| With the development of autonomous driving technology,drivers can be liberated from complex driving tasks to a certain extent,and their riding posture will exhibit diversity.Studies have shown that future autonomous vehicles are expected to have a large seat recline angle to relax.Posture is an important factor affecting the effect of restraint system on occupant protection,traditional restraint systems may no longer be suitable for tilting posture.Therefore,the occupant protection issues brought about by reclining seats have important significance.Based on the above background,the main research purpose is to analyze the influence of driver posture on collision injury and the influence of driver’s knee angle on the upper torso through simulation.Research on driver torso angle recognition method based on depth camera.Explore active passive integrated protection methods based on active seat adjustment strategies.Firstly,develop driver injury research under different postures.Based on NHTSA’s Yaris,a frontal 100% overlap rigid barrier collision restraint system model was constructed,and the effectiveness of the model was verified through benchmarking analysis with experimental results.The driver simulation model in different postures was established,and the driver motion response and injury of different postures were compared and analyzed.The results showed that with the increase of the torso angle of the dummy,the head,neck and leg injuries increased significantly.Secondly,explore the influence of different knee angles on the upper torso of the driver.The simulation models of different knee joint angles for drivers with two seat backs were constructed.The effects of different knee angles on the torso were studied from two aspects: motor response and injury.The results showed that changes in knee angle had less effect on the torso.Thirdly,the driver’s torso angle is recognized based on a depth camera.A depth camera combined with OpenPose algorithm is used to acquire key points on a threedimensional surface.Based on the three-dimensional surface key points,the shoulder and hip point cloud contours were extracted.And the mathematical analysis method was used to offset the three-dimensional surface key points to the human body.Finally,the threedimensional coordinates were used to identify the torso angle.Subsequently,a volunteer experiment was conducted,and the results showed that the recognition error was within 5°.Finally,explore an integrated protection method based on active seat adjustment.Through injury threshold analysis,an active seat adjustment strategy is formulated.Integrated simulation of active seat adjustment and collision.According to the analysis results,the joint optimization platform built by modeFRONTIER and MADYMO software was used to optimize the active seat adjustment parameters.The optimization results show that the WIC value decreases.A seat adjustment model was built in the pre-crash phase to explore the influence of the angular velocity of seat rotation on the head.In summary,excessive torso angle of the driver can seriously affect the protective effect of the restraint system,while increasing the risk of neck shoulder belt interaction.Taking into account the danger of tilting posture and the impact of different knee joint angles on the upper torso,the recognition algorithm can accurately identify the torso angle with the driver’s torso angle as the recognition target.For tilting posture,an integrated protection method based on active seat adjustment can adjust the driver’s posture during the pre-crash phase to reduce injury. |
