Development Of Intelligent Cruise Control Algorithm Based On Preceding Vehicle Lane Change Intention Identification

With the increasing problem of road traffic,humans are paying more and more attention to autonomous driving technology.In recent years,the adaptive cruise system,as the key technology of advanced driving assistance system,has become increasingly mature and gradually entered the lives of the general public.However,the traditional adaptive cruise system only controls the longitudinal motion of the vehicle.In the case where the vehicle in front is running slowly for a long time and the lane changing conditions are good,autonomous lane changing cannot be performed,which greatly reduces driver usage and satisfaction with the adaptive cruise system.At the same time,on the condition of the lane change of the leading car in the adjacent lane,the traditional adaptive cruise system can not comprehensively consider the main target in the lane and the leading car in adjacent lane.Under this condition,the conventional adaptive cruise control system will experience a large fluctuation in longitudinal acceleration,which greatly reduces the comfort and may even be dangerous.Therefore,it is of great significance to identify the lane change intention of the leading vehicle and to study the autonomous lane change control of the vehicle.This paper develops an intelligent cruise algorithm based on the identification of the lane change intention of the leading vehicle.The main content includes the selection of the main target based on the identification of the lane change intention of the leading vehicle,the autonomous lane change decision and lane change trajectory planning,the longitudinal and lateral motion control of intelligent cruise vehicles,the verification of proposed algorithm in the co-simulation platform based on Matlab/Simulink,CarSim and Prescan software.The specific research has following aspects:(1)This paper first divides the area of interest based on the target characteristic data detected by the millimeter-wave radar and the lane line polynomial fitting data detected by the camera sensor,and filters the targets outside the area of interest.According to the prediction data of "Effective target" output by the Kalman filter at the previous cycle and the radar data in the region of interest at the current cycle,valid measurement of "effective target" is already filtered.According to whether there is valid measurement data,the life cycle of all "effective targets" is calculated.The target state estimation method based on the interactive multi-model and the life state of the "effective target" are used to estimate the state of each target,and the support vector machine is used to identify the lane change intention of each "effective target".Based on the status of the "effective target" and the identification result of the lane change intention,the final main target is selected.(2)Aiming at the condition that the leading vehicle drives slowly for a long time and the lane changing conditions are satisfied,in this paper,the autonomous lane change decision and lane change trajectory planning algorithm are studied.The autonomous lane change decision part is mainly divided into: calculation of unsatisfactory degree of current lane,judgment of speed advantage of adjacent lanes,and feasibility analysis of autonomous lane change.The dissatisfaction calculation of current lane is mainly divided into two parts: the expectation of speed and the expectation of spacing.According to the calculation results of dissatisfaction of current lane,the intention of lane change is established.Under this premise,it is necessary to judge the speed advantage of the adjacent lanes and ensure that the target lane has a speed advantage relative to the original lane.This is of great significance for the autonomous lane change decision of the intelligent cruise algorithm.Finally,this paper analyzes the feasibility of autonomous lane change.Based on the assumption that the vehicle is in a uniform acceleration state and the surrounding traffic vehicles are in a uniform speed state,the collision safety of the host vehicle with the leading vehicle of the original lane,the leading vehicle of the target lane and the following vehicle of the target lane is analyzed.By establishing inequality constraints on lane change time and longitudinal acceleration,the feasibility of autonomous lane change is transformed into the question of whether the uncertain parameters have feasible solutions.The autonomous lane change trajectory planning is mainly based on the fifth-degree polynomial curve to plan the lane change trajectory for straight lane change and constant curvature curve lane change scenarios.After comprehensively considering the impact of the lane change process on the road flow and the speed error and the spacing error between the host vehicle and the leading vehicle after the lane change is completed,the indefinite parameters are optimized and the autonomous lane change trajectory is obtained..(3)After obtaining the state of main target and the results of autonomous lane change decision and lane change trajectory planning,the intelligent cruise vehicle's lateral and longitudinal motion control algorithm is researched.Firstly,according to the results of lane line polynomial fitting data,autonomous lane change decision and autonomous lane change trajectory planning results,a reference trajectory based on the body coordinate system of the vehicle is generated.In order to follow the reference trajectory,a lateral dynamic model of the vehicle is established.The steering wheel angle is calculated by using the LQR controller,and the steady-state lateral distance error is eliminated by applying the feedforward steering wheel angle.Based on the status of the main target(relative longitudinal speed and relative longitudinal distance),the LQR controller is selected to calculate the target longitudinal acceleration of the vehicle.According to the target acceleration,vehicle speed,road slope and other information,the vehicle inverse dynamics model is established,and the feedforward actuator control amount is calculated.The PID controller is used for feedback correction to achieve effective control of vehicle longitudinal acceleration.(4)The co-simulation platform is built based on Matlab/Simulink,CarSim and Prescan software to verify the effectiveness of the proposed algorithm.The scenario and sensor models are established in Prescan,the high-precision vehicle dynamics model is established in Carsim,the simulation environment integration and control algorithm are established in Matlab/Simulink.The simulation results show that the intelligent cruise algorithm based on the lane change intention identification proposed in this paper can effectively comprehensively consider the main target in the current lane and "effective target" in adjacent lanes.And under the condition that the leading vehicle drives slowly for a long time and the lane changing conditions are satisfied,the host vehicle can change lane autonomously.