| A large number of interactive real-world traffic simulations can give decision makers great supports and references during making traffic plans and constructions. However, some appropriate data is necessary when making simulations. How to get the data cheaply, conveniently, comprehensively and how to extract the inner content such as traffic model, as the input of simulation process with a right method are big problems needed to be solved. This paper proposes a method for traffic simulation based on sensing data, mainly using human electronic footprint data together with bigdata simulation platform to extract a traffic model. Here is the work that completed.First, a method for extracting traffic model and traffic parameters with human electronic footprint data is proposed. The details will be descripted including the calculation of traffic speed and flow, vehicle trajectory, OD matrix, travel plans and so on.Second, the architecture of bigdata simulation platform has been analyzed, including needs analysis, detailed design. The traffic simulation experiment is mainly carried out on this platform.Third, the experiment on traffic simulation using matsim together with the bigdata simulation platform has been completed. This paper pre-processes the human electronic footprint data and GIS data, forming networks, buildings, travel plans and count data,then making them as the input of matsim, and finally getting the visual results.Human electronic footprint data are always acquired by existing public infrastructures, so the data contains not only a high density of information about human behaviors, as well as high-coverage, low-cost and easily accessible features. Using this data and relying on the bigdata simulation platform can make the simulation process simpler, make the simulation results more complete and accurate, and make the simulation executors get rid of the work on preliminary data acquisition and pretreatment, which makes them more focused on traffic simulation itself. |
PDF Full Text Request