Taxi is one of the main carriers of urban transportation, it is of great significance to improve its dynamic and economy performance by matching the parameters of vehicle powertrain system and optimizing energy management based on the construction of representative driving cycle, which can reflect the driving characteristics of taxi. Considering the long operating time and wide operating range of taxi, the test data of two taxis for one week was collected in this study. Firstly, Fourier filter was used to preprocess the test data. Through processing of taxi operation data, the difference of passenger carrying time and mileage between the weekdays and weekends, along with the travel law of the peak and off-peak periods in Xi'an were obtained. Then divided the test data into short strokes, extracted the characteristic parameters related to velocity, acceleration and time of each short strokes, reduced the dimension by principal component for the characteristic parameters and then used cluster analysis. Finally, constructed the driving cycle of Xi'an taxi by Markov-Monte Carlo simulation.

Driving Cycle