文件列表:

BIVEC2019WH_v2.pdf (628704, 2019-06-05)
FIGURES/ (0, 2019-06-05)
FIGURES/Animation.PNG (28232, 2019-06-05)
FIGURES/Comparison.PNG (57405, 2019-06-05)
FIGURES/Flow_Observations.PNG (65657, 2019-06-05)
FIGURES/Flow_Simulation.PNG (44590, 2019-06-05)
FIGURES/NoData.PNG (30320, 2019-06-05)
FIGURES/NoMoreDetToDel.PNG (4767, 2019-06-05)
FIGURES/Observations.PNG (63898, 2019-06-05)
FIGURES/ObservationsMain.PNG (41305, 2019-06-05)
FIGURES/ObservationsOff.PNG (34283, 2019-06-05)
FIGURES/ObservationsOn.PNG (34466, 2019-06-05)
FIGURES/RemoveDetector.PNG (5615, 2019-06-05)
FIGURES/RemoveLocation7.PNG (15121, 2019-06-05)
FIGURES/SelectCorridor.PNG (7709, 2019-06-05)
FIGURES/SelectDataAtDetector.PNG (25026, 2019-06-05)
FIGURES/SelectLinkComp.PNG (40503, 2019-06-05)
FIGURES/SelectLinks.PNG (5838, 2019-06-05)
FIGURES/SelectPoints.PNG (5893, 2019-06-05)
FIGURES/SelectTime.PNG (16475, 2019-06-05)
FIGURES/Speed_Observations.PNG (63840, 2019-06-05)
FIGURES/Speed_Simulation.PNG (59493, 2019-06-05)
FIGURES/TravelTimeComp.PNG (24389, 2019-06-05)
MATLAB/ (0, 2019-06-05)
MATLAB/CREATEMODEL.m (5328, 2019-06-05)
MATLAB/LOGGHEvvc.m (5582, 2019-06-05)
MATLAB/LTM_SC_v3.m (5663, 2019-06-05)
MATLAB/LTM_SC_v3_mex.mexw64 (295936, 2019-06-05)
MATLAB/VISUALIZECORRIDOR.m (5783, 2019-06-05)
MATLAB/VISUALIZECORRIDOR_NoDB.m (4812, 2019-06-05)
MATLAB/add_virtual_detector.m (2675, 2019-06-05)
MATLAB/aggregateLanes.m (1393, 2019-06-05)
MATLAB/animateSimulation.m (2998, 2019-06-05)
MATLAB/buildFullOD.m (2854, 2019-06-05)
MATLAB/buildODmatrix.m (2339, 2019-06-05)
MATLAB/buildTFmatrix.m (2388, 2019-06-05)
MATLAB/collectData.m (4540, 2019-06-05)
MATLAB/collectData_noDB.m (5747, 2019-06-05)
MATLAB/connectServices.m (1430, 2019-06-05)
MATLAB/createNetworkModel.m (5148, 2019-06-05)
... ...

