Research On Forecasting Method Of Short-Term Tourist Flow Considering Spatial Corelation

With the rapid development of the world economy,China's tourism industry has been expanding rapidly in the past decades.Although the rapid development of tourism has brought huge economic benefits,the surge in tourist flow in scenic spots has caused a series of challenges to the natural and social environment.Accurate short-term estimation and prediction of tourist flow can strengthen the management ability of scenic spot,minimize the risk resulted from the surge in tourist flow,and ensure the sustainable development of the scenic spot.Therefore,it is of great significance to establish a scientific and accurate short-term tourist flow forecasting model to achieve accurate short-term tourist flow forecasting.Short-term tourist flow has complicated characteristics such as non-linearity and randomness,which are greatly influenced by short-term factors such as weather and temperature.At the same time,due to the spatial correlation,the correlation among scenic spots can be really complicated.Artificial intelligence model has an excellent ability to deal with non-linearity,which provides a feasible solution for complex short-term tourist flow forecasting.On the basis of studying the related theories of spatial correlation among scenic spots,this paper constructs a scenic spot spatial correlation model,and proposes a short-term tourist flow forecasting model considering the spatial relationship among scenic spots based on the scenic spot spatial correlation model.The main contents of this paper are as follows:(1)Summarize the researches related to tourist flow forecasting.This paper summarizes current researches in terms of tourist flow forecasting from three aspects:spatial correlation among scenic spots,influencing factors and forecasting models.We summarize the related theories of spatial correlation among scenic spots from four aspects:radiation theory,symbiosis theory,competition theory and tourism route organization theory,and argue the necessity to establish a short-term tourist flow forecasting model considering the spatial relationships around the scenic spot.(2)Construct the spatial correlation model of scenic spots.This paper explains the concepts of spatial association of scenic spots,defines the quantification model of spatial association among scenic spots,and analyses the influencing factors of tourist flow among scenic spots from three aspects:accessibility of scenic spots,location relationship of scenic spots and attribute relationship of scenic spots.Based on the data of tourist flow among scenic spots,a model to figure out the relationships among these influencing factors and tourist flow of scenic spots is constructed by using logistic regression.(3)Propose a tourist flow forecasting method considering the spatial correlation of scenic spots.This paper analyses the influencing factors of short-term tourist flow forecasting and combines the influence caused by the surrounding scenic spots with the traditional short-term tourist flow forecasting model.Establish a short-term tourist flow forecasting model taking into account the spatial correlation of scenic spots and using LSTM model.(4)Verify the accuracy of the spatial correlation model and the tourist flow forecasting model.This paper takes 27 scenic spots above 3A level in Nanjing as the research objects.The experiment demonstrates that the spatial correlation model has an admirable accuracy and the tourist flow forecasting model considering the spatial correlation among the scenic spots has a better result than the traditional forecasting model.