Research On AEB Control Strategy Based On Driver Characteristic Identification

The research direction of this paper is based on the current hottest research topic of “humancar driving”,focusing on the identification method of driver characteristics,applying the identified driver characteristics to the strategy development of active safety technology AEB system,AEB The early warning part was optimized,and the safety distance model of the AEB collision avoidance part was adjusted to be applicable to the test vehicle.The main researches of this paper are as follow:(1)By analyzing the advantages and disadvantages of several identification models,the fuzzy neural network is selected as the driver identification method,and the driver's braking intention identification model is built by Matlab's ANFIS interface.The three parameters of the pedal opening and closing degree and its change rate and the braking deceleration are set as the identification parameters of the driver's braking intention,the relevant data is obtained by the driving simulator,and the data is trained and inspected to obtain The ideal test results are proved by the verification results.(2)The AEB control strategy is designed,and the early warning level after fuzzy identification is applied to the early warning part of AEB.The existing safety distance model is compared and analyzed,and the Simulink model of Honda safety distance is established.After adjustment,it is suitable for this test vehicle.And established a braking strategy to simulate the automatic emergency test conditions specified by E-NCAP,analyze the simulation results,and verify the rationality of the AEB control strategy.(3)Introduced the laser tracking-based vehicle tracking and identification method,completed the transformation of ABS pump,applied the improved AEB strategy to the real vehicle,and classified the hazard level into first,second and third according to the fuzzy identification results.Level,where the hazard level is not early warning,the hazard level is 2 mild warning,and the hazard level is triple warning.In addition,the obtained experimental personnel follow the vehicle behavior as the basic data,establish the fuzzy neural network model to identify the braking behavior and the alarm value calculation,and finally according to the dangerous alarm situation and false alarm rate obtained by the test.system assesment.