Research On The Injury Prediction Algorithm And The Terminal Design Of Advanced Automatic Crash Notification System

Advanced Automatic Crash Notification(AACN)system is one of the postaccident safety technologies.AACN system can detect whether the collision accident occurs and record the information of the collision accident through the on-board data instrument,and then predict the injury of the occupants in the vehicle according to the occupants' injury prediction algorithm.If the injury of occupants is serious,AACN system will call for help and send the information of accident to the medical emergency.It can improve the rescue efficiency and save more seriously injured occupants' lives.Automatic Crash Notification(ACN)system can only judge whether the collision accident occurs,it can't predict the injury of occupant,so it is lack of practicability.Compared with ACN system,AACN system uses occupant injury as the trigger signal,so it has important theoretical and practical significance for the research of occupants' injury prediction algorithm and AACN system terminal.In order to establish the prediction algorithm of driver injury,the accident data of the U.S.highway traffic safety administration were counted.According to the accident data,the factors that affect the driver's injury were determined.Then,the prediction model of driver's injury was established based on Naive Bayes and Logistic regression.The sensitivity of the prediction model was analyzed and the best threshold of the model was determined.Finally,according to the prediction model,the algorithm of driver injury prediction was established.In order to establish the prediction algorithm of rear occupant injury in frontal impact,a 100% overlap rigid barrier frontal impact test was carried out at 50km/h,the impact acceleration data and dummy injury index were collected,and the simulation model of rear occupant restraint system was established based on LS-DYNA software,then the simulation model was loaded with the impact acceleration data,and the dummy of simulation and test were compared injury index to verify the reliability of the simulation model.In order to obtain more crash acceleration data,the front crash acceleration data under different initial speeds were generated according to the vehicle simulation model,and then the acceleration data were loaded into the rear occupant restraint system simulation finite element model to get the rear occupant injury values under different speed changes,and the rear occupant injury prediction model was obtained through regression analysis.Finally,according to the prediction model,the prediction algorithm of rear occupant injury in frontal collision was established.In order to integrate the injury prediction algorithm of drivers and rear passengers into the terminal of AACN system,the overall algorithm of AACN system was designed.Then the hardware and software of AACN system terminal were designed.Finally,the simulation acceleration data was used to verify the AACN system terminal.The verification results showed that the designed terminal could accurately identify collision and predict occupant injury,and send accident information to the outside.