The Characteristics Of Online Car-hailing Travel And Demand Forecast Based On Didi Data

The analysis of residents’ travel behavior is a very important basic work in urban comprehensive transportation system planning and urban construction planning,and it is also an effective basis for formulating transportation policies.Therefore,it is very important to study the characteristics of urban residents’ travel activities.Based on the GPS data of car-hailing in Xi’an,this paper studies the law of time and space characteristics of car-hailing trips,and then performs data mining on hotspot areas for passengers to make it more convenient for car-hailing drivers to find passenger sources.It provides transportation planning and management for government departments.Provide data support.First,based on the GPS data pre-processing of the car-hailing,the C# programming language is used in the My SQL database to extract the time and location data of the car-hailing,and then the total travel volume at different times,the travel volume at each time period,and the morning and evening travel Perform statistical analysis on residents’ travel laws during peak hours and travel time.Secondly,tap the online hotspots for passengers and passengers.According to the spatial distribution characteristics of online car-hailing demand,the spatial distribution of online hot-spot areas of online car-hailing passengers in different time periods during the working day is analyzed,and DBSCAN spatial clustering is performed on the location information of online car-hailing passengers at different time periods.Calculate the particle coordinates in the hotspot area for passengers,and further analyze the hotspot area for passengers traveling.Third,research on short-term forecasting of residents’ travel volume,taking travel demand as the target of short-term forecasting of car-hailing travel demand,and performing correlation analysis on the factors that influence the forecast,and building an improved GA-BP based on the characteristics of demand forecast The neural network model makes short-term predictions for residents’ travel,and carries out error analysis to verify the validity of the model.Finally,some government supervision and management suggestions for car-hailing are given,and the future research direction and content are prospected.