Diversified sign passenger positioning method for automobile safety virtual evaluation

Abstract

The invention relates to the technical field of ergonomics, in particular to a diversified physical sign passenger positioning method for automobile safety virtual evaluation, which comprises the following steps: S1, constructing a diversified physical sign passenger positioning sample data set which comprises physical sign parameters, vehicle interior parameters and constraint system parameters; s2, using the passenger positioning sample data set to construct and verify a passenger positioning parameter prediction model; s3, human body finite element model positioning is carried out based on the passenger positioning parameter prediction model; and S4, checking and correcting a positioning result of the human body finite element model. According to the method, a standardized positioning method can be provided for passenger postures in different physical signs and different interior environments, so that the reliability and comparability of a virtual collision test are improved, and meanwhile, the processing efficiency before simulation is also improved.

Description

Technical Field

[0001] The invention relates to the technical field of ergonomics, and in particular to a diversified vital sign occupant positioning method for automobile safety virtual evaluation. Background Art

[0002] At present, the passive safety collision test of vehicles is mainly carried out based on physical experiments based on standard dummy models. With the advancement of virtual simulation technology, the virtual evaluation system is gradually maturing. In the evaluation of the virtual evaluation system, the use of the Human Body Finite Element Model (HBFEM) is an important tool for evaluating the safety performance of vehicles. The initial position and posture of the occupant and the relative relationship with the internal structure of the vehicle have an important impact on the risk of injury during the collision. Therefore, a reasonable occupant positioning scheme can not only improve the authenticity and credibility of the simulation, but also provide valuable reference for vehicle safety design.

[0003] However, there are still some problems in the positioning scheme for multi-body occupants (i.e. people of different ages, genders, body shapes, etc.). Traditional HBFEM is usually based on a standard adult human model, using human body dimensions of specific percentiles (95th percentile for men, 50th percentile for men and women, 5th percentile for women), so the protection effect assessment for non-standard body occupants (such as children, the elderly, obese people, etc.) may be inaccurate.

[0004] At the same time, the traditional HBFEM human body model only considers occupants in a standard sitting posture. However, in actual situations, the occupants may make dynamic adjustments during the seating positioning process. There is no unified standard, and a single standard sitting posture human body model cannot cover the occupants' actual sitting posture, and the evaluation results are not comprehensive enough.

[0005] Therefore, it is urgent to build a positioning method for developing diversified vital sign occupant models to match real traffic accident scenarios and improve the accuracy of virtual assessments. Summary of the invention

[0006] The present invention provides a diversified vital sign occupant positioning method for automobile virtual safety assessment. Through parametric modeling, a standardized positioning method can be provided for occupant postures with different vital signs and in different interior environments, thereby improving the reliability and comparability of virtual collision tests, while also improving the processing efficiency before simulation.

[0007] This application provides the following technical solutions:

[0008] A diversified vital sign occupant positioning method for automobile safety virtual assessment includes the following steps:

[0009] S1. Constructing a sample data set of occupant positioning with diversified physical signs, wherein the sample data set of occupant positioning includes physical sign parameters, vehicle interior parameters and restraint system parameters;

[0010] S2, using the occupant positioning sample data set to build and verify the occupant positioning parameter prediction model;

[0011] S3, positioning the human body finite element model based on the occupant positioning parameter prediction model;

[0012] S4. Check and correct the positioning results of the human body finite element model.

[0013] Furthermore, the physical parameters include gender, weight, height, and weight coefficient; the vehicle interior parameters include seat height, distance between the steering wheel and the sole reference point, and seat cushion angle; the restraint system parameters include seat reference point, steering wheel center point, steering wheel diameter, steering wheel angle, seat track angle, and accelerator pedal center point.

