Research On Autonomous Emergency Braking Control Strategy Based On Driving State Recognition

With the rapid development of the automobile market,safety technology has become the focus of automobile companies' continuous investment and research.How to ensure the safety of all passengers in the complex driving environment is the development direction of automobile safety technology.Autonomous Emergency Braking is an active safety configuration that assists the driver in emergency braking at critical moments.Information and sensing technology is adopted to improve the driver's perception ability and assist the driver to avoid accident or reduce accident injury.The safe distance model is the foundation of AEB system,and the appropriate safe distance model plays an important role in achieving good effect of the system.In view of the existing safety distance models,which have difficulties in parameter selection,low accuracy and neglect of driver individuation differences,this paper conducts the following research.Firstly,this paper takes the safe distance model as the research focus.Firstly,the Back Propagation Network was carried out on the driving condition identification.According to the correlation coefficient method to select input characteristic value for maximum speed,average speed and idle time,average acceleration,maximum acceleration and minimum acceleration,the six parameters as the input of neural Network output level for driving cycle,each driving condition of average impact degree should be the standard under the condition of the driving traffic identification cycle and predicting period of 120 s and 6s respectively.According to the corresponding standard drivers' transportation average impact degree,combined with the driving condition identification coefficient formula obtained on state recognition,according to the driving condition 0.5 and 1.0 threshold divided into conservative type,standard type and radical type,cycle driving state recognition and prediction period were analyzed respectively,9 s and 3 s,recognition accuracy is higher.In the end,coefficient of driving condition recognition on the safety distance model of the minimum safety distance and driver reaction time to modify these two parameters.The next,Carsim/Simulink co-simulation model was established.According to the C-NCAP standard test case,different driving states under multi-speed scenarios were simulated and analyzed to verify the effectiveness of AEB control strategy based on driving state recognition.The results show that the autonomous emergency braking system based on driving state recognition can better coordinate the critical safety distance and critical alarm distance,which is more in line with the driver's expectation of the system.Finally,the experimental platform is built based on a certain company's 77 GHZ millimeter-wave radar.Before the experiment,first ensure that the radar installation accuracy is in line with the requirements,then the active safety test was carried out.Experiments verify the accuracy of the Carsim/Simulink co-simulation model established in this paper,and the accuracy meets the requirements.It also shows that the control strategy based on driving state recognition has certain practical significance.