The Design Of Automobile Anti-collision Warning System Based On Driving Tendency

The traffic accidents not only caused a large number of casualties, also also bring huge economic losses to people. According to the survey, more than 90% of the accident every year was caused by the driver. The one reason is the driver’s improper operation or distraction, another reason is the existing warning system does not warn in time, which there is some false alarm rate. As a result, drivers do not have enough time to deal with the accident. According to the data shows, the difference of individual driving tendency lead to the required response time is not the same. Even the same driver, under different driving conditions, the reaction condition is also different, which in turn requires reaction time is also different. So further studying of driver’s dynamic driving tendency identification method and the real-time warning is very significance. On the basis of researching domestic and foreign driver tendency and vehicle collision warning system, this paper does some research as follows.Firstly, research of the driving tendency identification method. This paper studied the relationship between the driver’s driving behavior(frequency of braking, frequency of acceleration, depth of braking and depth of acceleration) and driving tendency through vehicle experiment, and the driving tendency was quantified by driving tendency of granularity. Then the identification methods of drivers’ real-time driving tendency was established by the method of fuzzy inference, with the drivers’ driving behaviors as input variable and the driving tendency of granularity as output variable, and the drivers’ reaction time of the warning algorithm was corrected by the driving tendency of granularity. At last, the warning algorithm compared with the classic warning algorithm.Secondly, the warning/brake model was established. Warning/braking algorithm based on the longitudinal collision was established, according to the relationship between the driver’s driving behavior and the longitudinal time of collision avoidance, and then the driving tendency was quantified by driving tendency of granularity, which was used to revised the reaction time of the warning algorithm. The revised warning algorithm compared with the classic warning/braking algorithm.Thirdly, design of hardware and software of the warning system. Hardware design is mainly to circuit principle diagram and designing, welding and debugging the circuit board according the circuit principle diagram. The software design is mainly to write C language code and debug the program with the Code Warrior software of Freescale company. Finally making the compiled program write into the Freescale chip on the circuit board, then making the corresponding circuit test and function testAt last, making real vehicle experiment through using of the experimental platform designed. Three drivers with different driving tendency, which were selected by static driving tendency psychological questionnaire, were carried on the real vehicle experiment. Analysis of experimental data shows that warning system can not only predict the warning/brake, but also, to a certain extent, be adapt to the drivers’ individual driving characteristics, and has low false alarm rate. it has important significance to ensure the driving safety of vehicles and reduce the number of traffic accidents.