[0014] Further, the S1 includes:

[0015] S11, obtaining the vital sign parameters of the occupant;

[0016] S12, obtaining vehicle interior parameters and restraint system parameters;

[0017] S13, measuring occupant positioning parameters in different vehicle interior environments;

[0018] S14. Integrate the volunteers' physical parameters, vehicle interior parameters, restraint system parameters and occupant positioning parameters to establish an occupant positioning sample data set.

[0019] Further, the occupant positioning parameters include size parameters and angle parameters, the size parameters include the X-axis size of the hip joint relative to the plantar reference point, the X-axis size of the hip joint relative to the H point, and the Z-axis size of the hip joint relative to the H point; the X-axis size of the eye relative to the plantar reference point, the Z-axis size of the eye relative to the heel point behind the accelerator pedal, the X-axis size of the eye relative to the hip joint, and the Z-axis size of the eye relative to the hip joint; the X-axis size of the ankle relative to the plantar reference point; the X-axis size of the ankle relative to the accelerator pedal, and the Z-axis size of the ankle relative to the heel reference point behind the accelerator pedal;

[0020] The angle parameters include the angle between the hip joint and the eye, the knee angle, the head angle, the neck angle, the chest angle, the abdomen angle, and the pelvic angle.

[0021] Further, the S2 includes:

[0022] S21. Divide the occupant positioning sample data set into two parts, using 80% of the data as the model building data set and 20% of the data as the verification data set.

[0023] S22, constructing an occupant positioning parameter prediction model based on stepwise regression technology, wherein the occupant positioning parameter prediction model is a set of multiple prediction sub-models, wherein the multiple prediction sub-models are respectively used to predict each occupant positioning parameter defined in S13, and a specific significance level is used as a criterion for selecting input variables, and finally the optimal prediction sub-model is selected;

[0024] S23. Evaluate the prediction effect.

[0025] Furthermore, the prediction sub-model is a regression prediction equation with optimal coefficients, the input variables of the regression prediction equation are selected physical parameters, vehicle interior coefficients, and restraint system parameters, and the output of the regression prediction equation is the occupant positioning parameter.

[0026] Further, the S3 includes:

[0027] S31, predicting the hip position of the occupant using the prediction sub-model;

[0028] S32, predicting the position of the eye relative to the hip using the prediction sub-model;

[0029] S33, Fitting the torso and limbs.

[0030] Further, the S31 includes:

[0031] The prediction model is used to predict the front and rear hip positions, and the offset vector from the hip to point H is calculated to obtain the movement path of the hip in the XZ plane. According to the predicted front and rear hip positions, the points on the path are determined as the predicted hip positions. The Y coordinates of the left and right hip joints are equidistant from the center line of the seat, and the torso segment is located in the XZ plane of the center line of the seat;

[0032] The S32 includes:

[0033] The prediction model is used to calculate the position of the eye relative to the hip. The Z coordinate of the predicted eye position is equal to the canthus landmark point, and the X coordinate is equal to the infraorbital landmark point.

[0034] The S33 includes:

[0035] First, adjust the trunk above the L5-S1 intervertebral disc to achieve the predicted pelvic angle, and remeasure the abdominal angle, chest angle, neck angle, and head angle; then keep the pelvic angle and below unchanged, adjust the abdominal angle (adjust the lumbar spine) to the predicted angle, and recalculate the chest angle, neck angle, and head angle; keep the abdominal angle and below unchanged, adjust the thoracic spine to the predicted chest angle, and recalculate the neck angle and head angle; keep the chest angle and below unchanged, adjust the cervical spine to the predicted neck angle, recalculate the head angle, adjust the head to the predicted head angle, and complete the trunk positioning;

[0036] The driver's hands are placed at 2 o'clock and 10 o'clock on the steering wheel, and the lower limbs are placed according to the relative positions of the ankle and hip joints, the coordinate origin, and the predicted values ​​of the knee angles calculated. Other unspecified positions are judged to see whether they conform to the real human posture.

