Identification Of Important Intersections And Prediction Of Traffic State Of Urban Road Network

Aiming at solving the present problem as low efficiency and lack of precision of prediction methods of traffic state of road network,this paper puts forward a kind of prediction method based on the important intersections and mainly analyzes the evaluation of intersection importance,the identification of important intersections and prediction of road network traffic state,respectively.The main contents are as follows:1.A method of evaluation of intersection importance based on the decision on multiple attributes is presented.Firstly,this paper analyzes indicators of intersection importance evaluation in traditional traffic engineering theory and in complex network theory.Then,it establishes a comprehensive evaluation model by the introduction of analytical hierarchy process and computes the final value standing for the importance of each intersection,by which the evaluation is implemented.2.A method of evaluation of intersection importance based on the IPR algorithm is presented.Firstly,this paper introduces the PageRank sorting method used in the internet,and tells the implementation principle in detail.Then,it analyzes the defects when this algorithm is applied in the evaluation of intersection importance,and improves it with the combination with traffic dynamic characteristics.Finally,it uses the improved IPR algorithm to evaluate the intersection importance.3.A method of identification of the range of important intersections is given.Firstly,this paper introduces the importance vector and integrates different evaluation results as components.Then,it categorizes the whole intersections using cluster analysis method,thus selecting out all the important intersection in the road network.4.A method of prediction of traffic state of road network is presented.Firstly,this paper predicts the traffic state of important intersections with markov process method.Then,with the fuzzy comprehensive evaluation method and the combination with the weight of different intersection types,it implements the prediction process with high efficiency and accuracy.