Research On Forecasting Method Of Xi’An Road Short-time Traffic Flow Based On Data Mining

At present,the traffic system in xi ’an is becoming increasingly large and complex,and the inefficient road traffic system undoubtedly aggravates the traffic tension.The urban traffic problem has become an important factor that hinders the development of xi ’an and restricts the urban economic construction.Prediction in xi ‘an main road traffic flow for short periods of time,is one of the effective method to solve the problem of urban traffic,for vehicles to reduce traffic accidents,ease traffic congestion and improve traffic efficiency,and for the design of the traffic police department sign and marking,traffic lights time Settings,the arrangement of the traffic police and so on all has the very vital significance.This paper takes the short-term traffic flow prediction of the main road in xi ’an as the research object,and on the basis of combing the research status of the traffic flow at home and abroad,the principle of traffic flow prediction and the commonly used algorithm,analyzes the current situation and existing problems of the traffic flow prediction in xi ’an.On this basis,the short-term traffic flow prediction model based on SVM was constructed and implemented.The short-term traffic flow prediction model was constructed by means of SVM(support vector machine).The historical data of the traffic flow of the main road in xi ’an was collected.The short-term traffic flow prediction model based on KNN is constructed and implemented.Research of input flow,speed and share pretreatment method,analysis of freeway traffic flow time characteristic,in view of the dynamic similarity,periodic,and time correlation of traffic flow,process principle and algorithm of KNN algorithm analysis and design,is suitable for xi ’an trunk road traffic flow state of short-term traffic forecasting model.From the perspective of prediction effect,SVM and KNN can meet the requirements of short-time traffic prediction of the main road in xi ’an,and have high prediction accuracy,showing the potential of SVM and KNN in short-time traffic prediction.For the main road section of xi ’an,these two methods can be used to predict,improve the traffic condition of the busy section and avoid the peak traffic.It can be seen from the prediction effect and simulation performance indexes that the short-term traffic flow prediction method proposed in this paper based on SVM and KNN is of high prediction accuracy,which can meet the basic requirements of short-term traffic flow prediction of the main road in xi ’an.It can be predicted that the short-term traffic flow prediction of the main road in xi ’an through the method and model designed in this paper can improve the traffic efficiency,which has important reference value for the design of traffic police department signs and markers,the setting of traffic light time,and the arrangement of traffic police force.