Research On Lane Changing Model Of Unmanned Vehicle Under Intelligent Network

The continuous progress of sensing technology and artificial intelligence technology has promoted the rapid development of the transportation industry in the direction of intelligence and networking,and driverless cars have also become a research hotspot for major enterprises and efficiency.Changing lanes is one of the basic driving behaviors,and according to statistics,about 1/4 of traffic accidents are caused by changing lanes.Relying on intelligent network technology and advanced sensors,driverless cars can obtain traffic information around the vehicle in real time,provide data support for the vehicle’s lane change,and improve the safety and stability of lane change.In this thesis,the lane change model of unmanned vehicles in the intelligent networked environment is studied,and a lane change decision model is established from the two aspects of lane change safety and benefit,multi-objective optimization is carried out on the basis of the fifth polynomial curve function to complete the optimal lane change trajectory planning,and a tracking controller is established to track the optimal lane change trajectory.Specific research content is as follows.(1)The lane change decision model of driverless vehicle in intelligent networked environment is studied.By analyzing the vehicle’s lane change decision logic,the lane change decision is studied from three parts: car-following,lane change safety and lane change income.Firstly,based on the characteristics of intelligent networking technology,an IDM car-following model considering multi-vehicle information is established.Secondly,the virtual mass function is obtained by fitting China’s accident data and the vehicle safety potential field is established.Based on this,the minimum longitudinal safety distance calculation model is constructed.Finally,the lane value function and incentive criterion model are proposed to analyze the benefits of lane change and provide basis for lane change decision.(2)The lane-changing trajectory planning model for unmanned vehicle is studied.The trajectory planning is considered as the optimal control problem with free terminal time and constrained terminal state.The quintic polynomial function is obtained by minimizing the horizontal and vertical fluctuations of lane-changing vehicles.On this basis,the vehicle dynamic stability during lane-changing is analyzed,and the lanechanging time constraints that meet the vehicle dynamic requirements are obtained.Finally,the TOPSIS algorithm is used to optimize the lane-changing trajectory from four aspects: safety and comfort,and the optimal lane-changing trajectory is obtained.(3)The MPC trajectory tracking controller was studied.The model predictive control theory is elaborated and analyzed,and the objective function is designed considering the error of the state quantity and the control quantity on the basis of the linearization of the system state equation.The constraints of the model are proposed from the aspects of vehicle parameters and road surface conditions to improve the comfort and stability of the controlled vehicle.(4)The lane-changing model of unmanned vehicle is simulated and analyzed experimentally.Firstly,the feasibility of lane change decision model is verified by numerical simulation.Then the optimal lane change trajectory of the vehicle in the specified driving scenario is obtained by using the lane change trajectory planning model.Finally,Car Sim/Simulink joint simulation platform is established to model and simulate the designed trajectory tracking controller,which realizes the tracking of the optimal lane change trajectory.