Research On Forward Collision Warning Based On Intelligent Vehicle Motion State Estimation

In the face of increasing car ownership and huge traffic safety accident losses,Automotive Active Safety Technology has become an important indicator to measure the safety of intelligent vehicles.The development of intelligent vehicle forward collision warning technology can effectively prevent the occurrence of collision accidents,which is of great significance for reducing the loss of casualties and ensuring road safety.With the continuous development of technologies such as sensors,information processing and intelligent control,the vehicle anti-collision warning technology can integrate the information of self-driving and surrounding environment through intelligent algorithms,and consider the characteristics of the driver to improve the existing problems and make it more Intelligent and precise.In this paper,the intelligent vehicle in the city is taken as the research object,and the information acquisition,safety distance model and control early warning decision of the forward collision warning system are studied.The intelligent vehicle motion state estimation is combined with the collision warning system,and the driving environment is considered to improve the safety distance model,then the forward collision warning system integrating the target vehicle tracking and driving intention identification is established.Firstly,the motion state estimation of the forward target vehicle is studied,and the Unscented Kalman Filter algorithm based on the current statistical model is proposed.The relative distances and velocities under the two conditions of uniform acceleration and variable speed motion are estimated in MATLAB.The simulation results show that the current statistical model is closer to the motion law of the vehicle,and the estimation error of the proposed algorithm is less than that of the Extended Kalman Filter algorithm,which can effectively track the motion state of the preceding vehicle.Secondly,aiming at the insufficient adaptability of the safety distance model of the collision warning system,the minimum safety distance model is established based on the vehicle braking process and the moving state of the preceding vehicle.Using the fuzzy inference algorithm,the driver response time and the maximum braking deceleration under different weather status and road conditions are discussed.On this basis,the forward collision warning system is designed,the alarm distance and the dangerous alarm distance are calculated,the driving intention fuzzy identification model is introduced,and the three-level collision warning strategy under different driving intentions is proposed.Finally,using the Fuzzy toolbox and Simulink module of MATLAB,the calculation results of the minimum safe distance based on fuzzy inference and the calculation results of the traditional model are compared,and the influence of driving environment on the safe distance is verified,and the model with high adaptability is helpful to reduce the false alarm rate.The pre-simulation of PreScan/Simulink is used to establish the virtual test scenario under low-speed and high-speed conditions,and verify the effect of collision warning system after introducing vehicle motion state estimation.The simulation results show that the early warning system is more reasonable and reliable,and can successfully avoid collision accidents.