A Stop Planning Method over Big Traffic Data for Airport Shuttle Bus

With the growing volume of the airport passengers, public transit is needed for healthy and sustainable city development, in which airport shuttle buses play a key role in satisfying the demand. In this paper, a two-phase airport shuttle bus stop planning method is proposed based on taxi GPS data. It aims at providing convenient public transit to the airport by identifying optimal airport shuttle bus stop. In our method the first phase focuses on filtering the irrelevant "dirty" records. Then the remained data set is divided into two parts consisting of origin dataset and destination dataset. In the second phase, the k-means clustering algorithm is employed to identify representative points as candidate airport shuttle bus stops based on these two datasets. After that, taking advantage of traffic model and rules defined in traffic engineering, the candidate stops set can be further optimized. Finally, extensive experiments are conducted on a large-scale real-world taxi GPS data set to verify the practicality of our method on an In-memory database platform, HANA.