Unmanned Vehicle Gauss Pseudo-Spectral Method Dynamic Obstacle Avoidance Trajectory Planning And Tracking

Autonomous driving technology has become an important development direction of the automotive industry,and unmanned vehicles are being focused on and researched by universities and enterprises at home and abroad.Local trajectory planning and route tracking technology in autonomous vehicles play a very important role,so this thesis has launched an in-depth study on it.Taking into account the kinematic constraints and geometric collision relations of the vehicle in the process of movement,this thesis adopts the Ackermann vehicle steering model as the kinematic model of the unmanned vehicle,and uses the P criterion expression as the geometric model of the unmanned vehicle.By analyzing the geometric position of the vehicle,the necessary conditions for the safe driving of the vehicle in the lane are established.In this thesis,the possibility of collision between unmanned vehicles and obstacle vehicles is analyzed from multiple aspects,and the constraint conditions to prevent collisions are determined by the position relationship of the vehicles.The trajectory planning algorithm takes the planned shortest path as the objective function,which saves more energy when tracking the trajectory of the autonomous vehicle.The Gaussian pseudospectral method uses Lagrangian interpolation basis functions to approximate the state variables and control inputs in the control system.In order to make the interpolation function accurately approximate the state variables and control inputs in the control system,the discrete point of the interpolation function is Lagrangian The zero point of the first derivative of the Japanese orthogonal polynomial,which is solved using Newton iteration method.The performance function and constraint conditions of the optimal control problem have also been transformed accordingly.After discretization by pseudospectral method,the optimal control problem is transformed into a high-dimensional nonlinear programming problem.This thesis establishes the conversion relationship between the actual world coordinate system represented by latitude and longitude and the dynamic path planning coordinate system,which can combine the planning algorithm with the actual automatic driving vehicle.It also describes the method of obtaining the initial value of the intermediate state variables in the nonlinear programming problem.According to the avoidance characteristics of autonomous vehicles,the sigmod function is used to approximate the planned trajectory,and it and its derivative are used as the initial value of the search of the quadratic sequence planning algorithm,which can greatly improve the calculation speed of the search algorithm.In order to realize vehicle tracking,this thesis introduces the basic principles of pure preview tracking algorithm,and on this basis,it explains in detail the importance of the unmanned vehicle preview tracking model and preview distance.The fuzzy preview tracking controller is used to reduce the large lateral error of the pure preview tracking controller,and the MATLAB/Simulink simulation software is used to simulate its tracking ability on the sinusoidal road.At the end of the thesis,a simulation analysis of the trajectory planning is done,and the feasibility and real-time performance of the trajectory planning scheme in the thesis are verified,and the reliability of the trajectory tracking by the fuzzy tracking controller is also verified.