MPC-based Cooperative Control Strategy Of Path Planning And Trajectory Tracking For Unmanned Vehicles

With the development of artificial intelligent,unmanned vehicles have received more and more attention.From the earlier smart vehicles to the current intelligent connected vehicles,unmanned driving is gradually moving towards the lives of ordinary people.This thesis focuses on the local path planning problem and trajectory tracking problem for unmanned system,and builds a real vehicle platform for the driverless sightseeing vehicle.First,the problem of local path planning is investigated.The local path planning problem in unmanned driving can be cast into the dynamic obstacle avoidance problem.Combining the artificial potential field method with the model predictive control algorithm,a local path planning method based on decision process is proposed.Moreover,the application scenario is extended to the V2 X environment,and the traffic signal and overtaking time constraints are taken into account.A planning algorithm that facilitates engineering practice is put forward.Second,we study the trajectory tracking problem of unmanned driving.Based on the kinematic model,a general procedure to the model predictive control trajectory tracking problem using the particle swarm optimization algorithm is presented.The predictive model is used to determine the future state and the cost function is given as the fitness calculation formula for the particle swarm optimization.The optimal particle is iterated to find the optimal control law.Furthermore,considering the overall computing burden of the system,we use the properties of iterative optimization of particle swarm optimization algorithm and the characteristics of local prediction of model predictive control.The particle swarm optimization model predictive control is adopted in both the planning layer and the control layer to reduce the overall computation time of the system,which can be described as a progressive architecture.Co-simulation results verify the feasibility of the algorithm.Finally,the software and hardware implementation of the driverless sightseeing vehicle is suggested.From the perspective of application scenarios and needs,we build a driverless sightseeing car platform based on the common campus sightseeing vehicle.As for the hardware,STM32F103C8T6 MCU is utilized to realize the wirecontrolled modification of vehicle steering,throttle and brake.The robot operating system is adopted for the implementation of vehicle perception,planning and control algorithms.The overall hardware and software joint-debugging of the system shows the design feasibility and reliability of the system.