Analysis Of Urban Road Traffic Operation Based On Multi-source Data

With the continuous expansion of China’s urban population and the continuous improvement of people’s income,the number of motor vehicles has increased rapidly.The low-speed growth of road resources has been unable to meet the rapid growth of people’s travel demand.Traffic congestion has become a common problem faced by major cities.The construction of intelligent transportation system has become the first choice for major cities to control congestion.The existing traffic operation analysis system in Urumqi uses taxi GPS data to judge the state of road congestion,but the coverage of GPS data in the periphery of the city is poor due to the driving behavior of taxi drivers.The bus has a fixed driving line and driving time,which can cover the central and peripheral areas of the city,and can provide rich GPS data.If used,it can make up for the problem of insufficient coverage of taxi data.Therefore,this paper proposes to use the difference and complementarity of taxi and bus in urban traffic spatial distribution,improve the accuracy of road traffic flow parameter estimation through data fusion technology,and expand the data coverage.On this basis,the urban traffic operation is analyzed.This paper mainly studies from four aspects: multi-source data processing,fusion model construction and comparison,road network coverage analysis before and after data fusion and traffic operation analysis index construction.For the original taxi and bus GPS data,through data preprocessing,map matching,speed calculation and data noise reduction steps to obtain reliable GPS speed data.The data fusion model and multiple linear regression model based on BP(Back Propagation)neural network are constructed on the basis of data processing,and the model with higher fusion accuracy is selected by comparing the evaluation indexes.In order to determine the effective expansion of data coverage after multi-source data fusion,the road network coverage before and after data fusion is analyzed.Finally,based on the data fusion model,combined with the speed threshold of Urumqi road congestion level,the traffic operation index index is constructed.The results show that :(1)The fusion accuracy of multi-source data fusion model and multiple linear regression model based on BP neural network reaches 95.66 % and 92.27 %,respectively,which can better fuse multi-source data.The fusion effect of multi-source data fusion model based on BP neural network is better than that of multiple linear regression model.(2)Approximately a third of existing road conditions can be enhanced by integrating transit data.By integrating bus data during the daytime,the number of road sections covered by about 7 % can be increased on the basis of the proportion of existing taxi sections.(3)Traffic operation analysis index can be better applied to the analysis of urban road traffic operation in Urumqi.Through the analysis of the operation status of road sections and regions,it is shown that the operation status of road sections and regions is good,and there is no medium congestion or serious congestion.During the morning and evening peak periods,it is in a mild congestion state,and the rest time is basically smooth.The operation of road sections and regions conforms to the traffic operation law.