Research On Vehicle-to-pedestrian Accident Scenarios And Optimizing Of AEB Based On NAIS

Pedestrian is vulnerable group in road traffic accident,and the probability of serious injury or death is higher than other kind of traffic accidents.Automatic emergency braking?AEB?system can effectively reduce or avoid pedestrian injury.To evaluate the effectiveness of AEB system,many countries issued the new car assessment programme,but the test scenarios form foreign countries cannot reflect the features of road traffic in China,so the research based on China's road traffic accident scenarios has its necessity.Not only provide the reference to the pedestrian AEB test scenarios constraction in our country,but also provide a reference for pedestrian AEB system control strategy design.The research in this paper mainly includes the following four aspects:?1?228 cases of head-on collision accidents between passenger cars and pedestrians were screened out from the National Automobile Accident in-depth Investigation System?NAIS?database,and the static and dynamic variables related to the reconstruction of dangerous scenarios in each case were extracted.The Accident cases with similar characteristics were gathered into a group through cluster analysis.Then the chi-square test and the degree of pedestrian injury were used as the criteria to extract the typical features in each kind of scenario.Finally,five typical critical scenarios of pedestrian and passenger car collision were formed.?2?On the basis of typical critical scenarios,variables were specified and"obstacle"parameter was added to each type of scenario to form 10 types of test scenarios.PreScan software was used to model the test scenario.?3?Comparing and analyzing AEB collision avoidance control strategy,this paper focused on TTC model.In Simulink,AEB control module based on TTC model was built and added to the test vehicle.?4?Jointing PreScan test,the performance of collision avoidance algorithm was tested in different scenarios.The results showed that the vehicle equipped with AEB model could not avoid collision in all test scenarios.Its performance was poor,especially when the speed was high.In order to improve the collision avoidance performance,the algorithm was optimized.A pedestrian crossing road model was built for a single scenario,and the collision avoidance was realized through matching and optimization analysis of the AEB system's field of view?FoV?,brake trigger width?w?and AEB trigger time(TTC_max).In this process,boundary parameter conditions of AEB system to avoid collision or greatly reduce collision speed were obtained,and the mutual constraint relationship between w and TTC_max was obtained.In conclusion,this paper obtained typical pedestrian danger scenarios based on traffic accidents in China,established corresponding simulation test scenarios,and tested the collision avoidance performance of TTC model in various scenarios.The method applied in this paper provides a reference for the extraction of typical critical scenarios in China's roads and the development of AEB system test scenarios adapted to China's road traffic scenarios.Meanwhile,a technical scheme for the optimization of AEB key parameters is provide.