Research On Steering Control Method Of Autonomous Vehicle

With the development of science and technology and people's living standard continues to improve,more and more families have their own cars,the number of vehicle increase sharply,causing traffic environment worsening.and the consciousness of the driver determines the running condition of the vehicle.Mature autonomous vehicle control system is more reliable than the uneven driving technology,it can effectively relieve traffic pressure and reduce driver workload and improve road utilization.The steering control of autonomous vehicle plays an important role in the whole autonomous vehicle technology.A good steering control system can make the vehicle to be able to operate normally when turning.In this paper,the lateral control system of autonomous vehicle is studied.According to the actual situation of the road,a steering controller is designed,which can turn well when turning.The map of autonomous vehicle,in order to help autonomous vehicle can safely and smoothly drive on the road int the premise to comply with traffic regulations.in the electronic map building of autonomous vehicle,will refine the connection between the roads,the primary mission is build a connection relationship between traffic lane,it must be legal,include building change lanes point for different traffic lane of same segment,at the turn of the road junction,building turnin point for different traffic lane of different segment and building point for same lane of same of same segment.After the establishment of the relationship between the road points,according to Dijkstra algorithm,a program for the shortest path planning for autonomous vehicle is developed by Visual Studio programming software.For the coordinate information in the map,current civil coordinate data in the map and the coordinates of longitude and latitude of uncoverted WGS84 can not be used for the data of the real time attitude calculation,in this paper,the way point in the WGS84 coordinate system is transformed into a plane coordinate which is suitable for autonomous vehicle.After the establishment of the electronic map of the autonomous vehicle,the steering control system of the autonomous vehicle is based on the way points,When the autonomous car changes lanes or straight driving,By calculating the lateral deviation and angular deviation of the distance between the target point and the autonomous vehicle.lateral deviation and angular deviation are used as input to thelateral controller,the decision error is made in order to prevent lateral deviation and angular deviation data beyond the physical domain during steering,the lateral deviation and angle deviation quantization factor module is designed in the lateral control system,the value is limited to the fuzzy theory,the decision error is avoided in some special road segment due to data beyond the domain,input to lateral controller,output angle by scale factor in the end;When turning or turning around at the intersection,the road point is converted into the coordinate value of the vehicle motion coordinate system to calculate the front wheel angle.In this paper,the vehicle model is established in CarSim,the experimental section include straight line,change lane,turning and turning around,in Simulink,linear and lane changing module,turning and turning around module,system control center module are bulit,the output angle is introduced into CarSim,the simulation results show the effectiveness and reliability of the lateral control system designed in this paper.