Research On Path Planning And Following Control Based On Lane Changing Behavior Of Intelligent Vehicle

At present,the intelligence of automobiles has become the foothold and growth point of the development of today's automobile industry.The emergence of intelligent vehicle can make human travel more convenient,diversified,and safer.In the actual traffic environment,vehicle lane changing is one of the common behaviors that easily lead to traffic problems.In the accidents caused by lane changing,improper driver operation is the main incentive.In order to make the intelligent vehicle complete the lane changing operation well,make up for the human aspect and improve the rationality and safety of the lane changing.Based on the lane changing behavior of intelligent vehicle,this paper studies the lane changing decision model,lane changing path and velocity planning,and following control algorithm.Firstly,the research of lane changing decision model of intelligent vehicle is carried out.By analyzing the driver's lane changing mechanism,the decision process is divided into two parts: lane changing requirement and feasibility.The data of real driver lane changing is collected by the driving simulator,the risk perception RP coefficients with the surrounding obstacle vehicle is introduced as the characteristic parameter of the decision model.Use the RP of the front vehicle to design the value of the danger accumulation degree to characterize the lane changing demand,and use the remaining RP as the input of the SVM model to conduct a lane changing feasibility analysis of the target lane,then a reasonable and safe lane changing behavior decision model is established.After training with data,the model decision accuracy rate can reach 96%,and the model feasibility verification is performed.Secondly,after the intelligent vehicle completes the lane changing decision,it is necessary to plan the lane changing path and velocity.In order to make the lane changing path and velocity curve flexible while ensuring safety,to adapt to different driving requirements,a planning algorithm based on 3D cubic Bezier curve is proposed,the entire lane changing process is divided into two parts: the collision avoidance aligning phase.The segmented path planning is performed in the XY plane.The coordinates of the control points are obtained by satisfying the vehicle dynamicconstraints and boundary conditions,and the genetic algorithm is used to solve the lane changing destination by.And the Z axis coordinate is used to perform segmented velocity planning,the vehicle speed range is divided,and the classified velocity planning is performed,the constraints and optimization algorithm are used to solve the nonlinear programming problem to obtain the velocity curve.Through stitching,the integrated lane changing path and velocity curve are obtained.Thirdly,in order to make the intelligent vehicle follow the planned lane changing path and velocity,a vehicle dynamic model and a velocity following driver model are established,and a model predictive control algorithm is used to design the trajectory tracking controller.By constructing the CarSim and Simulink co-simulation platform for path following simulation analysis under different vehicle speeds,the results show that the planned path and velocity are effective and feasible and the following control algorithm has good tracking effect and robustness.Considering the real-time variability of the path and velocity,the fixed preview step is difficult to meet various conditions,so a variable preview step path following control algorithm based on multiple targets is proposed,the velocity,path curvature and deviation are taken as multiple targets,so a variable preview step strategy is designed to adapt to different lane changing conditions.The simulation results show that this strategy can improve vehicle path following accuracy and guarantee the driving stability.Finally,a HiL test platform for intelligent vehicle based on VCU control strategy is built,including the host computer system,real-time simulator system and D2 P rapid prototype controller.The Veristand is used to configure system.An intelligent vehicle lane changing path planning and following control algorithm model is built in the D2 P controller by the Motohawk and Simulink.The HiL test platform is used to perform simulation analysis at different vehicle speeds.The results verify the effectiveness of the planning and following control algorithm in the actual controller.