Building the exclusive driving cycle for an electric bus (EB) route is necessary for achieving accurate customized services and providing data services for operation. In this paper, we first preprocess the GPS data collected from an EB route in Meihekou and divide it into 2,845 short trips. Next, principal component analysis is used to make dimensionality reduction for feature parameters of short trips. Third, with multi-dimensional principal component factors as input, the K-mean clustering algorithm is used to select appropriate short trips for constructing an exclusive driving cycle for the EB route. Finally, to verify the necessity and effectiveness of the driving cycle constructed in this paper, we compared it with the China Automotive Test Cycle for urban bus (CHTC-B) and measured data. Results show that the exclusive driving cycle is closer to the operating characteristics of the actual status.