Research Of Nanchang Intelligent Traffic Flow Prediction Algorithm

A lot of traffic flow data in urban traffic information contains the traffic flow phenomena and rules,and it can affect the city traffic condition synthetically,scientifically. It’s the realization of scientific traffic management, formulate superstitious essential,and being the base of the decision-making of traffic.It is very valuable and important information resources, also is one of the main reference for the residents travel to know road.Due to the urban traffic flow information contains a large number of excavated effective traffic data information,can play a critical role for the actual traffic flow forecasting,so take what method to excavate its potential value can make use of the information is one of the hot spot of current research.In this paper, based on the practical history data of east lake area of Nanchang city to process traffic flow prediction algorithm research.Based on the existing traffic stream datum to handle and analysis,based on the understanding and acquaintance of the forecast algorithm to explore and research traffic stream forecast.Concrete research plan is to clean the data of traffic flow that i have got,then applicate program to statistics one day’s the traffic flow value of take 5 minutes as the unit of a street,then take the clustering algorithm of K- means algorithm to divid the data is into four types according to the flow value, and then in each class, according to the size of the time to get divid the flow values into 2, 3, 3, 2 classes.All of the classes a total of 10 categories, and then for each class, using the lesser support vector machine forecasting model for various periods to achieve atraffic forecast.This paper also has carried on the single most theravada support vector machine forecasting model to predict, and then comparedthe the result with the result of the combined algorithm model, and prove combination algorithm is scientific effectiveness.