Study On Local Path Planning Of Intelligent Vehicle Based On Gauss Pseudo-spectral Method

With the rapid development of smart cars,it is inevitable that smart cars will be used in the complex transportation system.In the current transportation system environment,the different types of roads,complex task and all kinds of traffic rules,often leads to a control process divided into multiple stages,and in different stages there are different constraints.Gaussian pseudo-spectral method is applied in this article,by combining the terminal state of the last phase and the initial state of the latter phase,we can guarantee the continuity between each stage of state.And then multi-phase control problem is solved as a whole.Furtheermore,it has faster computing efficiency than other methods.In this paper,we firstly use the function based on the P criterion to unify the various types of road,static and dynamic obstacles and intelligent vehicle boundary.Analysis of smart car in local path planning model is made.Boundary and path as well as obstacle avoidance constraints are considered.And the we add phase continuous constraint to establish general multistage local path planning in the form of the optimal control model.Lagrange interpolation is applied based on LG points to approximate state variables and control variables,thus differential equation and the constraints can be converted to algebraic equations.Finallly the optimal control problem is changed into a nonlinear programming problem(NLP),and we select the sequential quadratic programming method to solve it.Parking unparking typical working condition are taken for example to analysis.The vehicle dynamics model and the garage function are established.We did a detailed analysis of the three stages in the vehicle further using gaussian pseudo-spectral method,to solve the problem.The results verified the method for multi-stage problem solving ability.Combined with 5G communication technology,an optimized local path planning algorithm is proposed in this paper.In this paper,LQG method is used to carry out the research work on the key technology-intelligent vehicle path tracking algorithm.In the Prescan simulation environment and real road environment,the trajectory tracking experiment is verified.Real vehicle and simulation experiments indicate that the trajectory tracking algorithm based on optimal control can apply to small curvature straight track and large curvature different scenarios such as curve track,and it showed a stable,accurate tracking performance and speed-adaptive ability under the Gaussian white noise random disturbance.At the same time,four more scenarios-lane changing,pass the bus stop,lane change through the intersection and slow passing through the pavement are taken into consideration to further show the feasibility of this method by the Prescan simulation.