[0037] Further, the S4 includes:

[0038] S41. Check the angles of the head and limbs;

[0039] S42. Adjust the fitting result according to the inspection result.

[0040] Principles and advantages of the present invention:

[0041] The technical solution of the present invention first measures factors that have a significant impact on occupant positioning as input variables, and divides the input variables into occupant variables and vehicle variables. The occupant variables include gender, age, height, seat height, and weight; the vehicle variables include seat height, seat cushion angle, distance of the steering wheel relative to the sole reference point, pedal center, seat reference point position, steering wheel diameter, steering wheel center, steering wheel angle, etc.

[0042] Then, a regression prediction model is constructed based on statistical theory, and appropriate input variable combination data is selected to calculate the positioning parameters of each part of the occupant's body; based on the calculated body positioning parameters, the body parts of the human model are placed in sequence, and the undefined body parameters are determined by the inverse kinematics method. The positioning results are checked and corrected through the head angle and limb angle.

[0043] The present invention links human body characteristics with parameters such as vehicle interior. The prediction model can calculate corresponding positioning parameters before positioning according to the different body shapes of occupants and different vehicle interiors, which greatly improves the application scope of virtual evaluation. At the same time, the approximate position of the reference point is calculated before the model takes the seat, which improves the efficiency of simulation pre-processing.

[0044] In addition, compared with physical experiments, virtual evaluation systems can be closer to real traffic accidents, while various human models often have large differences. Compared with the relatively simple positioning parameters of physical dummies, the positioning of human models requires more and more detailed parameters. This solution forms a set of standardized methods for the positioning of human models in virtual evaluation, which improves the accuracy and scalability of virtual evaluation. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 Flow chart of the method for occupant positioning with diverse vital signs;

[0046] Figure 2 Schematic diagram for vehicle variable measurement;

[0047] Figure 3Flowchart constructed for the occupant positioning parameter prediction model. DETAILED DESCRIPTION

[0048] After the human body is seated, studies have shown that certain parameters such as eye position and hip position will have a significant impact on the positioning of the occupants. At the same time, some interior parameters and restraint system parameters in the car cabin, such as steering wheel position and accelerator pedal position, may also affect the positioning of the occupants. For example, studies have shown that the steering wheel position has a significant effect on the anterior-posterior position of the hip joint, elbow angle, knee angle and trunk inclination angle, and most of them are linear relationships; the seat cushion angle has little effect on posture, mainly affecting the hip joint position, trunk inclination angle, lumbar flexion angle and knee angle; the seat height has a significant nonlinear effect on the hip joint position, has no significant effect on other posture variables, and has no significant interaction with the seat cushion angle; gender has no significant difference in the effect on the test variables. These factors have a potential statistical relationship with the occupants, so the occupant's physical parameters, vehicle interior parameters and restraint system parameters are considered to be included to establish an occupant positioning model.

[0049] The following is further described in detail through specific implementation methods:

[0050] Embodiment 1

[0051] A diversified occupant positioning method for virtual vehicle safety assessment, such as Figure 1 As shown, the following steps are included:

[0052] S1. Construct a sample data set of occupant positioning with diverse physical signs. The sample data set of occupant positioning includes physical sign parameters, vehicle interior parameters and restraint system parameters. The specific steps include:

[0053] S11. Obtaining the occupant's vital signs parameters:

[0054] A certain number of volunteers are recruited. Considering the workload of subsequent data measurement, the volunteers recruited in this embodiment are about 15. The recruited volunteers have different types of physical signs, including gender, weight, height, and weight coefficient.

[0055] The physical parameters of each volunteer were recorded, including the height, sitting height, weight and weight coefficient of the volunteer. The unit of height and sitting height was millimeter, and the unit of weight was kilogram.

[0056] S12. Obtaining vehicle interior parameters and restraint system parameters:

[0057] Vehicle interior parameters reflect the factors that affect the positioning of passengers after they sit down, including seat height, distance between the steering wheel and the reference point of the sole of the foot, and seat cushion angle.

