Head adaptive protection method considering pedestrian biological characteristics in emergency state

Abstract

This invention discloses an adaptive head protection method considering pedestrian biometrics in emergency situations. Based on a finite element model of the pedestrian's emergency posture and the current vehicle speed, a head position prediction model is used to obtain different vehicle states at the time of collision. The collision position of the pedestrian's head under different vehicle states is determined. Based on the predicted intracranial pressure, the optimal vehicle state is selected, corresponding to the lowest intracranial pressure. The front bumper beam is adjusted according to the position of the front bumper beam corresponding to the optimal vehicle state. The brake pedal is controlled according to the vehicle speed corresponding to the optimal vehicle state. The steering wheel is controlled according to the vehicle angle corresponding to the optimal vehicle state to obtain the corresponding vehicle angle. This invention can minimize damage when a collision is unavoidable, thereby effectively protecting the pedestrian's head. It can automatically obtain the corresponding head injury situation based on different pedestrian vital signs and emergency postures, and adaptively adjust the vehicle state.

Description

Technical Field

[0001] This invention relates to the field of automotive safety technology, specifically to an adaptive head protection method that takes into account pedestrian biometrics in emergency situations. Background Technology

[0002] When a vehicle collides with a pedestrian, head injuries are often the most fatal. Therefore, effectively protecting pedestrians' heads during a collision is crucial for improving traffic safety.

[0003] Current pedestrian protection during collisions primarily focuses on actively controlling vehicle paths to avoid pedestrians. However, in most cases, collisions are unavoidable due to driver error or sensor system malfunction. Therefore, a system for mitigating human injury when a collision is unavoidable is currently lacking. Summary of the Invention

[0004] The purpose of this invention is to provide an adaptive head protection method that takes into account the biometrics of pedestrians in emergency situations, which can minimize damage and effectively protect the pedestrian's head when a collision is unavoidable.

[0005] According to one objective of the present invention, the present invention provides a head adaptive protection method that takes into account pedestrian biometrics in an emergency, comprising the following steps:

[0006] Step 1: The vehicle uses lidar to determine the safe distance to the pedestrian ahead. If the distance is less than the safe distance, the warning system activates a safety warning signal. The safe distance is the operating distance that the driver can control the vehicle to avoid the pedestrian through normal reaction.

[0007] Step 2: Input the safety warning signal into the vehicle's external camera, and the external camera will take a picture of the pedestrian in front of the vehicle that cannot be avoided;

[0008] Step 3: Transmit the pedestrian photo to the human body surface reconstruction model to obtain the human body surface point cloud. The human body surface reconstruction model is the widely used HMR or SMPLFLY model, which is used to extract feature information from the pedestrian photo and recover the corresponding human body surface point cloud.

[0009] Step 4: Input the acquired human body surface point cloud into the pedestrian emergency posture finite element model library. By minimizing the distance between the human body surface point cloud and the finite element model mesh nodes in the pedestrian emergency posture finite element model library, find the closest pedestrian emergency posture finite element model.

[0010] Step 5: Based on the found pedestrian emergency posture finite element model and the current vehicle speed, obtain the different vehicle states at the time of collision through the head position prediction model, including the pedestrian head collision position under the front bumper beam position, vehicle speed, and vehicle angle.

[0011] Step 6: Based on the deployment range of the pedestrian head protection airbag, determine the pedestrian head collision position under different vehicle states obtained in Step 5. If the head collision position is within the airbag deployment range, the corresponding vehicle state is the preferred state; if it is outside the airbag deployment range, the corresponding vehicle state is the dangerous state.

[0012] Step 7: Input the vehicle's preferred state into the intracranial pressure prediction model to obtain the pedestrian's intracranial pressure corresponding to different vehicle preferred states.

[0013] Step 8: Based on the predicted intracranial pressure, select the vehicle state corresponding to the minimum intracranial pressure as the optimal vehicle state.

[0014] Step 9: Adjust the front bumper beam according to the position of the front bumper beam corresponding to the optimal vehicle state. The front bumper beam is connected to the vehicle body through a locking structure and a universal joint controller. Under normal conditions, the locking mechanism is closed. When a collision occurs, the locking mechanism opens and the position of the front bumper beam is controlled by the universal joint controller.

