Abstract. Rapid increase of private cars leading to serious road congestion and air pollution problems, carpool has become an important choice for people to travel. In order to improve the service quality of carpool, trajectory matching is becoming a hot topic. However,traditional path matching algorithm based on Hausdorff distance has exposed two problems: (2) Ignoring waiting time of general users; (2) The direct calculation of the Hausdorff distance of the entire path cannot reflect the impact of special sections on the matching metric. To solve the above mentioned problems, this paper proposed an improved algorithm: (1) Hausdorff Distance is combined with turning point; (2) Turning point is taken as an effect way to segment trajectories. Compared to the classical algorithm MCT (Minimum completion time online mode scheduling algorithm), proposed algorithm get better performance.