[0058] The seat height is defined as the height difference between the seat reference point and the interior origin; the interior origin is defined as the intersection of a horizontal line through the accelerator pedal heel point AHP and a vertical line through the foot base reference point BOF. The accelerator pedal heel point AHP is defined as the intersection of the heel point of the H-point device with the floor surface (taking into account the amount of carpet collapse) when the shoe fixture is properly placed (when the sole remains on the accelerator pedal in a free state), and the foot base reference point BOF is defined as the contact point between the driver's shoe sole and the pedal, and its lateral position is located at the center line of the shoe sole;

[0059] The distance between the steering wheel and the foot sole reference point is defined as the relative distance between the steering wheel center and the interior trim origin in the X direction;

[0060] The seat cushion angle is defined as the angle between the seat cushion reference plane and the horizontal plane;

[0061] The restraint system parameters reflect the factors affecting the adjustment and positioning of the occupant after the vehicle's restraint system is seated, including SgRP (XYZ), steering wheel center (XYZ), steering wheel diameter, steering wheel angle, seat track angle, and accelerator pedal center (Y). The unit of the restraint system parameters is millimeter, where:

[0062] SgRP is defined as the seat reference point, which is the design reference point specified by the seat manufacturer. The seat reference point takes into account all seat adjustment forms (horizontal, vertical and tilt) and determines the final position of the seat during normal driving or riding. When 95% of the human models are placed on the seat as specified, the actual H point should coincide with the seat reference point.

[0063] The steering wheel angle is defined as the angle between the steering wheel and the YZ plane;

[0064] The steering wheel center is defined as the center point of the steering wheel geometry;

[0065] The steering wheel diameter is defined as the maximum width passing through the center of the steering wheel in the X-axis direction;

[0066] The seat track angle is defined as the angle between the seat track and the horizontal plane.

[0067] The vehicle interior parameters and restraint system parameters are measured as follows: Figure 2 shown.

[0068] S13. Measurement of occupant positioning parameters in different vehicle interior environments:

[0069] Different vehicle interior environments include the driver's side position on different vehicles. For example, if vehicle A, vehicle B, vehicle C, and vehicle D are tested at the same time, a total of 4 different interior environments are provided. In each interior environment, the volunteer adjusts the steering wheel and seat according to his or her own feelings after sitting down. After the volunteer fixes the sitting posture, the occupant positioning parameters of the volunteer are measured and recorded.

[0070] The occupant positioning parameters are determined according to the positioning feature points of the human body model. The present invention defines the positioning feature points of the human body model as follows:

[0071] Glabellar point: the unstressed skin surface point obtained by palpating the most anteriorly prominent point of the forehead in the midline at the level of the brow ridge;

[0072] Infraorbital point: The unpressed skin surface point obtained by palpating the lowest edge of the orbit (eye socket);

[0073] Tragus point: The unpressed skin surface point obtained by palpating the most anterior edge of the cartilaginous notch above the tragus (located at the upper edge of the external auditory canal);

[0074] Occipital point: The unpressed skin surface point on the posterior inferior prominence of the occipital bone. The hair is lightly pressed;

[0075] Canthus point: the unpressurized skin surface point at the outer junction of the upper and lower eyelids;

[0076] Spinal spinous process: the concave skin surface point at the rearmost of the corresponding spinous process, including the seventh cervical spinous process, the eighth thoracic spinous process, and the twelfth thoracic spinous process;

[0077] Suprasternal notch (manubrium): an uncompressed skin surface point on the midline above the upper edge of the manubrium jugular notch;

[0078] Substernal notch (xiphoid process): the uncompressed skin surface point below the midline of the sternum;

[0079] Anterior superior iliac spine (left and right): palpate proximally on the midline of the anterior surface of the thigh until the prominence of the anterior superior iliac spine is reached to identify the concave skin surface point;

