An Improved Artificial Potential Field Method Intelligent Vehicle Study On Obstacle Avoidance And Lane Maintenance

With the continuous improvement of people’s living standard,the number of cars shows an exponential growth,the road traffic problem is more and more serious,traffic accidents often occur.The Advanced Driving Assistant System(ADAS)plays a crucial role in relieving road traffic pressure and preventing further traffic accidents.ADAS includes many auxiliary driving functions.Vehicle obstacle avoidance and lane keeping are one of the hot issues for researchers to explore in ADAS.Therefore,this paper conducts research on vehicle obstacle avoidance and lane keeping system.(1)Establish the two-dof vehicle dynamics model,electric power steering system and the optimal driver model of side-acceleration preview.In order to make the intelligent vehicle follow the planned obstacle avoidance path,a two-degree-of-freedom vehicle model is firstly established to analyze the relationship between kinematics and dynamics among various parameters of the vehicle.Secondly,the steering wheel is controlled to meet the steering function required by vehicle obstacle avoidance.Then,based on the prediction-follow theory,the optimal pilot model of lateral acceleration for preview was built,and the simulation under different working conditions was carried out in MATLAB/Simulink software.The simulation results were compared with the tracking curve of the driver model in Carsim software to verify the accuracy of the established driver model,and the model was used as a simulation platform.(2)The traditional artificial potential field method is improved to build new attraction and repulsion potential fields,and then the two different obstacle avoidance and lane changing models are integrated to assist the intelligent car to avoid static and static obstacles.Artificial Potential Field Method(APF)has the characteristics of small amount of calculation,easy to be expressed by mathematical formula,and easy to realize real-time control,etc.The improved APF algorithm is selected as the control algorithm of vehicle obstacle avoidance and lane keeping control system.In order to overcome the local traps and unreachable defects of the traditional APF algorithm,the relative distance between the intelligent vehicle and the target point is increased in the repulsion field of the traditional APF algorithm.When the intelligent vehicle reaches the target point,the attraction generated by the target point and the repulsion generated by the obstacle are both zero,thus overcoming the defects of the traditional APF algorithm.In order to describe the driving environment information of intelligent vehicles more accurately,the repulsion potential field and obstacle repulsion potential field at the road boundary are constructed to prevent collision between intelligent vehicles and obstacles.In order to improve the safety and ride comfort of intelligent vehicles in obstacle avoidance,the sinusoidal and isokinetic migration model are combined to form the sinusoidal isokinetic migration model,which makes up for the disadvantages of the two models.(3)Drive the smart car along the lane center line to realize lane keeping function.Lane first build roads needed to keep the control system model,car-the relative position and lateral movement controller,secondly design way to adapt to the border in the vehicles are driven in a structured way,and make improvement on APF algorithm to adapt to the function of lane keep in smart car,and then improve the APF algorithm controller is designed,and finally make the smart car simulation under different conditions,through the contrast analysis of the advantages of different control algorithms.