Research On Forecast Method Of Intelligent Passenger Flow In Tourist Attractions

In recent years,With the rapid growth of economy,tour industry becomes pillar industry of the economy.Tourism flow forecasting is a crucial issue in the industry and generally seen to be one of the most complex function of tourism management.Especially with the implementation of the new official holiday regulation in China,the number of tourists in holidays has increased rapidly.The huge market has been boosting by the business of holiday tourism which make government have a headache.While at the same time tourism flow forecasting attracted the attention of academia and industry.Tourism flow is influenced by many factors like season,the human body comfort,the type of a week,holiday,special factor and economic.These factors make the tourism flow severely imbalance situation,so accurate tourism flow forecasting is a most important and difficult issue in tourism industry.With the accurate forecasted trends and patterns that indicate the sizes of future international tourism flow,the government and private sectors can have a well-organized tourism strategy and provide a better infrastructure to serve the visitors and develop a suitable marketing strategy to gain benefit from the growing tourism.Here a new intelligent method MSFNN for tourism flow prediction is proposed which combine statistical methods with BPNN model.This method is built on an analysis of the competition of the influence factors,volatility characteristics and a large number of literature.This paper has done the following work.Frist,Summarized the characteristics of different forecasting methods.Analysis many forecasting methods' characteristic like time series method ARIMA,support vector regression(SVR),grey prediction theory,Artificial Neural Network(ANN)and Deep Learning.Second,processing and analyzing multi-factor and scale of the scenic spots.This paper combine qualitative and quantitative methods to analysis the relationship between comfortable degree of human body,type of a week,tourism season and special factor with future forecasting day.According to historical passenger flow fluctuation,the passenger flow is divided into year,season,type of week three scale.Third,Special factor prediction(Holidays and Festivals)under the influence of the event.In order to avoid the effect of the same special factor event is inconsistent on Gregorian calendar and lunar calendar,this paper using fluctuation coefficient to predict the special factor event tourism flow and improve the universal usability of the model.Last,build the model of MSFNN.MSFNN is depend on Spearman,Monthly average method,Linear interpolation method,fluctuation coefficient and neural network to achieved the forecasting function.In this literature the historic daily tourism flow of Xi'an museum is used to confirm this method.After comparison whit ARIMA and single BPNN(only considering a single influence factor),the experimental results indicate that the proposed method is an effective approach which is more accurate than any other models and can be used as a suitable forecasting tool to solve the problem of scenic spot forecasting.