Prediction Of Traffic Flow Based On Cloud Model In Application Of Intelligent Tourism System

With the appearance and development of cloud model (a model for transformationbetween qualitative concept and quantitative value), it has been widely used in manyaspects, such as graphic processing, intelligent traffic, data mining and intelligenttourism. Also the vigorous development of tourism causes big problems for traffic. Howto combine traffic forecasting and tourism closely is the target and direction of thispaper.On one hand, a model for forecasting of short traffic flow base on cloud model wasproposed in this paper. With the historical and real-time data, the mean error of cloudmodel was4.8%, this result showed effectiveness of the forecasting mission. In thepaper, BP neural network was used to forecast the short traffic flow and its mean errorwas15.4%, which was much worse than that of cloud model. After that, with the resultsof the forecasting of short traffic flow, two-dimensional cloud model was used to realizethe signal controller for the intersection, and the optimized time of the intersectiontraffic light was displayed dynamically on the map.On the other hand, for the purpose of using the forecasting of short time trafficflow based on cloud model and the optimized model for the intersection, an intelligenttourism system for intelligent tourism has been constructed. Some basic functions ofmap interaction, such as Zooming in, Zooming out, Birds eye were realized. Besides,some other functions for tourism, such as, querying transportation information (buses,routes, stops and transfers), or the geographical position of the surrounding situation ina short distance, were also developed. For the purpose of tourism functions, the roadnetwork topology was built and the Dijkstra algorithm was developed to provide arecommended best-route for tourists. As a result, this system could provide referenceand support for travelling, and help them to reduce travel time and decrease travelexpense.