Research On Urban Hot Sections And Hotspots Extraction And Application

With the rapid development of China's urbanization,the problem of city's traffic congestion is brought due to the rapid growth of the urban population and motor vehicles quantity.Therefore,this paper studies the urban hot sections and the resident trip hotspot areas based on the taxi GNSS data to solve the problems caused by city's traffic congestion.The main research contents are emphasized in the following aspects:(1)The method of spatial correlation analysis was included to extract and analyze the traffic hot sections.Firstly,the global Moran's I index in each period's was calculated based on the spatial weight matrix and the traffic data of each section.Then,the hot sections of Dalian urban district city center in working's day and weekend were extracted respectively based on the local correlation statistics.Finally,the traffic status of the road sections was verified and simulated by using VISSIM software.The travel speed,the travel time delay and the travel time delay were taken as the judgment index.The discriminate matrix of three state evaluation sets was constructed and the weight matrix of three indexes was obtained.(2)The method of data field clustering was used to extract the daily hotspot and analyze the spatial and urban residents' temporal characteristics.The data field influence factor was selected by using sigma experience selection method.The hot areas' clustering centers were obtained by using the data field hierarchical spectrum method.And the taxi data drop-off points were clustered according to the subordinate relationship between the data points to the extract the resident trip hot areas.The hotspots were extracted,and the temporal and spatial distribution characteristics of the residents' daily travel time and hotspot were analyzed.And the K-means clustering method in SPSS was used to verify the hot spot extraction results based on data field clustering.(3)The taxi emergency dispatch points were configured based on the urban hot areas which have extracted.According to the travel speed and the length of each road between the hot areas,the road network structure topological map was constructed.The minimum time-distance matrix between urban hot areas was obtained by using Dijkstra algorithm.According to the principle that the time which the taxi in dispatching point traverses the hot areas were the shortest the taxi emergency dispatching point configuration model based on urban hot area was established.Finally,the taxi emergency dispatch points' urban hot area was established by using the genetic algorithms in Matlab.In this paper,the method of spatial correlation analysis and data field clustering were used to accurately extract the hot sections and hotspots in cities,and solve the problem that the parameters of a single clustering algorithm choose difficultly.Secondly,the extraction result of hot section could be verified by constructing discrimination model,which got the hot section affiliation degree of verified hot section was 0.512.The extraction result of hotspot was verified by K-means clustering in SPSS software.Finally,according to the extracted hotspots,the genetic algorithm was used to configure the taxi emergency scheduling points,which ensure that any hotspot area could be dispatched to a taxi within 3.225 minutes.The study results could not only provide a better theory and method for the residents' travel analysis,but also have important significance for the city's traffic guidance,traffic management and location-based services.