Research On Urban Road Traffic Operation State Prediction And Evaluation Method Based On Spatiotemporal Characteristics

Making accurate and effective predictions and evaluations of urban road traffic operation status is the basis for fully grasping the behavior of the road traffic system,rationally formulating traffic strategies,and effectively realizing the potential of the traffic infrastructure.The key to intelligent transportation system technologies such as operational efficiency.The key to the prediction of urban road traffic operating state is the prediction of traffic parameter data,and the traffic parameter data is actually a sequence of time and space data.Therefore,the research on the traffic operating state based on the characteristics of traffic data needs to consider both the temporal and spatial characteristics of the traffic parameters.In this paper,based on the characteristics of traffic flow data and the characteristics of spatio-temporal operation,the method of predicting and evaluating the traffic operation state considering the spatio-temporal characteristics of traffic parameters will be studied.The main research contents of the paper are as follows:(1)Analysis of the spatiotemporal characteristics of traffic parameters.For the preprocessed traffic parameter data,the time chart analysis,Pearson correlation analysis,Moran index analysis,mean standard deviation analysis and other methods were used to explore and analyze the temporal and spatial characteristics of traffic parameters.The evaluation model provides support.(2)Short-term prediction of traffic parameters.In view of the limitations of traditional data analysis methods that cannot meet the needs of analyzing various types of data,and in order to reflect the spatio-temporal characteristics of traffic parameter data,this paper proposes the LSTM-RBF prediction model.Through the comparison of LSTM examples,the optimal prediction interval is determined.The RBF method is used for example prediction and compared with the single LSTM and RBF prediction methods,which verifies the rationality and scientificity of the prediction method in this paper and the effectiveness of the single model.(3)Evaluation of urban road traffic operation status.Considering the uncertainty of traffic data,the ambiguity of the traffic state,the complexity of the traffic system,and the impracticality of the unified threshold,this paper proposes an evaluation method of urban road traffic operation status based on FCM and cloud model Cluster evaluation,fuzzy comprehensive evaluation and other commonly used traffic state evaluation methods are compared and analyzed as examples to verify the rationality and effectiveness of the evaluation method in this paper.(4)Case application analysis.Combined with the method of this paper,the traffic operation status prediction and evaluation of the case road network is carried out.Based on this,the traffic improvement is carried out.The operation efficiency of the road network after the improvement of the operation comparison is improved compared with the original.The application value of efficiency.