[0080] Posterior superior iliac spine (left and right): palpate along the iliac margin until the most posterior prominence (close to the sacrum) is found and the concave skin surface point is identified;

[0081] Pubic symphysis: The concave skin surface point at the anterior edge of the pubic symphysis determined by the participant by palpating downward under the midline of the abdomen until the pubic bone is reached. The participant is instructed to run a finger around the inferior edge of the symphysis to find the most anterior point;

[0082] Lateral femoral condyle: measured at the most lateral uncompressed skin surface point on the lateral femoral condyle, on the skin surface or through thin clothing;

[0083] Wrist: An uncompressed skin surface point on the dorsal aspect of the wrist, midway between the radial and ulnar styloid processes;

[0084] Acromion: The uncompressed skin surface point obtained by palpating the most anterior lateral edge of the acromion of the scapula;

[0085] Lateral humeral condyle: the uncompressed skin surface point on the lateralmost side of the humeral condyle;

[0086] Lateral malleolus: the uncompressed skin surface point on the outermost side of the fibular lateral malleolus;

[0087] Inside shoe point: On the inside of the right shoe, near the first metatarsophalangeal joint (approximately on the ball of the foot);

[0088] Heel Contact Point: The point in the center of the right foot's heel's contact area with the floor when the foot is in the normal driving position and in contact with the accelerator pedal.

[0089] The occupant positioning parameters specified in the present invention include size parameters and angle parameters, which are defined in detail as follows:

[0090] L1 is the X-axis dimension of the hip joint relative to the reference point on the sole of the foot, L2 is the X-axis dimension of the hip joint relative to the H point, and L3 is the Z-axis dimension of the hip joint relative to the H point; L4 is the X-axis dimension of the eye relative to the reference point on the sole of the foot, L5 is the Z-axis dimension of the eye relative to the heel point behind the accelerator pedal, L6 is the X-axis dimension of the eye relative to the hip joint, and L7 is the Z-axis dimension of the eye relative to the hip joint; L8 is the X-axis dimension of the ankle relative to the reference point on the sole of the foot; L9 is the X-axis dimension of the ankle relative to the accelerator pedal, and L10 is the Z-axis dimension of the ankle relative to the heel reference point behind the accelerator pedal;

[0091] θ1 is the angle between the hip joint and the eye, θ2 is the knee angle, θ3 is the head angle, θ4 is the neck angle, θ5 is the chest angle, θ6 is the abdomen angle, and θ7 is the pelvic angle.

[0092] S14, integrating the physical parameters of the volunteers, vehicle interior parameters, restraint system parameters and occupant positioning parameters to establish an occupant positioning sample data set. Based on 15 volunteers and 4 interior environments, the occupant positioning sample data set contains 60 data items.

[0093] S2. Use the occupant positioning sample data set to build and verify the occupant positioning parameter prediction model. The process is as follows: Figure 3 As shown, the specific steps include:

[0094] S21. Divide the occupant positioning sample data set into two parts, using 80% of the data as the model building data set and 20% of the data as the verification data set:

[0095] S22. Model construction and selection:

[0096] The occupant positioning parameter prediction model of the occupant is composed of a series of prediction sub-models, which are used to predict different occupant positioning parameters. In the prediction sub-model, the physical sign parameters, vehicle interior parameters and restraint system parameters are used as the input variables of the prediction sub-model, and the corresponding human body positioning model positioning parameters are used as the target output of the prediction sub-model. Different prediction sub-models use different combinations of input variables. Based on the stepwise regression technique, a series of prediction sub-models are automatically constructed to predict the various occupant positioning parameters defined by S13, including L1~L10 and θ1~θ7, and a specific significance level p value is used as the criterion for selecting input variables to enter and leave the model. If the p-value of the variable entering the model is less than 0.25 and the p-value of the variable leaving the model is greater than 0.10, the variable is selected as the input variable of the prediction sub-model to ensure that all items in these models have high significance. Finally, the constructed regression model is manually selected to obtain a concise model.

