Research On Vehicle Active Collision Avoidance System Based On Multi-factor Influence

As an important part of intelligent driving system,active collision avoidance system can effectively avoid rear-end collision,improve driving safety and reduce accident rate,which is very important to promote the development of intelligent transportation.As the most classical model in the risk assessment model,the safety distance model has the advantages of universality and safety.The traditional safety distance model is mainly based on the kinematic relationship between vehicles,and the initial state of the rear vehicle receiving the warning signal is often regarded as uniform driving,which is inconsistent with the actual situation.In addition,due to the complex force situation during the braking process of the vehicle,if the influencing factors are not considered enough,the reliability and safety of the active collision avoidance system under complex traffic conditions will be reduced.Therefore,in order to make the active collision avoidance system more flexible to adapt to complex and changeable working conditions,this thesis takes the safety distance model as the starting point,deeply analyzes the influence of the initial state of the rear vehicle,the road adhesion coefficient,the road slope,the vehicle load,the driver ’s characteristics and the motion state of the front vehicle on the safety distance,establishes a multi-factor fusion safety distance model,and considers the driver ’s personalized difference,and proposes an active collision avoidance control strategy that integrates the driving style.The main research contents and results of this thesis are as follows :(1)The influence of the initial state of the following vehicle,the road adhesion coefficient,the road slope,the vehicle load,the driver ’s characteristics and the different motion states of the preceding vehicle on the safety distance is analyzed.When the safety distance model is constructed,the initial state of the following vehicle receiving the warning signal is considered.There are three cases of uniform speed,acceleration and deceleration.Based on this,the kinematics and dynamics methods are used to analyze the force of each stage in the braking process,and a staged braking deceleration model is obtained.Combined with the motion state of the preceding vehicle,a multi-factor fusion safety distance model is established.Through simulation experiments,the specific influence and variation law of the initial state of the rear vehicle,road slope,road adhesion coefficient and vehicle load on the early warning safety distance are explored.The results show that when the vehicle is in braking condition,the change of early warning safety distance under the influence of four key factors is more obvious.The influence of road adhesion coefficient on early warning safety distance is the most significant.The change rate of early warning safety distance caused by different road adhesion coefficient can reach 228.74 %.The influence of vehicle load is the smallest,and the maximum change rate of early warning safety distance is only 13.23 %.The maximum change rate of early warning distance caused by road slope and initial state of rear vehicle is 39.21 % and49.53 % respectively.(2)Based on the NGSIM data set,the effective car-following data are screened out,and the driving style characteristic parameters are extracted.The factor analysis method is used to reduce the dimension of the selected characteristic parameters.The K-means ++ clustering algorithm is used to divide the driving style into three categories.According to their type characteristics,they are named as fluctuation cautious type,stable and moderate type,and fluctuation radical type.Based on the SVM algorithm,the driving style identification model is built,and the driving style is quantified to obtain the driving characteristic coefficient.Based on this,the buffer distance and driving reaction time are corrected to complete the design of active collision avoidance control strategy integrated into the driving style.(3)The longitudinal dynamics model of the vehicle is established by CarSim software,and the inverse dynamics model is constructed to transform the expected control acceleration.The hierarchical control method is used to design the control system.The fuzzy predictive control is used to design the upper control,and the PID is used to design the lower control.The simulation platform is built by combining Matlab / Simulink and Carsim to form a closed-loop simulation model of the active collision avoidance system,so as to design the active collision avoidance simulation experiments of different driving styles,and the active collision avoidance simulation experiments of different initial states,different road conditions and different loads of the rear vehicle under uniform speed and braking conditions.Comparative analysis of differential warning and control of active collision avoidance system.The results show that the three factors of road adhesion coefficient,the motion state of the front vehicle and the initial state of the rear vehicle have a great influence on the trigger time of each function in the active collision avoidance system.The influence of driving style and road slope on the trigger time of each function in the active collision avoidance system is relatively small,while the influence of vehicle load is the least significant.Among them,the difference between the warning time of light load vehicle and heavy load vehicle is only 0.33 s.