Unmanned driving method and system

The invention relates to the technical field of data processing, in particular to an unmanned driving method and system. The method comprises the following steps: firstly, obtaining vehicle position configuration characteristics and half-travel vehicle position configuration characteristics based on distribution characteristics of lane-changing failure vehicle distances of vehicles in second historical data of the lane-changing failure vehicle distances, and obtaining lane-changing failure atypical degrees of the vehicles according to similar conditions of the vehicle position configuration characteristics of different vehicles; and obtaining a lane changing failure descriptor according to the lane changing failure duration and the lane changing failure atypical degree of the vehicle, obtaining sample distances according to the lane changing failure descriptor and the vehicle position configuration characteristics, and clustering the sample distances to obtain different lane changing failure scene categories.