Font Size: a A A

Study On Passenger Traffic Prediction And Integrated Hub Allocation Of Economic Areas

Posted on:2012-11-07Degree:DoctorType:Dissertation
Country:ChinaCandidate:L ChenFull Text:PDF
GTID:1222330365471225Subject:Transportation planning and management
Abstract/Summary:PDF Full Text Request
Economic areas is a necessary stage of the economic development and a important organization for the modern economic development. It’s also a embodying vital force of the innovativeness. With the rapid growth in traffic demand brought by the rapid development of our country’s economic areas, because of the differences of demand characteristics between the inter-city and general city, the system configuration based on the traditional single city has objectively limited the expanding of the space structure of the economic areas. So the paper focus on the optimal and allocation’s study of inter-city transportation network. The research methods are the combination theory research and empirical research, the combination qualitative and quantitative method, the combination systematic study and layered study. The key research problems of research paper includes passenger traffic volume forecast of the inter-city, the selection, optimal and allocation of the transportation, the choice of the important node in economic zone, and the layout and planning model of integrated traffic terminals.The paper develops the study according to the following logical orders:Based on defined economic areas,the traffic system and other relevant concepts.this paper analysis the differences of the traffic network’s optimal and allocation between economic areas and the city, studies the important impact of the space structure to the traffic planning, also has a deep discussion on the restraining effect between the traffic planning and the resource allocation.In order to forecast passenger traffic within the metropolitan region, it proposed a prediction based on model support vector machine. Model data set would be divided into training set, test set and examine set, the determination of the loss function parameters and penalty factor was followed to the final prediction error criterion, and made the ε-insensitive loss function as the loss function, select the Gaussian kernel function into prediction, and when changing parameter value gradually, it would determined different parameters. When the model was used in metropolitan areas, predictive value can be well fitted with the actual. The application results showed that the model is more suitable for metropolitan passenger transport predictions than traditional methods, and it would be provide decision support for the metropolitan transportation planning. Regarding to the selection of transportation, through analyzing the each kind of transportation’s technical characteristic, the paper determines the principle of utility maximization, establishes the bi-level logit model that has been put into actual application in the economic areas. The application results showed that the error is acceptable statistically, and the bi-level logit model is reasonable.Based on analyzing the existence problem in the traditional research about the Layout and planning, he bi-level model based on the stochastic user equilibrium which has been established can able to reflect the interaction of supply and demand, it’s also different with the tradition static model. Considered the difference of the cost’s value between the understanding level and the actual, through designing the SUE(Stochastic User Equilibrium) between different transport mode of the inter-city, this paper establishes the layout and planning model of integrated traffic terminals which is more practical to the actual.,it"s also called bi-level model based on the stochastic user equilibrium.The rationality and validity. of the model has been confirmed by the example application. The paper designs the solution algorithm of the bi-level model based on the stochastic user equilibrium, and uses the genetic simulated annealing algorithm in the top level, but for the lower level, the F-W algorithm is used.
Keywords/Search Tags:traffic engineering, support vector machine, passenger traffic volume forecast, transportation network, bi-level model
PDF Full Text Request
Related items