Urban Road Traffic Status Analysis And Time Series Prediction Research

In the era of rapid development of science and technology,people rely on transportation for all activities.In order to improve travel efficiency,people constantly improve the utilization rate of transportation,but traffic congestion has become a pain point of travel.Traffic congestion not only increases people’s travel time,but also aggravates air pollution,which in turn affects people’s mood and work efficiency.Therefore,how to avoid traffic congestion has become a hot issue in the field of traffic control.In this paper,a traffic state prediction model is established for the traffic congestion problem.Firstly,the improved fuzzy C-means(FCM)algorithm is used to classify the traffic state and determine the classification standard of the traffic state;then the improved long and short-term memory network algorithm is used to predict the average speed of the road section;finally,the prediction results are clustered to obtain the predicted traffic state,and the performance of the model is analyzed by experimental comparison.(1)Processing and analysis of floating vehicle data.Aiming at the defects of floating vehicle data,this paper interpolates the missing data,deduplication and replaces the anomalous data,extracts the data characteristics,and then calculates the required traffic parameters.Temporal and spatial correlation analysis of traffic parameters is performed to improve prediction accuracy.(2)Improvements were made to the traffic state division algorithm.In this paper,the floating car dataset is processed,and the average speed and velocity variance of the road section are obtained as the characteristic parameters of the division.Referring to the Chinese urban road traffic congestion scale,the traffic state is divided into five categories,and the traffic state is divided by FCM algorithm and improved FCM algorithm,and the results of the two algorithms are compared.(3)Predict the average speed of the road segment.Firstly,the long and short time memory network(LSTM)algorithm is used to predict the average speed of the road section,and then the LSTM algorithm is improved by moth-flame optimization(MFO)algorithm,so that it can quickly find the optimal combination of parameters.According to data analysis,the input is determined to be 5*3 speed matrix(the average speed of the road segment at the first five moments of the current road section,upstream road section and downstream road section),and the output is the average speed of the current road segment at the next moment.(4)A traffic state prediction model is constructed.Firstly,the improved FCM algorithm is used to divide the traffic state,and the result is used as the traffic state division standard,and then the MFO-LSTM algorithm is used to predict the road section speed,and finally the traffic state is clustered according to the prediction result,and compared with the traffic state division standard,so as to realize the division and prediction of the traffic state.