From: Simon S. <sim...@ar...> - 2010-12-14 14:43:28
|
Dear Hanyang, a student and I developed an activity based demand generation program for SUMO last summer. You can find it in the SUMO repository. Our program has no graphical user interface; it reads the description of the population from a configuration file. On the other hand, it performs several kinds of activities and includes public transportation. In addition, it encodes the trip id in a way, that the car can be identified. So, you can follow the trips of one specific car over several days (it remembers the home and work position for every person/car). We decided to split the input into the algorithm from the user interface, as this is the common pattern behind the SUMO tools and as there are several graphical front ends for SUMO out there. I will offer a combination of a osm/net file and a population description file for two cities on the wiki (probably Januar next year). This makes it easier to start with it. Maybe we could cooperate on this topic (and fuse both projects). I am going to work on the code in Januar and February in any case. Sincerely, Simon Am Montag, den 13.12.2010, 04:22 +0000 schrieb Hanyang Zhang: > Hi There, > We are a research team at the University of Wollongong, NSW, Australia. We are working on a traffic simulation project using SUMO. We have created a tool to generate trips files from .osm files, and we would like to know if you are interested in accepting this tool as a part of your project. Our eventual goal is to generate a reallistic vehicular simulation environment for SUMO for use with Qualnet. > > The following paragraphs illustrate how our algorithms work. > > The basic idea is that OSM maps store the location information (e.g. geo coordinates of shops, museums, cinemas etc), and we can use it to create simulated traffic demand. For a given time slot, there are a certain amount of vehicles to emit. For each vehicle, we first decide its starting street using the location information (e.g. desity of residential buildings), and then specify the distance that the vehicle is going to travel with normal distribution, and at last we choose a destination based on the desity of commercial or industrial buildings. The tool is currently written in C#. > > The tool first extracts all locations stored within the map, and then classifies them into groups (e.g. shops, schools, banks, etc). For each group, the user needs to assign 3 factors to it: a living factor, indicating how many people live at the location; a working factor, indicating how many people work at the location; and shopping factor, which indicates the density of commercial outlets. These locations are then assigned to the nearest street (edge), and therefore each edge has 3 factors: working factor (the sum of working factors of all locations assigned to this edge), living factor and shopping factor. Then 6 types of traffic are generated, working area -> living area, living area -> shopping area, and so on. > > The higher the living factor of a street is, the more likely a trip from this street will occur. On the other hand, the higher the working factor of a street is, the more likely a commute trip will end at that street. > A day is divided into 24 hours, and the tool needs the poportion of trips emitted within each hour. I also set these up by hand, for example, at 8 am many people are going to work, and at 12pm many people go to food tores for food. > > The user also need to specify the average distance of the trips. For example, in Sydney the average commute distance is about 12km, so I use normal distribution N(12,2.4) to specify the distance of a trip. > > Once the trip file is created, we can use duarouter to generate the routes. > > Kind regards, > > Hanyang Zhang > ------------------------------------------------------------------------------ > Oracle to DB2 Conversion Guide: Learn learn about native support for PL/SQL, > new data types, scalar functions, improved concurrency, built-in packages, > OCI, SQL*Plus, data movement tools, best practices and more. > http://p.sf.net/sfu/oracle-sfdev2dev > _______________________________________________ > sumo-devel mailing list > sum...@li... > https://lists.sourceforge.net/lists/listinfo/sumo-devel |