Research On Speed Continuous Distribution And Visualization Based On Multiple-source Urban Traffic Big Data

Speed is an important indicator for evaluating the road conditions,which is not only the precondition of traffic planning,traffic design and traffic management,but also has an important impact on traffic safety.Most speed distribution studies are based on experimental data of spot speed in different roads.From the point of view of big data,a research method of speed continuous distribution based on multiple-source urban traffic big data is proposed,and the feasibility of the method through GPS data is validated.Firstly,the importance of grasping the speed distribution is explained,and the research status at home and abroad is introduced.The main research contents are defined,and the research methods of speed distribution and the data types of common multiplesource urban traffic big data are summarized.Secondly,a research method of speed distribution based on GPS data is proposed.The speed distribution type and speed distribution characteristics were studied respectively by GPS data of taxi and floating car.Through Stabilized Probability Plot and single-sample K-S test,the speed distribution types and characteristics of different grades and types of roads were obtained.Finally,using the Amap open platform to produce vehicle speed visualization.The results show that,the speed of freeways generally follows the normal distribution.When the data is skewed,log-normal distribution is used to better fit.The speed of urban expressway follows normal distribution.The speed of main road generally follows the normal distribution.When the data is skewed,the weibull distribution is used to better fit.The discreteness of speeds of expressways,main roads and highways is reduced in turn.The traffic disturbance has greater influence on the 15% speed than the average Speed and 85% speed.Ground roads with viaducts are more discrete than ground roads without viaducts.