A new idea was presented to deal with the nonlinear and time-varying characteristics of the intelligent vehicle (IV). It searched a mapping relationship between the input state space and output state space based on the vehicle running data. Furthermore the IV transverse locomotion pattern space was constructed. The construction method was deduced based on fuzzy c-means (FCM) by the well adaptive abilities of the fuzzy logic to deal with the nonlinear, time-varying and uncertainty problems. At last the method was simulated. The simulation result shows that the method can cluster the sampled data set effectively. Clustering set is clear, and the subspace is divided distinctly. It provides a theoretical basis for constructing IV locomotion pattern space