Driving Rule Acquisition and Decision Algorithm to Unmanned Vehicle in Urban Traffic

Lane-changing decision of multi-source information on the urban traffic environment is the key technology to unmanned vehicles achieve actual road driving, to extract the driver's lane-changing decision rules in the complex and dynamic environment, firstly, PreScan software was used and virtual urban traffic environment was built, 6-DOF vehicle dynamics model was based on the Simulink, the decision rules of driver lane-changing behavior was extracted through rough set. The results show that the relative speed of the experimental vehicle and leading vehicle maintains at around 4~7 m/s, and when the space distance between adjacent cars reaches 20~35 m, the driver begins to implement lane-changing, the results provide driving knowledge for unmanned vehicles online machine learning and theoretical basis for the depth study of lane-changing behavior of uncertain decision-making.