[0097] Based on the inventors' experience, the input variables used by the prediction sub-model include height, body mass index, ratio of seat height to body height, seat height, distance from the steering wheel to a reference point, and seat cushion angle, which are more likely to be used.

[0098] The final result is a set of regression prediction sub-models with optimal coefficients. For example, to predict the position size L6 of the eye relative to the hip joint, the prediction sub-model uses a linear equation. The relevant parameters of the linear equation include intercept, height coefficient, weight coefficient, sitting height ratio coefficient, etc. The output of the linear equation is the product of the required occupant positioning parameters and the corresponding coefficients, and the various items including the intercept are added together. The calculation formula for L6 is:

[0099]

[0100] Among them, X is the intercept, Stature is the height, K S is the height coefficient, Mass is the weight, K M is the weight coefficient, Sitting height is the sitting height, K SHS is the coefficient of sitting height to body height ratio. Taking the coefficients obtained from a set of experimental results as an example, the intercept X=84.8, the body height coefficient is 0.4659, the sitting height to body height ratio coefficient is -430.1, the seat height coefficient is -0.1732, the steering wheel and foot reference point distance coefficient is 0.4479, and the sitting point angle coefficient is -1.04.

[0101] S23. Evaluate the prediction effect:

[0102] Use the validation data set to evaluate the accuracy of the prediction sub-model. Feed the input variables in the validation data set into the prediction sub-model to obtain the prediction results, and calculate the root mean square error between the prediction results and the actual data of the occupant positioning parameters in the validation data set. If the calculated root mean square error is less than the threshold, it indicates that the model prediction accuracy meets expectations. If the calculated root mean square error is greater than or equal to the threshold, it indicates that the model prediction accuracy needs to be improved. In this case, the number of test samples can be expanded to increase the amount of data in the model establishment data set.

[0103] S3. Positioning of human body finite element model based on occupant positioning parameter prediction model:

[0104] The prediction model obtained in S23 predicts corresponding occupant positioning parameters, and the positioning of the human body finite element model is adjusted according to the predicted occupant positioning parameters.

[0105] The positioning sequence is to first directly predict the front and rear hip positions through the prediction model, then calculate the offset vector from the hip to the H point to determine the hip position, then calculate the position of the eye relative to the hip, and then use the inverse kinematics method to fit the torso and limb parts to the predicted position, and finally complete the positioning of the human finite element model. The specific process is as follows:

[0106] S31. Predicted hip position:

[0107] The prediction model is used to directly predict the front and rear hip positions, and the offset vector from the hip to point H is calculated to obtain the movement path of the hip in the XZ plane. According to the predicted front and rear hip positions, the points on the path are determined as the predicted hip positions. Ensure that the Y coordinates of the left and right hip joints are equidistant from the center line of the seat, so that the torso segment is located in the XZ plane of the center line of the seat.

[0108] S32, calculate the eye position:

[0109] The prediction model is used to calculate the position of the eye relative to the hip. The predicted eye position is the central eye point, whose Z coordinate is equal to the eye corner landmark point and whose X coordinate is equal to the infraorbital landmark point. The human finite element model is adjusted according to the prediction results to ensure that all torso segments are calculated in one plane.

[0110] S33, Fitting the torso and limbs:

[0111] Fit the trunk of the human finite element model: the adjustment sequence is from the pelvic angle to the head angle. First, adjust the trunk above the L5-S1 intervertebral disc to reach the predicted pelvic angle, and remeasure the abdominal angle, chest angle, neck angle, and head angle; then keep the pelvic angle and below unchanged, adjust the abdominal angle (adjust the lumbar spine) to the predicted angle, and recalculate the chest angle, neck angle, and head angle; keep the abdominal angle and below unchanged, adjust the thoracic spine to the predicted chest angle, and recalculate the neck angle and head angle; keep the chest angle and below unchanged, adjust the cervical spine to the predicted neck angle, and recalculate the head angle, adjust the head to the predicted head angle, and complete the trunk positioning.

