Research On Path Optimization For Road Transportation Of Dangerous Goods Based On Trajectory Data

Various dangerous goods such as oil and natural gas have become important materials to support economic development and service to the people's livelihood.However,dangerous goods often cause serious accidents once the dangerous situation occurs in the transportation process.Reasonable path optimization and selection of dangerous goods road transportation plays an important role in avoiding transportation risks.Path optimization research based on risk assessment and risk measurement has been relatively complete,but the data involved in the risk measurement model are often difficult to obtain comprehensively,what's worse,the data volume is large and difficult to update.With the development of Intelligent connected vehicles and the deep integration between big data technology and transportation,a large number of emerging information technology tools,such as satellite positioning system and GIS geographic information system,are being continuously applied to the management of dangerous goods transportation.Nowadays,the application of the satellite positioning system and the construction of the transportation safety supervision system have accumulated a large amount of data.How to effectively integrate these data and information technologies into the dangerous goods road transportation service is one of the hot topics of research.Due to the special transportation field and relatively independent management,the trajectory data of dangerous goods vehicles are not widely used in existing research.It is conducive for dangerous goods transportation vehicles to explore the lmovement law and integrate the experience of all dangerous goods transportation vehicles in the whole network,which is also one of the important ways to balance transportation risks and benefits.Therefore,taking Dalian as an example,this paper attempts to analyze the characteristics of transportation behavior by mining the trajectory data of dangerous goods transportation vehicles,and then integrate it into the construction of an empirical safety layered road network,where the optimal path can be obtained.Firstly,this paper introduces the general situation of the research area and the sources of various data,and elaborates the pretreatment methods for the problems existing in the track data of dangerous goods transportation vehicles,which is convenient for further extraction of subsequent transportation features.Secondly,through the massive trajectory data of dangerous goods vehicles,the spatial and temporal characteristics of its operation are extracted and analyzed,and the group characteristics of dangerous goods vehicles in Dalian are described,including three aspects,i.e.the transport time characteristics,origin-destination points and transportation path spatial distribution,and the characteristics of the empirical sections..Thirdly,on the basis of the characteristics of the empirical sections,combined with the mobile risk of dangerous goods,an empirical safety layered road network index system for the transportation of dangerous goods is constructed,and then the hierarchical road network is classified by probabilistic neural network method.Fourthly,an improved A*algorithm considering accident risk is proposed,and the route optimization of dangerous goods transportation is implemented by combining the hierarchical searching strategy.The feasibility and effectiveness of this method are verified by comparing the obtained transportation scheme with the actual transportation scheme through the examples.