Study On The Methods Of Multi-scale Prediction For Expressway

In recent years, Traffic congestion is serious, traffic congestion with the system timedelay, energy waste and environment pollution, traffic accidents and other problems alsowill increase, and the mouth of the serious influence to the people often travel. Trafficcongestion has become the world urgently needs to solve traffic problems.For expressway, with the popularity of private cars, traffic jam on the expressway isincreasing, timely, accurately identify the road traffic incidents and heavy traffic has beena focus and difficult point in research of traffic condition monitoring, and with theincrease of traffic load and traffic forecasting has become one of hotspots in the field ofdynamic traffic management research, the emergence of a large number of results.However, restricted by the traffic data acquisition and processing method, the trafficcondition prediction research in terms of efficiency, effectiveness and economy there arebig room to improve.Expressway is an important transportation link between cities, large capacity, longdistance fast transportation services for the city, has an important traffic collection ofhydrophobic function. Expressway plays an important role in the urban traffic, is thefocus of traffic management. Traffic parameter data to predict information have differentapplication way, application of different way for information data to predict the specificneeds of traffic parameters is not the same. To forecast the traffic parameter data when animportant question to consider is to predict the time scale, therefore, different time scaleswere expressway traffic parameters short-term multi-step prediction has high value ofresearch and application.Through the above analysis shows that the expressway multiple time scale trafficparameters prediction has a very important role, this paper selected topic has higheracademic significance and practical value. In this paper, based on pipe project inshandong province expressway intelligent transportation information platform is the keytechnology research and application demonstration"(issue of), multiple time scales to theexpressway operation situation forecast were studied. The main research results of thispaper specifically embodied in the following aspects: 1) Put forward the expressway running state predict method for determining thescaleShandong high-speed through the historical data of space and the time characteristicanalysis of specific data, combining forecasting principle and the method to determinethe scale of, is presented for the shandong expressway running state prediction ofshort-term, medium-term and long-term prediction of scale.2) Based on kalman filter prediction method of short-term traffic flow parametersFirst of all, studied the kalman filter for traffic parameters prediction method, basedon kalman filtering, short-term traffic parameter prediction model is established;Short-term operation, then discriminant expressway in order to accurately analyzed theexpressway traffic running state division method; And, in the example of a road inshandong expressway traffic flow, for example, has carried on the short-term forecastanalysis.3) Established the expressway running situation of medium and long-term forecastmethodAnalysis on characteristics of expressway traffic flow parameters, determines therunning situation of the expressway, long-term prediction, then combined with the trafficparameters change characteristics were established based on the weighted average of themedium-term prediction method research of traffic flow parameters and traffic flowparameters of the long-term forecast method based on exponential smoothing; Finally,analyzes instances of these two methods respectively, the results show that the predictionmodel has good application value.