Analysis And Prediction Of The Time And Space Distribution Characteristics Of Cable Cars Passenger Flow In Mountain Scenic Area

The research of passenger flow is an important topic in the current scenic area management research,which involves many problems such as passenger flow distribution analysis,passenger volume prediction and so on.The time distribution trend and spatial distribution characteristics of passenger flow play a vital role in the management and resource scheduling of scenic area.The result of passenger flow analysis and prediction is the basis and foundation for scenic area managers to make decisions and provide tourism services.For mountain type scenic area,because the space position of the cable car is adjacent to the main scenic spots and main entrance in the scenic area,the cable car passenger flow has high research value.In view of the unique spatial characteristics of mountain scenic area,this dissertation combines the spatial and temporal distribution characteristics of the cable car passenger flow,and then puts forward the analysis method of the spatial and temporal distribution characteristics of the cable car passenger flow in mountain scenic area and the prediction method of time sharing cable car passenger flow.The main research work are as follows:(1)In this dissertation,the spatial and temporal distribution characteristic analysis method of time-sharing cable car passenger flow based on clustering is proposed,and the time distribution of passenger flow in each space location is analyzed by peak index.The time distribution trend of passenger flow per hour in each place is obtained by hour,and K-means clustering is used to group the space sites,in order to put locations in a group in which the distribution trend of passenger flow is similar in time.The spatial distribution characteristics of passenger flow in each period are obtained,and then the spatial and temporal distribution of passenger flow is summed up.Considering the change of passenger flow caused by the promotion activities in the scenic area,the concept of "Panholiday" is introduced in this dissertation,and the distribution rules of passenger flow under different circumstances are found,and the case of Huangshan scenic area is analyzed as a case.(2)In order to solve the effects of sudden situation and untimely management scheduling,this dissertation constructs the Mountain Type Spacial-Temporal Artificial Neural Network(MT-STANN),and proposes a prediction method of time-sharing passenger flow based on space-time neural network.K-means spatial clustering is carried out for spatial locations with different spatial and temporal distribution characteristic,and MT-STANN prediction models are constructed for different groups of spatial locations.On the basis of Back propagation neural network(BP),the nested spatial weight matrix is constructed,and the neural network model is initialized with this matrix.The input and output of the network are determined according to the results of spatiotemporal feature analysis,and then the MT-STANN network is trained and constructed.In this dissertation,the prediction results of ANN,MT-STANN and Support vector machine(SVM)model are compared.It is found that the MT-STANN model has better fitting results and better prediction accuracy.The experiment shows that the MT-STANN prediction model can effectively predict the time-sharing data of the passenger flow in mountain scenic area,and the forecasting method of time-sharing passenger flow based on space-time neural network is convenient and helpful to the managers' decision and scheduling,and can also reduce the impact of sudden events and delay scheduling.The results have both theoretical and practical significance.