Research On The Factors Of Traffic Congestion In Beijing And Solutions

In recent years,Beijing traffic congestion turns to be more and more serious,thus causing complex and critical environmental pollution and bringing huge influence onto the life and work of urban residents.Consequently,many researchers or scholars are keen on the issue of traffic congestion,and wish to release the pressure on road traffic.Based on these fundamental principles,we treat Beijing city as the case study,illustrating the situations,characters,and factors of Beijing traffic jam.In conclusion,this thesis constructs novel models for the traffic flow prediction,these results can provide reference values for the Beijing traffic.First of all,in this thesis,we collect many case studies from other successful cities who have experience in the traffic governance.These successful cases offer many useful and accessible suggestions and policies,which can be generalized to the traffic situation in Beijing.After that,we crawled from official website the real-time dynamic data regarding Beijing traffic,which is periodically refreshed every five minutes.Then,the traffic features on the time,space,and intensity are attained from the analysis of the data.Moreover,according to the situations and features in Beijing traffic,interacted and influenced factors are collected,and the decision tree model is built to evaluate the importance of above factors,thus offering a couple of suggestions to improve the traffic congestion happened in Beijing city.In addition,the prediction of traffic flow is implemented by the models of Recurrent Neural Networks and Long-short Time Memory Networks,which can predict the future situation based on current situation of traffic flow,thus helping urban residents to plan their schedules and routines in advance.The purpose of this is that when the congestion influencing factors cannot be solved temporarily and the car ownership cannot be reduced,the traffic flow parameters can be predicted to achieve the role of diversion,thereby alleviating the traffic congestion.Finally,three conclusions are drawn from the above experiments: firstly,there exist notable features of traffic congestion in Beijing on time,space,and intensity,and the imbalance is one of the most significant characters;secondly,by inspecting in the empirical analysis of the decision tree modeling on traffic congestion factors in Beijing city,eight factors,including travel time,road length,the number of lanes,area population,bus stop density,intersection density,the number of traffic lights,disposable income per capita,are affects traffic congestion more than other factors,in addition,travel time is the most significant factor.Moreover,these factors can be adjusted to efficiently alleviate the situation of traffic congestion;thirdly,the most important point is the short-term prediction of the two traffic flow parameters,the traffic congestion index and the average speed,predict the changes in the traffic congestion index and the average speed in a short period of time in the future,in order to understand the future traffic congestion in advance,it can ease traffic congestion.