[0112] Fit the limbs of the human finite element model: for the upper limbs, place the driver's hands on the steering wheel at 2 o'clock and 10 o'clock; for the lower limbs, place them according to the relative positions of the ankle and hip joints, the coordinate origin, and the predicted values ​​of the knee angles calculated; other unspecified positions can be consistent with the real human posture.

[0113] S4. Check and correct the positioning results of the human finite element model:

[0114] Since there are certain differences between the human body finite element model and the real human body in terms of seating conditions, the values ​​of the occupant positioning parameters calculated using the occupant positioning parameter prediction model may be unreasonable. Therefore, it is necessary to check and correct the occupant positioning parameters. The specific steps are as follows:

[0115] S41. Check whether the head angle (whether the Frankfurt plane position is reasonable) and the angle of the limbs are significantly different from the normal sitting posture of the human body. At this time, corrections are made based on the historical database of the test results of the actual positioning parameters of the human body.

[0116] S42, modify the positioning parameters:

[0117] If an unreasonable head angle is detected, for example, the angle of the eyes is not aligned with the front of the vehicle, priority is given to ensuring that the head position maintains a reasonable Frankfurt plane position, and then adjusting the head angle to meet the regression model calculation results.

[0118] If an abnormal angle between the limbs is detected, the interior position such as the front and rear position of the seat, the steering wheel angle, etc. can be adjusted, and other positioning parameters can be recalculated.

[0119] The above are only embodiments of the present invention. The invention is not limited to the field involved in this implementation case. The common knowledge such as the known specific structure and characteristics in the scheme is not described in detail here. It should be pointed out that for those skilled in the art, several deformations and improvements can be made without departing from the structure of the present invention, which should also be regarded as the protection scope of the present invention, and these will not affect the effect of the implementation of the present invention and the practicality of the patent. The scope of protection required by this application shall be based on the content of its claims, and the specific implementation methods and other records in the specification can be used to interpret the content of the claims.

Claims

1. A diversified vital sign occupant positioning method for automobile safety virtual assessment, characterized by: The steps include: S1. Constructing a sample data set of occupant positioning with diversified physical signs, wherein the sample data set of occupant positioning includes physical sign parameters, vehicle interior parameters and restraint system parameters; S2, using the occupant positioning sample data set to build and verify the occupant positioning parameter prediction model; S3, positioning based on the occupant positioning parameter prediction model; S4. Check and correct the positioning results of the human body finite element model.

2. The method for locating occupants with diversified vital signs for automobile safety virtual assessment according to claim 1, characterized in that: The physical sign parameters include gender, weight, height, and weight coefficient; the vehicle interior parameters include seat height, distance between the steering wheel and the sole reference point, and seat cushion angle; the restraint system parameters include seat reference point, steering wheel center point, steering wheel diameter, steering wheel angle, seat track angle, and accelerator pedal center point.

3. The diversified vital signs occupant positioning method for automobile safety virtual assessment according to claim 1 is characterized by: The S1 includes: S11, obtaining the vital sign parameters of the occupant; S12, obtaining vehicle interior parameters and restraint system parameters; S13, measuring occupant positioning parameters in different vehicle interior environments; S14. Integrate the volunteers' physical parameters, vehicle interior parameters, restraint system parameters and occupant positioning parameters to establish an occupant positioning sample data set.

