Research On Dynamic Path Planning And Control Of Driverless Vehicle

Unmanned vehicles can effectively improve traffic safety and congestion issues,and have broad application prospects.With the rapid development of sensing technology and computer technology,more and more enterprises and universities both domestically and internationally have increased their research and development efforts on autonomous vehicles.As one of the core technologies of autonomous vehicles,path planning and trajectory tracking are the technical foundations for achieving autonomous vehicle driving.Therefore,according to the needs of the research topic,this article conducts research on the global path planning,dynamic obstacle avoidance trajectory planning,and trajectory tracking control of unmanned vehicles in dynamic environments.Ant colony optimization algorithm is the most widely used biomimetic optimization method in global path planning,but it has disadvantages such as slow convergence speed and easy to fall into local optima.Therefore,a new heuristic function and global pheromone updating strategy are used to propose an ant colony optimization algorithm for global path planning of unmanned vehicles.Then smooth the generated path using non-uniform rational B-spline curves.The simulation results validate the feasibility and effectiveness of the ant colony optimization algorithm in scenarios with obstacles.Compared with traditional ant colony algorithms,improved ant colony algorithms search for shorter path lengths and require fewer iterations to converge to the shortest path.When autonomous vehicles are driving in complex road environments,it is also necessary to have reliable and stable methods to avoid moving obstacles.A local path planning based on the linear time-varying model predictive control theory is proposed.The monorail vehicle dynamics model is used as the prediction model,and the vehicle dynamics constraints are considered to predict the travel path of vehicles and obstacles.A reasonable obstacle avoidance function is designed to make the planned local path as close to the global reference path as possible while the vehicle avoids obstacles.In order to ensure that the vehicle accurately tracks the upper layer reference local path,the trajectory tracking controller selects a more accurate dual track vehicle dynamics model as the prediction model,and obtains the control quantity(vehicle front wheel angle)from the vehicle state quantity obtained from the sensor according to the linear time-varying model predictive control theory to control the vehicle to track the reference path as far as possible.Integrate local path planning hierarchical controllers based on linear time-varying model predictive control theory,and build the joint simulation platform of Car Sim and Simulink verifies the safe obstacle avoidance of driverless vehicles in the scene where both curves and dynamic obstacles exist.The simulation results show that the vehicles can avoid obstacles safely,but also meet vehicle dynamics and kinematics constraints,indicating the feasibility of the local path planning layer and controller design theory proposed in this paper.