[0015] Step 10: Control the brake pedal according to the vehicle speed corresponding to the optimal vehicle state to obtain the corresponding vehicle speed;

[0016] Step 11: Control the steering wheel according to the vehicle angle corresponding to the optimal vehicle state to obtain the corresponding vehicle angle.

[0017] Furthermore, in step 4, the pedestrian emergency posture finite element model library is established according to the following steps:

[0018] S401 is recruiting volunteers of different ages, genders, heights, and BMIs.

[0019] S402, volunteers pose in emergency positions in the event of a vehicle collision;

[0020] S403, obtains surface point clouds of different volunteers in different emergency postures through surface scanning;

[0021] S404 uses existing mature commercial human finite element models, such as THUMS or GHBMC, as the benchmark model, and establishes finite element models of pedestrian emergency postures corresponding to different body surface point clouds through mesh deformation technology.

[0022] S405, collects and constructs a pedestrian emergency posture finite element model library by summarizing different pedestrian emergency posture finite element models.

[0023] Further, in step 5, the head position prediction model is established according to the following steps:

[0024] S501, construct different DOE simulation matrices, with matrix influence factors representing human models in different emergency postures, front bumper beam position, vehicle speed, and vehicle angle.

[0025] S502 uses finite element simulation to obtain the head collision position under different combinations of matrix influence factors;

[0026] S503 uses a deep neural network model as its base model, matrix influence factors as input, and head collision location as output. It obtains a head position prediction model through iterative training.

[0027] Furthermore, in step 6, the pedestrian head protection airbag is located at the junction of the vehicle's windshield and hood, and is elongated in shape, automatically deploying upon impact.

[0028] Further, in step 7, the intracranial pressure prediction model is established according to the following steps:

[0029] S701, construct different DOE simulation matrices, with matrix influence factors representing human models in different emergency postures, front bumper beam position, vehicle speed, and vehicle angle.

[0030] S702, through finite element simulation, obtains the head collision position under different matrix influence factor combinations;

[0031] S703 uses a deep neural network model as its base model, matrix influence factors as input, and intracranial pressure as output. It obtains an intracranial pressure prediction model through iterative training.

[0032] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0033] This invention provides an adaptive head protection method that takes into account pedestrian biometrics in emergency situations. It is adaptable to different collision conditions, has strong robustness, and can minimize damage when a collision is unavoidable, thereby effectively protecting the pedestrian's head. It can automatically obtain the corresponding head injury information based on the pedestrian's different physical characteristics and emergency posture, and adaptively adjust the vehicle status. Attached Figure Description

[0034] Figure 1 This is a schematic diagram of the casting partitions according to an embodiment of the present invention. Detailed Implementation

[0035] 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, and 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.

[0036] All components of this invention are general standard parts or parts known to those skilled in the art, and their structures and principles can be learned by those skilled in the art through technical manuals or conventional experimental methods.

[0037] Example 1

[0038] like Figure 1 As shown:

[0039] An adaptive head protection method that takes into account pedestrian biometrics in emergency situations includes the following steps:

[0040] Step 1: The vehicle uses lidar to determine the safe distance to pedestrians ahead. If the distance is less than the safe distance, the warning system activates a safety warning signal. The safe distance is the operating distance that the driver can control the vehicle to avoid the pedestrian with normal reaction.

[0041] Step 2: Input the safety warning signal into the vehicle's external camera, and the external camera will take a picture of the pedestrian in front of the vehicle that cannot be avoided;

[0042] Step 3: Transmit the pedestrian photo to the human body surface reconstruction model to obtain the human body surface point cloud. The human body surface reconstruction model is a widely used model such as HMR and SMPLFLY, which can extract feature information from the pedestrian photo and recover the corresponding human body surface point cloud.

[0043] Step 4: Input the acquired human body surface point cloud into the pedestrian emergency posture finite element model library. By minimizing the distance between the human body surface point cloud and the finite element model mesh nodes in the pedestrian emergency posture finite element model library, find the closest pedestrian emergency posture finite element model.

