This paper selected five driving school coaches with different driving experience and gender as real vehicle test drivers,four typical car steering conditions test data with four different vehicles is collected.And this paper use principal component analysis and clustering analysis to classify and identify drivers intention.The paper analyzes the advantages and disadvantages of the three different clustering methods in the direction of drivers steering intention,and it can be analyzed that the Gath-Geva algorithm is more advanta-geous than the fuzzy C-means algorithm and Gustafson-Kessel algorithm by clustering center data analysis and clustering objective evalua-tion index analysis.