| As tourism has become a pillar industry of the global economic system,tourism transportation has become a hot research topic.At the same time,car-hailing has become an important mode of urban transportation.In urban tourism transportation,there is a mismatch between the supply and demand of car-hailing in some tourist attractions.In order to solve this problem,this study realizes the extraction of car-hailing orders in scenic spots and conducts the analysis of the spatial and temporal characteristics of them based on multi-source data,and puts forward the classification method of car-hailing dependence in scenic spots.On this basis,a multi-scenic spot car-hailing demand prediction model is constructed.The main contents and conclusions include the following aspects:(1)The first part of this research is the analysis of spatial and temporal distribution characteristics of car-hailing order extraction in scenic spots.This part extracts and counts the car-hailing data in scenic spots based on the car-hailing data and the Beijing tourist scenic spot data,and then analyzes the spatiotemporal characteristics of the scenic spot’s car-hailing.Firstly,the time distribution characteristics of car-hailing travel demand in scenic spots found that the car-hailing in scenic spots is obviously periodic and the time factor has a great influence.Secondly,in terms of spatial distribution characteristics,the study found that the overall spatial distribution of car-hailing in scenic spots is very unbalanced,and the level of scenic spots and the geographical location of scenic spots will significantly affect the distribution of car-hailing.Finally,comparing the travel demand of car-hailing in scenic spots under different weather conditions,it is found that severe weather has a greater impact on the travel of car-hailing.the analysis of spatial and temporal distribution characteristics provides a basis for the analysis of scenic spot dependence and the selection of prediction model input.(2)The second part of this study establishes a classification method for the dependence of car-hailing in scenic spots.Based on the analysis of the spatial and temporal characteristics of car-hailing travel in scenic spots,this part clarifies the concept of dependence on car-hailing in scenic spots.The index system of car-hailing dependence in scenic spots is established,and a classification method of car-hailing dependence in scenic spots is proposed,and this method is used to classify the main scenic spots in Beijing into high,medium and low according to the dependence of car-hailing.The study found that high-dependency scenic spots are mainly popular scenic spots located in the popular scenic spots from the third ring road to the fifth ring road in urban areas;medium-dependency scenic spots include the popular scenic spots in the third ring road and the main scenic spots in urban areas;low-dependency scenic spots are mainly include natural scenic spots far from urban areas.The dependency classification provides support for the construction of feature set of the demand prediction and the proposal of optimization strategies for each scenic spot with different dependency level.(3)The last part of this study establishes a multi-scenic car-hailing travel demand forecasting model.Aiming at the problem of multi-scenic car-hailing travel demand forecasting,the Long Short-Term Memory(LSTM)model is used to process the time variable characteristics of the scenic car-hailing,and the Convolutional Neural Network(CNN)is used to realize the fusion of the spatial characteristics of the scenic spots and weather factors.This paper uses Bayesian optimization(BO)for neural network parameter selection,and proposes BO-LSTM-CNN for multi-scenic car-hailing travel demand prediction.Next,the demand forecast of some scenic spots in Beijing is carried out,and a variety of performance evaluation indicators are selected for evaluation.The results verify the validity and applicability of the model method proposed in this paper.Finally,an optimization strategy of car-hailing service in scenic spots based on short-term demand forecast is proposed for different scenic spots.Based on multi-source data,this study establishes a classification method for the dependence of car-hailing in scenic spots and a multi-scenic spot car-hailing demand prediction model,which is meaningful for understanding the current car-hailing situation of Beijing’s major scenic area and for the subsequent improvement of car-hailing service level in scenic spots. |
PDF Full Text Request