when rating the comfort of the seats, the ratings are based on the softness of the seat cushion, the softness of the backrest, the cushion cushioning, the backrest cushioning, the seat cushion support, and the backrest support.

[0011] Furthermore, before constructing the training dataset, the human feature parameters, seat foam properties, and scoring data are normalized.

[0012] Furthermore, the artificial neural network model is any one of the following neural network structures: BP, SVM, and LSTM.

[0013] This invention also provides a system for correlation analysis and foam optimization design based on the relationship between foaming and subjective evaluation of seat comfort, comprising the following modules: The optimization module is used to input the anthropometric parameters of the target user group and the set score of ride comfort into the trained correlation model of seat foam properties and ride comfort, so as to obtain the optimal physical property parameters of car seat foam that can be used to guide the material selection and design of seat foam.

[0014] Furthermore, it also includes the following modules: a dataset construction module, used to collect human characteristic parameters and seat foam material properties data of the selected target user group, and to have the selected target user group score the seat comfort to construct a dataset, wherein the scoring is based on subjective evaluation scoring rules, which are determined based on the seat comfort of the foam body performance; The training module is used to construct the correlation between seat foam material properties and ride comfort using an artificial neural network model. The training dataset is imported to train the artificial neural network model, resulting in a correlation model between seat foam material properties and ride comfort.

[0015] The present invention also provides a computer device.

[0016] The present invention also provides a computer-readable storage medium.

[0017] Compared with the prior art, the present invention can achieve at least the following beneficial effects: (1) Existing seat comfort evaluation rules in the industry reflect the comprehensive feeling of design factors in multiple fields such as seat surface design, ergonomic layout, material properties, and frame structure. Decoupling these factors is very difficult, so it is often difficult to establish a correspondence between objective seat design parameters and comfort feelings. This invention proposes a subjective evaluation scoring rule for seat comfort based on the performance of foam body. This scoring rule can provide feasibility and convenience for establishing a quantitative relationship between seat foam design parameters and comfort.

[0018] (2) This invention achieves a precise quantitative description of the nonlinear relationship between foam design and comfort by using objective parameters of seat foam design and direct data on riding comfort, as well as introducing an artificial neural network model with self-learning capabilities, thus providing technical support for seat product design centered on the actual comfort experience of users. Attached Figure Description

[0019] Figure 1This is a flowchart illustrating a method for correlation analysis and foam optimization design based on subjective evaluation of foaming and seat comfort, provided by an embodiment of the present invention.

[0020] Figure 2 This is a schematic diagram of the measurement of foaming compression-unloading characteristic data in an embodiment of the present invention.

[0021] Figure 3 This is a schematic diagram of the seat foaming parameter design based on a neural network model in an embodiment of the present invention. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0023] Please see Figure 1 The present invention provides a method for correlation analysis and foam optimization design based on the relationship between foaming and subjective evaluation of seat comfort, comprising the following steps: Step 1: Select the target user and collect human feature parameters.

[0024] In this step, target users are selected based on the distribution patterns of the Chinese anthropometric database. Based on the selected target users, their height and weight characteristics are obtained.

[0025] In one embodiment, the target user group for the seat is selected from at least 20 people. The specific parameters for the method of selecting the body type of the people are based on Table 1, and the height and weight characteristic parameters of the selected target user group are collected.

[0026] Table 1. Composition of the target user group

[0027] Step 2: Collect data on the physical properties of the seat foam.

[0028] In this step, through foaming performance testing and analysis, we obtain physical property data such as the thickness, density, hardness, hysteresis loss rate, indentation ratio, and resilience of the seat cushion foam and backrest foam.

[0029] In one embodiment, for selected seat foam components, foam properties and compression-unloading characteristics are tested and data is collected, with the following control requirements: a. Foam thickness test Select no fewer than 5 test points in the seat cushion and backrest area, measure and record the foam thickness value of the test points.

[0030] b. Foaming density test According to the requirements of GB / T 6343 Determination of Apparent Density of Foamed Plastics and Rubber.

[0031] c. Obtain the hardness value of the foam through foam compression-unloading characteristic testing. According to the requirements of GB / T 10807 for the determination of hardness of flexible foam polymer materials.

[0032] In one embodiment, such as Figure 2 As shown, the indenter is slowly pressed down along the normal direction of the foamed sample surface at a speed of (100mm±20mm) / min until it reaches 75%±2.5% of the initial sample thickness. Then, it is unloaded at the same speed to obtain the desired result. Figure 2 The force-displacement curves for the foaming compression-unloading process are shown. The displacement at the initial 3N preload is set as the origin of the coordinate system, with the loading force on the Y-axis and the displacement on the X-axis. The curves indicate the pressure values ​​at 25%, 65%, and 75% of the sample thickness.

[0033] In one embodiment, based on the force-displacement curve of the compression-unloading process, the pressure value corresponding to when the indenter presses into the sample thickness of 25% is taken as the foam hardness value.

