The Research Of Vehicle Dangerous Working Conditions Warning Technology Based On Multi Information Fusion

With the rapid development of highway construction in our country and the increase of vehicles number, the number of road traffic accidents incressing in a high speed too. Because of the complex road conditions in our country, traffic accidents often happened unexpected or suddenly. Because it’s too late for drivers to react or drivers don’t pay enough concentration when they are driving, they can’t predict accidents risk timely thus fail to do the effective remedial measures, lead to the traffic accidents. As a result, it’s very important to give warnings to the driver before vehicle’s entering dangerous working conditions. Nowadays, warning system has become one of research hotspot of scholars both at home and abroad.Building reasonable model is an important part of vehicle warning algorithm. This paper choose rollover and collision as dangerous working conditions of vehicle, the 3-DOF rollover dynamic model and vehicle position model are built respectively and set up the simulation model in Matlab/Simulink. Then, the rollover model and position model in typical working conditions are simulated and compared with Trucksim model and Prescan model respectively to verify the rationality.Collision warning or rollover warning and other warning systems need to acquire various state parameters in real time of vehicle by on-board sensors. The warning system has been challenged because the sensor accuracy is not high, or some state parameters can’t be measured directly due to cost limits. Based on Kalman filtering which is one of the multi information fusion method, rollover dynamic model and vehicle position prediction model, the vehicle states are estimated in view of rollover warning system and collision warning system, in order to improve parameters precision, estimate some state parameters unmeasured, and improve the system reliability and reduce the cost. The vehicle rollover warning algorithm based on TTR and the vehicle cooperative collision algorithm based on TTC are researched, achieve the purpose of give driver warnings in rollover and collision conditions. Finally, active rollover prevention control and active collision prevention control are simulated to verify the rollover dynamic model and vehicle position prediction model.