Research On Path Planning Methods For Driverless Car

With the support of national policies for driverless driving and the rapid development of 5G technology and other network technologies,driverless cars are an inevitable trend of the development of the automotive industry in the future.However,due to the sharp increase in the number of cars,the transportation sector is faced with traffic safety,road congestion,energy consumption,environmental pollution and other problems.It is extremely urgent to research and develop energy-saving and environment-friendly clean energy electric vehicles and realize stable and safe unmanned driving.Artificial intelligence is an innovative technology that can change the world,and autonomous driving is one of the most important technologies in this field.Autonomous driving is the integration of multiple technologies,including modules such as vehicle positioning and environment awareness as well as path planning and motion control.Planning an optimal path in a complex scene is the basis of unmanned driving.Ensuring the trackability of the path is the key to realize unmanned driving.To establish an efficient controller to track and control the trajectory accurately is the ultimate goal of unmanned driving.This paper designs an implementation scheme of unmanned path planning,and verifies the feasibility of path planning through speed planning and trajectory tracking control.The main work is as follows:(1)Unmanned path planning in static environment.First of all,to improve the artificial potential field method,the repulsion of artificial potential field method is decomposed into two,in order to improve the artificial potential field method is easy to fall into local optimum and target inaccessible shortcomings,secondly the artificial potential field method with fast search random tree and introducing gravitational forces adjustment factor,finally increase the elliptic restricted areas to reduce the full range of RRT algorithm and point of huge computation,Simulation results show that the improved method can efficiently plan the optimal path.(2)Research on speed planning of unmanned driving.Establish dynamic driving scene,will be based on the improved method to get the optimal path using b-spline curve path smooth,using S-T figure velocity planning algorithm mapping obstacles,match the speed and path to get smooth trajectory of obstacle avoidance,based on the digital simulation show that velocity of Matalab planning trajectory can achieve dynamic obstacle avoidance process of scenario.(3)Research on tracking control of the planned trajectory.Firstly,the vehicle kinematics model is established,which is transformed into a linear discrete system as the predictive model of the model predictive controller through Taylor transformation and Euler transformation.Then,the objective function and constraints are established,and the model predictive control problem is transformed into a quadratic programming problem.MATLAB is used to build the model predictive controller.Finally,trajectory tracking control in dynamic scene is realized based on model predictive controller.(4)Since it is difficult to realize the real driverless vehicle experiment,Prescan,an unmanned driving system simulation software,is used in this paper to build a driving scene that reflects the real driving situation,and a simulation system is built jointly with MATLAB Simulink to realize the co-simulation of trajectory tracking control in dynamic scenes.The co-simulation results show that the proposed method can realize the track tracking control of obstacle avoidance lane change in dynamic scenarios and has good smoothness while satisfying the vehicle kinematics model.