Study On Trigger Decision Technology For Vehicle Accident Automatic Call System

Vehicle accident automatic call system is an effective post-crash safety technology.It can reduce the accident response time and improve the rescue efficiency effectively,so it is of great significance to alleviate the severity of injuries or save the life for accident victims.After the vehicle accident,vehicle accident automatic call system not only needs to detect the accident,but also needs to identify the types of accident and evaluate the injury condition of vehicle occupant,so that more information about the accident is able to be provided for the call center to help accident rescuer distribute relief resources reasonably and medical staff make better healthcare decisions.This requires vehicle accident automatic call system to make various decisions when triggering to reflect the real situation of accident,so the research about key decision technologies when vehicle accident automatic call system triggers was carried out in this article.The involved key decision technologies when vehicle accident automatic call system triggers were introduced,including vehicle accident detection technology,vehicle body damage evaluation technology after collision and occupant injury evaluation technology after collision.Also,evaluation indexes and implement methods about these technologies were given.The collision trigger algorithm which is composed of collision detection algorithm and collision direction recognition algorithm was designed for vehicle accident automatic call system.Collision detection algorithm deals with the longitudinal and transverse acceleration signals of vehicle body based on moving window principle and this algorithm compares the computational result with the preset threshold to determine whether collision accident happened.Collision direction recognition algorithm identifies the collision direction according to principal direction of force.Three-axis acceleration sensor,NI data acquisition card and laptop with NI LabVIEW software were used to build a set of data acquisition system.The threshold of collision detection algorithm was determined by using this data acquisition system to collect vehicle body acceleration signals of a sport utility vehicle in various working conditions.To verify the reliability of the designed collision trigger algorithm,trolley collision tests were conducted combining with the vehicle accident automatic call system terminal hardware designed by our research team.The test results showed that the collision detection algorithm was able to detect the collision accident rapidly and collision direction recognition algorithm could recognize the collision direction accurately.428 vehicle collision accident cases from National Highway Traffic Safety Administration Special Crash Investigation database were collected.The main factors influencing driver's injury reaching to MAIS3+ were analyzed based on Logistic regression model and the analysis result turned out that change in vehicle velocity,collision direction and use of safety belt were the main factors influencing driver's injury reaching to MAIS3+.Then the Logistic regression model between driver's injury risk and the main factors was constructed and this model could be used for vehicle accident automatic call system to predict driver's injury condition.