| China is one of the countries that suffered the largest number of road traffic accidents in the world. Analyzing the causes of road traffic accident and human body damage is significant for making a reasonable plan of preventing and treating accident. Meanwhile, the study of injury biomechanics is often limited by lots of conditions for the experimental cost and growing public ethical awareness. So compared with post mortem human surrogate, traffic accident reconstruction is a more ideal way for the study of injury biomechanics. To solve the problems that existing in traffic accident reconstruction, such as boundary missing and measurement error, this paper starts with the technology of getting the evidence from road traffic accident, uses UAV aerial technology to achieve the orthographic map of accident scene, and analyzes EDR data from vehicle. What’s more, through reconstructing typical accident cases, this study explores the effect of vehicles, effect mechanism of vehicles and human body to provide some references for the responsibility affirmation, the human body injury protection, traffic injury treatment and the design of automotive safety, which is significant for economy and society.The main studying methods and results are as follows:(1) To solve the problems that existing in the investigation of accident scene such as time-consuming, low precision and difficultly to verify again, we first proposed a solution at home which is using UAV aerial to collect the evidence of accident scene. First, this research assembled a four-rotor aircraft as the aerial shoot platform, and studied the key problems such as camera calibration and image geometric rectification. Meanwhile, experiments evaluated the precision of image geometric rectification and measurement precision of aerial image; Second, to draw the traffic accident spot diagram in real-time, a scene image drawing system was made which is based on aerial image and be compared with traditional survey method; Finally, we tentatively do the research on the aerial image mosaic and 3D modeling that is based on slope photography. The results indicate that the RMSE of the correction orthographic map is 3.6 pixels; in the photography height range between 11 m and 112 m, the relative error range can be controlled between-2.5% and 2.5%; the scene image drawing system that is based on aerial image is better than traditional survey method on the drawing efficiency and data accuracy. So it’s an important method for the investigation of accident scene aerial image to use aerial image to fix and measure evidence.(2) It’s difficult to get lots of important data such as collision speed and ?V for traditional deep investigation. To solve this problem, an analytical tool of EDR data was first used to analyze the EDR data of Toyota’s cars in China. To compare the speed before collision and reference speed from video or speedometer of EDR data through analyzing 11 different cases of Toyota’s traffic accident, and the results are as follows: mean of different speeds is-0.42km/h; the standard deviation is 3.85km/h; the mean of relative error is-1.1%; the standard deviation of relative error is 6%. At the same time, this research analyzed some parameters about 76 cases of Toyota’s traffic accidents, like collision speed, situation of drivers’ wearing seat belts, longitudinal ?V, braking time, etc. The results of statistical analysis are as follows: the driver’s average braking time is 1.38 seconds before the collision occurs; there is significant correlation of driver’s MAIS with vehicle longitudinal ?V; though seat belt can reduce injury level for drivers, the proportion of wearing seat belt correctly is only 51.3% in the sample. So it indicate that EDR data can provide a reliable collision speed for reconstruction, at same time, the seat belt use and longitudinal ?V are vital parameters for injury process reconstruction.(3) For aerial image and EDR data play an important role in traffic accident reconstruction, this study will introduce that through several cases’ accident scenarios and process reconstruction. The results are as follows: It is intuitive to analyze the relationship among factors when aerial orthographic map was applied to reconstruct the accident scenarios digitally, which can be helpful to reconstruct the accident progress; PC-CRASH and ARAS360 can be used to reconstruct the accident collision point, accident type and driver’s braking action while aerial orthographic map and data from deep investigation were synthesized; Using EDR data as the initial condition and combining it with data from deep investigation, dynamic response of vehicle and pedestrians can be reconstructed accurately with MADYMO and LS-DYNA, and the results conform with the information of video.(4) This study extracted some data from EDR data of 8 cases as the input criterion of driver restraint system boundary, like trigger time of airbag and pretensioner, collision speed, situation of drivers’ wearing seat belts and longitudinal ?V, and reconstructed the injury process of driver to get the driver’s kinematics response. From the response of the multi-rigid-body dummy head and chest to calculate HIC36, HIC15 of head and VC, CTI of chest, according to the curve of injury risk and tolerance of head and chest, to estimate the injury of driver’s head and chest. The results are as follows: in head-on collision accident, it’s easy to get driver’s kinematics response by using EDR data as the initial and boundary condition to reconstruct the injury process of driver, and that was matched up with the results of deep investigation; The HIC36, HIC15 of dummy head and VC, CTI of dummy chest from injury process reconstruction of driver can be used to estimate injury level of driver’s head and chest, and that was consistent with the facts of driver’s injury situation. These results will lay the foundation for classification of medical treatment that is quick and matched injury level through EDR data to evaluate driver’s injury after the accident.
|
PDF Full Text Request