Research On The Key Techniques Of Urban Road Traffic Congestion Identification

The urban road traffic congestion becomes one of main social questions for each big city in the world, not only has brought many inconvenient for people's routine work and life, will also restrict the economical growth, and accelerate urban environment worsening and so on, serious influence city sustainable developments. Timely and accurately finding traffic congestion has took place (or soon occurs) in the road network, which has the important meaning regarding the formulation reasonable effective traffic congestion management strategy.Urban road traffic congestion state identification is a very complicated process, which includes many contents and technologies. But the paper mainly focuses on the following three key technologies: traffic information collection technology, traffic data forecasting technology, and traffic congestion state distinction algorithm.First of all, the paper studies the definition, classification, and related properties of urban road traffic congestion, and analyzes the deep level reasons from the economic point of view.Second, this paper analyzes and compares to various traffic information collection means. It studies the arrangement of each kind of traffic detector, of which gives out the general layout step. In addition, the paper proposes one kind of multi-objective programming model in the disposition density optimization method of stationary traffic detector.Then, in the short time forecast aspect of traffic flow parameters, three forecast techniques is proposed: the improvement exponential smoothing forecast technique, based on the BP neural network's forecast technique, as well as the muiti-model fusion forecast technique. And when calculates the weight, proposes the determination length of stride smallest absolute value law.Finally, the paper carries on the summary and the appraisal to classical and improved urban road traffic state identification algorithm, gives out the working steps of recognition algorithm. Through the re-application of the identification results of individual algorithms, it designs the object-oriented road traffic congestion identification algorithm voting fusion.