Functional Architecture Construction in the Road Emergency Scene of Autonomous Transportation System Based on Text Clustering

The autonomous transportation system (ATS) is the next development direction of intelligent transportation. To construct the functional architecture of ATS scientifically, an algorithm of functional architecture is presented. First, the road emergency scene is selected, and the functions are analyzed in this scene by the text frequency, the text vector space, and the Density Peak Clustering algorithm. Second, the Silhouette Coefficient is applied to evaluate the clustering results, which determines the definition of multi-layer structures for the functional architecture of ATS, namely “autonomous perception, autonomous learning, autonomous decision, and autonomous response.” Finally, the integrated analysis results of the case study are displayed. The mean of the clustering coefficient is 0.78. The results show that the proposed algorithm could promote the construction of functional architecture and improve the development of the autonomous transportation.