Spatio-temporal Analysis And Cost Modeling Of Trip Data Based On License Plate Recognition Big Data

License plate recognition data is massive,with high coverage and space-time characteristics.From the license plate recognition data,the travel information of most vehicles traveling on the urban road network can be obtained,so that urban traffic managers can better understand the vehicles travel laws and the concentration in various regions,and put forward more targeted management measures and reasonable car development strategies to alleviate the pressure both on traffic congestion and the ecological environment from a large number of motor vehicles,large urban population,and lack of land resources.This can effectively promote the construction of public transit metropolis and realize the sustainable development of urban transportation.This paper proposes a technical process based on license plate recognition data processing: First,preprocess the license plate recognition data,determine the travel time threshold T of the vehicle by the difference method and the interval method,split the vehicle travel trajectory,and split the travel trajectory into several sub-trajectories to obtain the single trip information of the vehicle;.Second,analyze the travel time characteristics,travel frequency,travel distance,travel time,etc.of local cars,online ride-hailing and nonlocal cars.Third,streets as the unit,perspective from traffic count and traffic flow,using the attraction method and greatest advantage flow method to study the concentration level of local cars,online ride-hailing and nonlocal cars.Fourth,divided car travel costs into personal costs and social costs based on the current domestic and international transportation cost estimation theories and methods,including vehicle use cost,travel time cost,traveler’s physical loss cost,road facility cost,traffic congestion cost,environmental pollution cost and traffic accident cost,and then according the car travel characteristics to improve those 7 parts cost measurement methods,and a car travel cost estimation model was established.Finally,taking Shenzhen as an example,estimating the travel costs of local cars and online ride-hailing,the result shows that the personal cost of the total travel costs forlocal cars and online ride-hailing in Shenzhen is relatively close in 2018.