[0034] The foaming collapse ratio is calculated according to the following formula (1): ………………………(1) In the formula: ： Indentation ratio; Fc25%: The pressure (N) when the sample thickness is 25% indented; Fc65%: The pressure (N) when the sample thickness is pressed in to 65%.

[0035] The hysteresis loss rate of foaming is calculated according to the following formula (2): ………………(2) In the formula: ： Lag loss rate (%); A(oabcd): The area enclosed by oabcd; A(oabeo): The area enclosed by oabeo.

[0036] d. Rebound test According to the requirements of section 5.1 of GB / T 6670, "Determination of Rebound Performance of Flexible Foam Polymer Materials by Falling Ball Method".

[0037] The rebound rate is calculated according to formula (3): …………………………………………(3) In the formula: R Rebound rate; h 1: Rebound height value (mm); h Falling height (mm).

[0038] Step 3: Selected target users rate the seat comfort, which is based on subjective evaluation scoring rules, and the subjective evaluation scoring rules are determined based on the performance of the foam body to determine the seat comfort.

[0039] In this step, subjective evaluations of seat comfort are conducted to obtain ratings for seat cushion softness, backrest softness, seat cushion cushioning, backrest cushioning, seat cushion support, and backrest support.

[0040] In one embodiment, for the foamed seat cushion and backrest being tested, selected target users subjectively evaluate the comfort of the seats, obtaining scores for seat cushion softness, backrest softness, seat cushion cushioning, backrest cushioning, seat cushion support, and backrest support. The control requirements are shown in Table 2. A ten-point scoring method is used to divide the seat comfort into 7 levels, with each level corresponding to a score in the corresponding segment.

[0041] Table 2 Seat Comfort Scoring Requirements (Subjective Evaluation Scoring Rules)

[0042] Step 4: Normalize the human body feature parameters, seat foam properties data, and scoring data to establish a training dataset.

[0043] This step cleans and normalizes the data collected in steps 1, 2, and 3 to create a training dataset.

[0044] In one embodiment, the normalization calculation formula is as follows:

[0045] in: =

[0046] =

[0047] In the formula, the sample set X = { } , = 1, …, , The sample mean. For the number of samples, This represents the sample variance.

[0048] Normalizing data can make the model converge to the optimal solution more easily, speed up the solution process, and improve the model's accuracy. In other embodiments, however, some deep learning models (such as the NFNet model) can achieve high accuracy without normalization; furthermore, different normalization algorithms can be selected based on different needs.

[0049] Step 5: Build and train a correlation model between seat foam properties and ride comfort.

[0050] An artificial neural network model is used to construct the correlation between seat foam properties and ride comfort, and is trained according to requirements. The training dataset obtained in step 4 is imported to train the artificial neural network model, and training / validation accuracy curves are plotted. Through continuous iterative training, the artificial neural network model achieves the required prediction accuracy. In one embodiment, the artificial neural network model can employ any neural network structure such as BP, SVM, or LSTM.

[0051] The construction and training of the correlation model between seat foam properties and ride comfort is a key step in seat comfort evaluation and optimization. In one embodiment, the control requirements are as follows: The height and weight of the occupants (target users), foam thickness, foam density, foam hardness, foam indentation ratio, foam hysteresis loss rate, and foam rebound rate are defined as the input values ​​of the correlation model, denoted as x1, x2, x3, x4, x5, x6, x7, and x8, respectively. The seat cushion softness, backrest softness, seat cushion cushioning, backrest cushioning, seat cushion support, and backrest support ratings are defined as the output values ​​of the correlation model, denoted as y1, y2, y3, y4, y5, and y6, respectively. Based on the above data definition method, training and validation datasets are created.

[0052] To create an artificial neural network, configure the number of hidden layer neurons, learning rate, neuron activation function, minimum mean squared error target, minimum performance gradient, and maximum number of training iterations according to actual needs. In one embodiment, a three-layer neural network is created, using the Levenberg-Marquardt function as the training function, with a learning efficiency of 0.8 and a control error of 1 × 10⁻⁶.-6 The maximum number of training iterations was 50,000. The three-layer neural network was trained to obtain the correlation model between the seat foam material properties and the ride comfort.

[0053] Table 3 Validation Table of Neural Network Training Results

[0054] Step 6: Using the correlation model of seat foam properties and ride comfort trained in Step 5, input the anthropometric parameters of the target user group and the ride comfort setting score, and output the optimal physical property parameters of the car seat foam (i.e., Figure 1 The recommended design parameters for foaming are provided to guide engineers in selecting and designing materials for seat foaming by using optimal physical property parameters.

[0055] In one embodiment, by using a trained neural network model and inputting the human characteristic parameters of the target user group (selection principles are the same as in step 1) and the seat comfort setting score, the optimal parameters of the seat foam thickness, density, hardness, indentation ratio, hysteresis loss rate, and rebound rate can be output to guide seat engineers in selecting and optimizing foam materials.

[0056] In one embodiment, a system for correlation analysis and foam optimization design based on the relationship between foaming and subjective evaluation of seat comfort is provided, comprising the following modules: The optimization module is used to input the anthropometric parameters of the target user group and the set score of ride comfort into the trained correlation model of seat foam properties and ride comfort, so as to obtain the optimal physical property parameters of car seat foam that can be used to guide the material selection and design of seat foam.

