Autonomous Vehicle Trajectory Planning And Motion Control Under Complex Attachment Driving Conditions

As the global automotive industry undergoes profound changes,intelligent driving technology for automobiles has become an essential research component for innovation and development in the automotive industry.Intelligent driving system has great advantages in reducing the number of accidents,reducing the severity of unavoidable accidents and improving travel efficiency.It can be generally divided into four parts according to system functions: environment perception,decision making,trajectory planning and motion control.As an important part of intelligent driving system,trajectory planning and motion control will directly affect the safety,stability and comfort of automobile driving.In the real urban traffic,the vehicle is likely to drive at high speed in extremely complex road attachment conditions,such as conditions with local snow or dynamic switching between different road surfaces,which poses a great challenge to the adaptiveness of the planning and control algorithm of the intelligent driving system.Therefore,in order to improve the adaptive capability of the system to complex driving conditions,improve driving stability and safety,and realize the coverage of driving conditions,this paper designs and implements a trajectory planning and motion control system scheme based on double-layer cascade model predictive control(MPC),and the research mainly includes the following three aspects.1.For the safe driving problem of lane changing and collision avoidance of selfdriving vehicles,a longitudinal-lateral coupled trajectory planning and motion control algorithm is designed under a double-layer cascaded model predictive control architecture.Which corrects the vehicle state by using the characteristics of doublelayer MPC with feedforward-feedback correction,predictive capability,and rolling optimization.Firstly,the trajectory planning layer designs the cost function and constraints by constructing a nonlinear kinematic model with longitudinal and lateral coupling,solves them optimally to obtain the optimized longitudinal and lateral acceleration,then obtains the predicted trajectory of dynamic planning as the output;secondly,the motion control layer designs a controller based on linearized vehicle dynamics,solves the vehicle acceleration and front wheel steering angle for drive or brake control and steering control of the vehicle,and realizes the tracking of dynamic reference trajectory and vehicle speed;finally,the proposed trajectory planning and motion controller is simulated and verified that the algorithm can effectively ensure the continuity,smoothness,reasonableness and safety of intelligent driving.2、 Considering the complexity of vehicle dynamics under mixed attachment road conditions,a motion controller based on nonlinear model predictive control is designed.Firstly,the nonlinear dynamics model and tire model are established to accurately predict the state of the vehicle,which ensures the dynamic tunability of the motion control process of the autonomous vehicle and the adaptiveness under mixed road conditions;secondly,the multi-dimensional performance evaluation indexes of longitudinal position tracking capability,lateral position tracking capability,speed tracking capability and side slip angle suppression ability are designed;finally,the setting of Finally,the performance of the controller is verified by setting an extreme driving condition,"snake condition",and the improved motion controller is shown in the form of "radar diagram" to improve the tracking performance.3.In order to improve the tracking control effect and the coverage of complex driving scenarios,a trajectory planning and motion controller based on the attachment stability factor is designed.Firstly,for the problem of uncertainty of tire-road contact surface under complex conditions,the attachment stability factor-based trajectory identification method for intelligent vehicles is adopted;secondly,the trajectory planning strategy based on the attachment stability factor is designed,which adjusts the longitudinal and lateral acceleration constraints to be optimized by the weight factor,and then can dynamically adjust the trajectory so that the intelligent vehicles can cover more driving scenarios and effectively improve the lane change driving efficiency.At the same time,the efficiency of lane change driving is effectively improved;finally,a motion control switching strategy based on attachment stability factor for regular driving and emergency driving conditions is given,and the effectiveness of the proposed controller for improving driving efficiency,adaptive capability and driving safety is verified through several typical checkerboard,docking and docking mixed conditions of vehicle snake driving and lane change driving.