Research On Taxi Travel Demand In Taxi Pick-up Hotspots Areas Of Beijing

Recent years,the economy develop rapidly.The travel activities of residents’become more complex and the travel demand grows up substantially.As an important part of public transportation,taxi plays an important role in satisfying the residents’short trip demand,increasing the diversity of travel modes and promoting urban economic development.However,the randomness of travel about residents leads to the spatiotemporally unbalanced distribution of demand,the blindness of vacant taxis when looking for the passengers causes higher empty-load ratio.Thus,the contradiction between supply and demand of taxi becomes increasingly prominent.This paper analyzed the spatial-temporal distribution of taxi travel demand and excavated the taxi pick-up hotspots by the taxi GPS data of Beijing,then analyzed the influence of multiple factors and presented the taxi travel demands forecasting model based on the pick-up hotspots areas.Firstly,based on the taxi GPS data preprocessing,the map-matching algorithm is selected to correct the taxi track points which deviate from the actual road network.Then,extracted the taxi off-board data by programming and calculated the scope of the study area total taxi travel demand in a certain period of time,and then,analyzed the spatiotemporal distribution characteristics of taxi travel demand on working day and rest day.Secondly,based on the spatiotemporal distribution characteristics of taxi travel demand,the working days is divided into three periods:morning peak,evening peak and night peak.The Density-Based Spatial Clustering of Applications with Noise algorithm for excavate and analysis of taxi pick-up hotspots areas at different periods of working day,the ArcGIS tool is used to realize the visualization.And then,the spatiotemporal distribution characteristics of taxi travel demand in taxi pick-up hotspots areas is analyzed.Finally,analyzed the influencing factors of taxi travel demand based on the spatiotemporal characteristics of taxi travel demand in taxi pick-up hotspot areas.Based on the accumulation of historical taxi demand and forecasted the demand on future period of typical pick-up hotspots in Beijing.A local weighted regression model,genetic algorithm optimization of neural network model and combination forecasting model is applied in the field of taxi demand forecasting.The model evaluation index is selected to evaluate the prediction effect of the three prediction models.This paper proved the validity and applicability of the model through forecasting the typical taxi pick-up hotspots in Beijing.Mining the spatio-temporal characteristics of taxi travel demand and taxi pick-up hotspot,forecasting the demand of taxi can provide support for taxi scheduling,reduce the vacancy rate of taxis and increase the travel cost of residents.And then,it can reduces air pollution,fuel waste and traffic congestion problem,alleviate the imbalance between supply and demand of taxis.