4. The diversified vital signs occupant positioning method for automobile safety virtual assessment according to claim 3 is characterized by: The occupant positioning parameters include size parameters and angle parameters, wherein the size parameters include an X-axis size of the hip joint relative to a reference point on the sole of the foot, an X-axis size of the hip joint relative to the H point, and a Z-axis size of the hip joint relative to the H point; an X-axis size of the eye relative to a reference point on the sole of the foot, a Z-axis size of the eye relative to a heel point behind an accelerator pedal, an X-axis size of the eye relative to the hip joint, and a Z-axis size of the eye relative to the hip joint; an X-axis size of the ankle relative to a reference point on the sole of the foot; an X-axis size of the ankle relative to an accelerator pedal, and a Z-axis size of the ankle relative to a heel reference point behind an accelerator pedal; The angle parameters include the angle between the hip joint and the eye, the knee angle, the head angle, the neck angle, the chest angle, the abdomen angle, and the pelvic angle.

5. The method for locating occupants with diversified vital signs for automobile safety virtual assessment according to claim 4 is characterized by: The S2 includes: S21, dividing the occupant positioning sample data set into two parts, using 80% of the data as a model building data set and 20% of the data as a verification data set; S22, constructing an occupant positioning parameter prediction model based on stepwise regression technology, wherein the occupant positioning parameter prediction model is a set of multiple prediction sub-models, wherein the multiple prediction sub-models are respectively used to predict each occupant positioning parameter defined in S13, and a specific significance level is used as a criterion for selecting input variables, and finally the optimal prediction sub-model is selected; S23. Evaluate the prediction effect.

6. The method for locating occupants with diversified vital signs for virtual vehicle safety assessment according to claim 5, characterized in that: The prediction sub-model is a regression prediction equation with optimal coefficients, the input variables of the regression prediction equation are selected physical parameters, vehicle interior coefficients, and restraint system parameters, and the output of the regression prediction equation is the occupant positioning parameter.

7. The method for locating occupants with diversified vital signs for automobile safety virtual assessment according to claim 6, characterized in that: The S3 includes: S31, predicting the hip position of the occupant using the prediction sub-model; S32, predicting the position of the eye relative to the hip using the prediction sub-model; S33, Fitting the torso and limbs.

8. The method for locating occupants with diversified vital signs for automobile safety virtual assessment according to claim 7 is characterized by: In the step S31, the prediction model is used to predict the front and rear hip positions, and the offset vector from the hip to the H point is calculated to obtain the movement path of the hip in the XZ plane. According to the predicted front and rear hip positions, the points on the path are determined as the predicted hip positions, and the Y coordinates of the left and right hip joints are equidistant from the center line of the seat, so that the torso segment is located in the XZ plane of the center line of the seat; In the step S32, the prediction model is used to calculate the position of the eye relative to the hip, the Z coordinate of the predicted eye position is equal to the eye corner mark point, and the X coordinate is equal to the infraorbital mark point; In the step S33, the trunk above the L5-S1 intervertebral disc is first adjusted to reach the predicted pelvic angle, and the abdominal angle, chest angle, neck angle, and head angle are remeasured; subsequently, the pelvic angle and the angle below are kept unchanged, the abdominal angle is adjusted (the lumbar angle is adjusted) to the predicted angle, and the chest angle, neck angle, and head angle are recalculated; the abdominal angle and the angle below are kept unchanged, the thoracic vertebrae are adjusted to the predicted chest angle, and the neck angle and the head angle are recalculated; the chest angle and the angle below are kept unchanged, the cervical vertebrae are adjusted to the predicted neck angle, and the head angle is recalculated, and the head is adjusted to the predicted head angle to complete the positioning of the trunk; The driver's hands are placed at 2 o'clock and 10 o'clock on the steering wheel, and the lower limbs are placed according to the relative positions of the ankle and hip joints, the coordinate origin, and the calculated predicted values ​​of the knee angle. Other positions are judged whether they conform to the real human posture.

9. The diversified vital signs occupant positioning method for automobile safety virtual assessment according to claim 8 is characterized by: The S4 includes: S41. Check the angles of the head and limbs; S42. Adjust the fitting result according to the inspection result.

Images