Extracting trip demand for static demand responsive transit service based on taxi trips

Static demand-responsive transit (SDRT) is an innovative service model, which provides public transport services for passengers who have similar origin and destination locations, and departure time for a certain period at generally higher price. This paper proposes a method to extract potential trip demand for SDRT from taxi order data. Passenger OD flow similarity is analysed from the perspective of spatial and geometric features, including the parameters of spatial geography, trip distance, and trip angle. OD flow similarity function is constructed, which serves as the clustering criterion within a K-nearest neighbour agglomerative hierarchical clustering framework. Thus we may locate the potential passenger OD demand to direct the efficient design of SDRT route. A case study is conducted based on taxi order data collected in Suzhou, China over one day, where the proposed method manages to extract the potential passenger demand of 22 clusters for profitable SDRT.