# OpenTrafficCenter A basic implementation of a traffic monitoring and control center of the Flemish Highway system in Matlab developed by the **L-Mob Research Center** at the **KULeuven** ## Introduction The application allows users to visualize traffic data and create a simple traffic model for any corridor along the highways of Flanders (Belgium). The application can be easily extended to visualize the impact of different traffic management scenarios such as the opening of a peak hour lane. The traffic data is provided by the open data platform of Flanders. The application uses live traffic data observed by double loop detectors along highway roads. It is externally repackaged and made available on http://www.itscrealab.be as a direct download or through Matlab's SQL interface. The road network was created by combining a reference map of Flanders (GRB) and extracts from https://www.openstreetmap.org. The application is built in two phases 1) Select and visualize data along a selected corridor 2) Setup and run a dynamic traffic model ## Detailed description ### Select and visualize data along a selected corridor Run the main file **VISUALIZECORRIDOR.m** or **VISUALIZECORRIDOR_noDB.m** to start the application. It is recommended to run this file section by section following the inline instructions, the outprints displayed in the command window or the text inside pop-up figures. The general steps of phase 1 are: 1) Load the network into Matlab 2) Select a corridor (route between 2 points on the network) 3) Select a time window for the observations 4) Create a network model (add on/off-ramps to the selected corridor) and collect the data 5) Visualize traffic state on the corridor in a space time diagram 6) Visualize the on-ramps 7) Visualize the off-ramps ### Setup and run a dynamic traffic model After collecting and visualizing the traffic data a simple dynamic traffic model can be built using the CREATEMODEL.m file. It is recommended to run this file section by section following the inline instructions, the printouts displayed in the command window or the text inside pop-up figures. The traffic simulation is based first order kinematic wave theory of a single commodity flow. The application implements the Link Transmission Model of the KULeuven (https://www.mech.kuleuven.be/en/cib/traffic/downloads). The general steps of phase 2 are: 1) Derive Demand matrices and turning fractions 2) Set simulation time interval 3) Set standard parameters of the traffic model 4) Refine the capacity at specific locations (bottlenecks) 5) Run the simulation 6) Visualize the simulated traffic state on the corridor in a space time diagram 7) Compare simulated travel times with observed travel times 8) Animate the result 9) Visualize flows and counts for observed roads ## Example: some notes ### Data gathering Extract the csv data files and place them in your working directory allong with all the matlab files. ### Main file Run the file VISUALIZECORRIDOR_NoDB.m in matlab. For more control Run this file section by section. ```MATLAB VISUALIZECORRIDOR_NoDB ``` ### Select a corridor - When prompted to select points select **From a List**.  - Pick the first corridor (**E314 Leuven Lummen**)  ### Select a time window The supplied data is gathered on 11th of December 2018. Pick the appropriate time for the evening peak (**15:00** to **20:00**). Data for different dates can be found on http://db.itscrealab.be/download/loop_detectors/  ### Identifying bad detector locations - Scroll through the data detector location by location and observe that for **location #7** no data is available  - Close the window with the data and select **From a List** when prompted to remove a detector  - Select detector location #7 named **Complex nr 20 ? Wilsele ? De Vunt** and click on **ok**  - No other detectors should be removed  ### Visualizing the processed data Next 6 figures are plotted of the selected corridor. 1) An overiew of the corridor on the map. Click on the blue diamont points to get an overview of the data for the detectors at that specific location  2) A smoothed space-time diagram of speeds allong the corridor. This is an excellent way of visualizing the congestion patterns allong the corridor. The black horizontal lines are detector locations. Inbetween detector locations data is interpolated using a filtering technique that takes into account spatio-temporal relations in traffic (according to Treiber-Helbing 2002).  3) A smoothed space-time diagram of flows allong the corridor.  4) An interactive overview of all detectors on the main corridor  5) An interactive overview of all detectors on the on-ramps  6) An interactive overview of all detectors on the off-ramps  ### Create demand Now open the file CREATEMODEL.m. All of the data is processed and is used in this m-file to create a model representation of the selected corridor. For more control Run this file section by section. ```MATLAB CREATEMODEL ``` The model uses the detector counts to create demand at each origin (source) and splitting rates at every diverge node in the network. ### Setting the standard values of the supply The model recuires additional attributes for every link. Based on the number of lanes, a standard capacity and jam density is assigned to each link (line 28 & 29 in the code). The maximum spillback speed is derived from these values to conform the assumption of a triangular diagram. Additionally you can set the capacity of specific links manually (for example by looking at capacity outflow in the data). ```MATLAB capacity_per_lane = 2100; kjam_per_lane = 100; ``` When promped to change the capacity of link click on **No** such that only the standard values are used  ### Running the simulation and inspecting the result After the model is finished different different figures are opened to visualize the result: 1) On top is an animation of the density in the network over time. Press space-bar to start the animation. If you close the figure the animation is stopped and the next figure is highlighted  2.a) An overview of detector locations in the network. Zoom into a specific zone, hit any key on the keyboard and click on a link to visualize the difference between the observations and simulation at a specific location.  2.b) A new figure will be opened for each link that you click  3) A comparison between simulated travel times and observed travel times allong the corridor  4) A smoothed space-time diagram of flows allong the corridor.  5) A smoothed space-time diagram of speeds allong the corridor.  ## Outlook The presented example shows a reasonable fit to the data with standard model parameters. An even better fit is possible by tweaking the capacity of the bottleneck or the standard values (like jam density) of different highway sections. The interaction rules at merges for redistributing the available downstream space is are rather simplistic and based on entry capacities just as described by the general node model (Tampère, Corthout, Cattrysse & Immers 2010). On highways traffic often clears the right lane at on-ramps such that this capacity based priority distribution isn't accurate. It is recommended to set appropriate on-ramp capacities to account for this. The model can now be used the evaluate scenarios like the opening of peak hour lanes downstream of the current bottleneck. In the illustrated case such measures would reduce congestion substantially. The performance of the corridor can be evaluated and compared for different days to show variability in bottleneck throughput. Note that the throughput at weaving sections can be dependent on the amount of the competing flows which is often variable over the peak period making the capacity of such sections variable over time (the current implementation is limited to a fixed capacity value over the simulation interval). Demand variations can also be observed for different days to compute elasticity or responsiveness. As such these tools are usefull for gaining insights in the behavioral response in case of accidents or road works.

近期下载者：

相关文件：

评论：[我要评论] [举报此文件]

收藏者：