Cubic Spline Path Planning Based On Dynamic Parameter Adjustment And Path Tracking

The automobile industry has developed rapidly in recent years,especially new energy vehicles have been widely accepted by the public.The process of automobile digitization and intelligence has been greatly promoted,and autonomous driving technology has become a research hotspot of automobile enterprises and universities.Autonomous driving technology and assisted driving technology can not only reduce the occurrence of traffic accidents,but also can improve the traffic efficiency of cities by combining automatic driving technology with big data technology.This paper studies the path planning,path following and path smoothing of autonomous vehicles,designs a cubic spline path planning method based on dynamic parameter adjustment,designs a model predictive controller based on dynamics and kinematics,and designs a trajectory Optimization.The main content completed is as follows:(1)A cubic spline path planning method based on dynamic parameter adjustment.In this paper,the parameters of the cubic spline path planning algorithm are dynamically adjusted based on the vehicle rollover and sideslip models.This makes the lateral acceleration of the vehicle less than the critical acceleration calculated by the rollover and sideslip model when driving on the planned path,so as to reduce the occurrence of rollover accidents.The dynamic parameter adjustment module realizes the dynamic adjustment of the parameters of the cubic spline path planning algorithm,and mainly includes a horizontal planning range dynamic adjustment module,a vertical range adjustment module and a vertical range expansion module.After a series of candidate path sets are generated by the cubic spline path planning algorithm based on dynamic parameter adjustment,the evaluation function selects the best path from the candidate path sets.(2)Model predictive control(MPC)algorithm based on vehicle kinematics model and dynamic model.In this paper,a controller is established based on the vehicle dynamics model and the vehicle kinematics model,respectively,and the MPC algorithm based on the vehicle dynamics model and the vehicle kinematics model is simulated under the double lane change condition.The simulation results show that both have better path following effects.Finally,the simulation results of the MPC algorithm based on the vehicle dynamics model and the vehicle kinematics model are compared and analyzed.(3)Trajectory smoothing based on quadratic programming and gradient optimization.In this paper,the quadratic programming algorithm is used to realize the smooth optimization of the road centerline.Then,this paper analyzes the path optimization method based on gradient optimization,and analyzes the solution of nonlinear optimization problems using the steepest descent method and the conjugate gradient method.This paper proposes a solution to the problem of in-situ steering that may occur when shifting forwards and backwards during parking,and the simulation results verify the effectiveness of the algorithm.(4)The co-simulation of the cubic spline path planning algorithm and the control algorithm with dynamic parameter adjustment is realized.In this paper,driving conditions based on specific speed changes are simulated.In order to better analyze the effectiveness of the cubic spline path planning algorithm for dynamic parameter adjustment.Based on the driving scenarios under different vehicle speed conditions and the evaluation indicators of rollover and sideslip,further simulation comparisons were carried out.It mainly includes straight road scenes and intersection driving scenes.The simulation results verify that the improved cubic spline path planning algorithm can reduce the occurrence of vehicle sideslip and rollover accidents.