[0044] In this step, the pedestrian emergency posture finite element model library is established according to the following steps:

[0045] S401 is recruiting several volunteers with different physical characteristics (age, gender, height, BMI).

[0046] S402, volunteers pose in emergency positions in the event of a vehicle collision;

[0047] S403, obtains surface point clouds of different volunteers in different emergency postures through surface scanning;

[0048] S404 uses existing mature commercial human finite element models, such as THUMS and GHBMC, as the benchmark model, and establishes finite element models of pedestrian emergency postures corresponding to different body surface point clouds through mesh deformation technology.

[0049] S405, collect and construct a pedestrian emergency posture finite element model library by summarizing different pedestrian emergency posture finite element models;

[0050] Step 5: Based on the found pedestrian emergency posture finite element model and the current vehicle speed, obtain the pedestrian head collision position under different vehicle states (front bumper beam position, vehicle speed, vehicle angle) during the collision through the head position prediction model.

[0051] In this step, the head position prediction model is established according to the following steps:

[0052] S501, construct different DOE simulation matrices, with matrix influence factors representing human models in different emergency postures, front bumper beam position, vehicle speed, and vehicle angle.

[0053] S502 uses finite element simulation to obtain the head collision position under different combinations of matrix influence factors;

[0054] S503 uses a deep neural network model as the base model, matrix influence factors as input, and head collision position as output. It obtains a head position prediction model through iterative training.

[0055] Step 6: Based on the deployment range of the pedestrian head protection airbag, determine the pedestrian head collision position under different vehicle states obtained in Step 5. If the head collision position is within the airbag deployment range, the corresponding vehicle state is the preferred state; if it is outside the airbag deployment range, the corresponding vehicle state is the dangerous state.

[0056] In this step, the pedestrian head protection airbag is located at the junction of the vehicle's windshield and hood. It is long and narrow and automatically deploys upon impact.

[0057] Step 7: Input the vehicle's preferred state into the intracranial pressure prediction model to obtain the pedestrian's intracranial pressure corresponding to different vehicle preferred states.

[0058] In this step, the intracranial pressure prediction model is established according to the following steps:

[0059] S701, construct different DOE simulation matrices, with matrix influence factors representing human models in different emergency postures, front bumper beam position, vehicle speed, and vehicle angle.

[0060] S702, through finite element simulation, obtains the head collision position under different matrix influence factor combinations;

[0061] S703 uses a deep neural network model as the base model, matrix influence factors as input, and intracranial pressure as output. It obtains an intracranial pressure prediction model through iterative training.

[0062] Step 8: Based on the predicted intracranial pressure, select the vehicle state corresponding to the minimum intracranial pressure as the optimal vehicle state.

[0063] Step 9: Adjust the front bumper beam according to the position of the front bumper beam corresponding to the optimal vehicle state. The front bumper beam is connected to the vehicle body through a locking structure and a universal joint controller. Under normal conditions, the locking mechanism is closed. When a collision occurs, the locking mechanism opens and the position of the front bumper beam is controlled by the universal joint controller.

[0064] Step 10: Control the brake pedal according to the vehicle speed corresponding to the optimal vehicle state to obtain the corresponding vehicle speed;

[0065] Step 11: Control the steering wheel according to the vehicle angle corresponding to the optimal vehicle state to obtain the corresponding vehicle angle.

[0066] The head adaptive protection system proposed in this invention has a simple structure, is easy to control, adapts to different collision conditions, and has strong robustness. It can minimize damage when a collision is unavoidable, thereby effectively protecting the pedestrian's head. It can automatically acquire the corresponding head injury information based on the pedestrian's different physical characteristics and emergency posture, and adaptively adjust the vehicle's status accordingly.