[0057] It also includes the following modules: a dataset construction module, which is used to collect human feature parameters and seat foam material properties data of the selected target user group, and to have the selected target user group score the seat comfort to construct a dataset, wherein the scoring is based on subjective evaluation scoring rules, which are determined based on the seat comfort of the foam body performance; The training module is used to construct the correlation between seat foam material properties and ride comfort using an artificial neural network model. The training dataset is imported to train the artificial neural network model, resulting in a correlation model between seat foam material properties and ride comfort.

[0058] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the method provided in the foregoing embodiments.

[0059] In one embodiment, a computer-readable storage medium is provided, the computer-readable storage medium storing a computer program, characterized in that, when executed by a processor, the computer program implements the method provided in the foregoing embodiments. During the project development phase, when it is necessary to select and optimize the materials for seat foaming, a trained neural network model can be used. By inputting the human characteristic parameters of the target user group and the seat comfort setting score, the optimal parameters of seat foaming thickness, density, hardness, indentation ratio, hysteresis loss rate, and rebound rate can be output to guide seat engineers in selecting and optimizing foaming materials.

[0060] This invention, through a subjective rating rule for seat comfort based on the properties of the foam itself, overcomes the shortcomings of existing subjective evaluation methods in achieving indicator decoupling and accurate assessment of foam performance. By introducing a self-learning artificial neural network model, it constructs and optimizes the correlation between foam performance and seat comfort, solving the problem of describing the nonlinear relationship between seat material properties and comfort scores. Quantitatively predicting the comfort score of a seat based on the material properties of a specific type of foam can provide technical support for the efficient and precise optimization design of seat foam.

[0061] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined in this invention may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for correlation analysis and foam optimization design based on subjective evaluation of foaming and seat comfort, characterized in that, The anthropometric parameters of the target user group and the set scores for ride comfort are input into a pre-trained correlation model of seat foam properties and ride comfort. This yields the optimal physical property parameters for automotive seat foam that can guide the material selection and design of seat foam. The correlation model of seat foam properties and ride comfort is obtained through the following steps: Collect human characteristic parameters from the selected target user group; Collect data on the physical properties of seat foam; The selected target user group rates the seat comfort, which is based on subjective evaluation scoring rules, and the subjective evaluation scoring rules are determined based on the performance of the foam body to determine the seat comfort. A training dataset was constructed based on human feature parameters, seat foam properties, and rating data. An artificial neural network model was used to construct the correlation between seat foam material properties and ride comfort. The training dataset was imported to train the artificial neural network model, resulting in a correlation model between seat foam material properties and ride comfort.

2. The method for correlation analysis and foam optimization design based on subjective evaluation of foaming and seat comfort as described in claim 1, characterized in that, The human body characteristic parameters include height and weight.

3. The method for correlation analysis and foam optimization design based on subjective evaluation of foaming and seat comfort as described in claim 1, characterized in that, The physical properties of the seat foam include the thickness, density, hardness, hysteresis loss rate, indentation ratio, and rebound rate of the seat cushion foam and backrest foam.

4. The method for correlation analysis and foam optimization design based on subjective evaluation of foaming and seat comfort as described in claim 1, characterized in that, When rating the comfort of a seat, the rating is based on the softness of the seat cushion, the softness of the backrest, the cushion cushioning, the backrest cushioning, the seat cushion support, and the backrest support.

5. A method for correlation analysis and foam optimization design based on subjective evaluation of foaming and seat comfort according to any one of claims 1-4, characterized in that, Before constructing the training dataset, the human body feature parameters, seat foam properties data, and rating data are normalized.

6. A method for correlation analysis and foam optimization design based on subjective evaluation of foaming and seat comfort according to any one of claims 1-5, characterized in that, The artificial neural network model is any one of the following neural network structures: BP, SVM, and LSTM.

7. A system for correlation analysis and foam optimization design based on subjective evaluation of foaming and seat comfort, characterized in that, Includes the following modules: The optimization module is used to input the anthropometric parameters of the target user group and the set score of ride comfort into the trained correlation model of seat foam properties and ride comfort, so as to obtain the optimal physical property parameters of car seat foam that can be used to guide the material selection and design of seat foam.

8. The foaming optimization design system based on correlation analysis of foaming and subjective evaluation of seat comfort according to claim 7, characterized in that, It also includes the following modules: a dataset construction module, which is used to collect human feature parameters and seat foam material properties data of the selected target user group, and to have the selected target user group score the seat comfort to construct a dataset, wherein the scoring is based on subjective evaluation scoring rules, which are determined based on the seat comfort of the foam body performance; The training module is used to construct the correlation between seat foam material properties and ride comfort using an artificial neural network model. The training dataset is imported to train the artificial neural network model, resulting in a correlation model between seat foam material properties and ride comfort.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method according to any one of claims 1-6.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the method described in any one of claims 1-6.

Images