Is the cut-in behavior in china dangerous? Research and analysis based on Chinese driving behavior characteristics in highway

Cut-in scenarios are a common and complex traffic scenario faced by autonomous vehicles in the future, but there are few studies on cut-in scenarios. Based on the naturalistic driving databases we constructed, this paper extracts 3,044 Chinese highway cut-in scenarios and uses them for cluster analysis to study different cut-in characteristics. In the 11 kinds of cut-in scenarios obtained by cluster analysis, four kinds of critical cut-in scenarios and seven typical cut-in scenarios are established by analyzing the minimum TTC and the lateral acceleration in different cut-in processes. At the same time, considering the impact of special areas on the cut-in behavior, the scenario distribution of different special areas in 11 kinds of cut-in scenarios is analyzed, and according to clustering similarity, the most critical and the most typical cut-in scenarios distribution characteristics for different special areas are obtained. These characteristics can indicate the risk issues in Cut-in scenarios, and provide a reference for the algorithm design of autonomous vehicles.