[0067] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A head adaptive protection method considering pedestrian biometrics in emergency situations, characterized in that, Includes the following steps: Step 1: The vehicle uses lidar to determine the safe distance to the pedestrian ahead. If the distance is less than the safe distance, the warning system activates a safety warning signal. The safe distance is the operating distance that the driver can control the vehicle to avoid the pedestrian through normal reaction. Step 2: Input the safety warning signal into the vehicle's external camera, and the external camera will take a picture of the pedestrian in front of the vehicle that cannot be avoided; Step 3: Transmit the pedestrian photo to the human body surface reconstruction model to obtain the human body surface point cloud. The human body surface reconstruction model is the widely used HMR or SMPLFLY model, which is used to extract feature information from the pedestrian photo and recover the corresponding human body surface point cloud. Step 4: Input the acquired human body surface point cloud into the pedestrian emergency posture finite element model library. By minimizing the distance between the human body surface point cloud and the finite element model mesh nodes in the pedestrian emergency posture finite element model library, find the closest pedestrian emergency posture finite element model. Step 5: Based on the found pedestrian emergency posture finite element model and the current vehicle speed, obtain the different vehicle states at the time of collision through the head position prediction model, including the pedestrian head collision position under the front bumper beam position, vehicle speed, and vehicle angle. Step 6: Based on the deployment range of the pedestrian head protection airbag, determine the pedestrian head collision position under different vehicle states obtained in Step 5. If the head collision position is within the airbag deployment range, the corresponding vehicle state is the preferred state; if it is outside the airbag deployment range, the corresponding vehicle state is the dangerous state. Step 7: Input the vehicle's preferred state into the intracranial pressure prediction model to obtain the pedestrian's intracranial pressure corresponding to different vehicle preferred states. Step 8: Based on the predicted intracranial pressure, select the vehicle state corresponding to the minimum intracranial pressure as the optimal vehicle state. Step 9: Adjust the front bumper beam according to the position of the front bumper beam corresponding to the optimal vehicle state. The front bumper beam is connected to the vehicle body through a locking structure and a universal joint controller. Under normal conditions, the locking mechanism is closed. When a collision occurs, the locking mechanism opens and the position of the front bumper beam is controlled by the universal joint controller. Step 10: Control the brake pedal according to the vehicle speed corresponding to the optimal vehicle state to obtain the corresponding vehicle speed; Step 11: Control the steering wheel according to the vehicle angle corresponding to the optimal vehicle state to obtain the corresponding vehicle angle.

2. In the head adaptive protection method considering pedestrian biometrics in an emergency situation according to claim 1, in step 4, the pedestrian emergency posture finite element model library is established according to the following steps: S401 is recruiting volunteers of different ages, genders, heights, and BMIs. S402, volunteers pose in emergency positions in the event of a vehicle collision; S403, obtains surface point clouds of different volunteers in different emergency postures through surface scanning; S404 uses the existing mature commercial human finite element model as the benchmark model and establishes finite element models of pedestrian emergency postures corresponding to different body surface point clouds through mesh deformation technology. S405, collects and constructs a pedestrian emergency posture finite element model library by summarizing different pedestrian emergency posture finite element models.

3. In the head adaptive protection method considering pedestrian biometrics in an emergency situation according to claim 1, in step 5, the head position prediction model is established according to the following steps: S501, construct different DOE simulation matrices, with matrix influence factors representing human models in different emergency postures, front bumper beam position, vehicle speed, and vehicle angle. S502 uses finite element simulation to obtain the head collision position under different combinations of matrix influence factors; S503 uses a deep neural network model as its base model, matrix influence factors as input, and head collision location as output. It obtains a head position prediction model through iterative training.

4. In the head adaptive protection method for considering pedestrian biometrics in an emergency state as described in claim 1, in step 6, the pedestrian head protection airbag is located at the junction of the vehicle's windshield and engine hood, is elongated, and automatically deploys upon impact.

5. The head adaptive protection method considering pedestrian biometrics in an emergency situation according to claim 1, in step 7, the intracranial pressure prediction model is established according to the following steps: S701, construct different DOE simulation matrices, with matrix influence factors representing human models in different emergency postures, front bumper beam position, vehicle speed, and vehicle angle. S702, through finite element simulation, obtains the head collision position under different matrix influence factor combinations; S703 uses a deep neural network model as its base model, matrix influence factors as input, and intracranial pressure as output. It obtains an intracranial pressure prediction model through iterative training.

Images