Study On Simulation Of Identification Of Driving Hazard Based On Vehicle Operating Parameters

With the rapid development of highway in China,traffic accidents have become one of the most serious accidents that threaten the safety of people’s lives and property.The active safety technology of automobiles has gradually become the focus of attention.Identification of driving hazards is an important research content of active safety technology,which can effectively avoid vehicle collisions and reduce traffic accidents.This paper studies the identification of driving hazards based on vehicle operating parameters,hoping to accurately identify and record the dangerous states encountered during vehicle operation,thereby improving driving safety.The dissertation conducts relevant research on the identification of driving hazards based on vehicle operating parameters.First of all,using UC-win/Road software to build a vehicle driving scene of two-way six-lane based on the highway operating environment.The dangerous state of the test scene is set to the sudden emergency braking of the moving model during the process of driving the vehicle model following the target moving model,and the tester takes measures to avoid collision.Collect the operating parameters of the vehicle model and the time to traffic collision in this process.The vehicle operating parameters include vehicle longitudinal acceleration,lateral acceleration,body yaw rate,steering wheel rotation rate and braking force,and thus constitute an independent candidate set X.Based on Set X,three traffic conflict time models were established using multiple linear regression analysis and multiple nonlinear regression analysis methods,respectively.Based on the independent variable set X,the first model of the time to conflict M1 was established using the multiple linear regression method.Based on model M1,stepwise regression was used to improve M1.After the improvement,a second model of the time to conflict M2 is obtained.Using the grey correlation degree to process the independent variable set X,a model M3 between the time to conflict and independent variable set is established.Through the use of measured data for comparison,it was found that M3 can identify the time to conflict well.Finally,the model of the time to conflict M3 is improved.In order to reduce the risk of“false positives” in M3,the dangerous state of driving is screened by time window,and the identification model considering the time window is extended to the urban fast road.The test data in the simulation environment are tested in Matlab.The results show that theidentification model of vehicle dangerous state also has a high recognition rate for the expansion situation and it can identify when the vehicle is in a dangerous state.