User Guide

DLSIM Users Guide

Working Document Version 1.0

Please feel free to send any questions, feedback, and corrections to Dr. Cafer Avci (cafer.avci@aalto.fi) or Dr. Xuesong (Simon) Zhou (xzhou74@asu.edu) by adding comments in this document.

Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.3 or any later version published by the Free Software Foundation; with no Invariant Sections, no Front-Cover Texts, and no Back-Cover Texts. A copy of the license is included in http://www.gnu.org/licenses/fdl.htmlwww.gnu.org/licenses/fdl.html

Introduction

1.1. Motivation

Motivated by a wide range of transportation network analysis needs, static traffic assignment (STA) and dynamic traffic assignment (DTA) models have been increasingly recognized as a set of important tools for assessing operational performances of those applications at different spatial resolutions (e.g., network, corridor and individual segment levels) and across various analysis temporal regimes (e.g., peak hours, entire day and second-by-second). The mathematical modeling and related volume-delay functions are described in Appendix.

The advances of STA and DTA are built upon the capabilities of integrated flow assignment and simulation models in describing the formation, propagation, and dissipation of traffic congestion in a transportation network.

Bridging the gap from macroscopic static assignment to mesoscopic dynamic assignment

Planning practitioners have recognized the full potential of DTA modeling methodologies that describe the propagation and dissipation of system congestion with time-dependent trip demands in a transportation network. In April 2009, the TRB Network Modeling Committee conducted a DTA user survey through the FHWA TMIP mail list, which identified the following top 5 technical barriers:

DTA requires more data than are available or accessible to most users (47%)
Setting up a DTA model consumed inordinate resource (44%)
Cost/benefit of implementation is unclear (45%)
DTA tools take too long to run (35%)
The underlying modeling approaches are not transparent (35%)

The development goal of DTALite aims to provide an integrated open-source package for strategic traffic analysis that includes both static traffic assignment and dynamic traffic simulation to reflect the impact of road capacity constraints. The underlying volume-delay models include BPR functions and its extension of BPR-X. Three traffic stream models, namely, point queue model, spatial queue model and simplified kinematic wave models, are embedded in the mesoscopic simulator to describe queueing behavior at bottlenecks with tight capacity constraints.

Adopting open network standard of GMNS

General Travel Network Format Specification is a product of Zephyr Foundation, which aims to advance the field through flexible and efficient support, education, guidance, encouragement, and incubation. Further
details can be found in

https://zephyrtransport.org/projects/2-network-standard-and-tools/
Integrated graphic user interface and analysis package

NeXTA (Network eXplorer for Traffic Analysis) is another open-source graphic user interface (GUI) for transportation network analysis, while the lower-case “e” stands for education with broader impacts. With both open-source traffic assignment/simulation engine (as a simple Windows console application) and graphic user interface, the software suite of DTALite + NeXTA aims to

provide an open-source code base to enable transportation researchers and software developers to expand its range of Strategic Traffic Assignment capabilities to various traffic management analysis applications.
present results to other users by visualizing traffic flow dynamics and traveler route choice behavior in an integrated 2D environment.
provide a free education tool for students to understand the complex decision-making process in transportation planning and optimization processes.

parallel computing on shared memory multi-core computer

Emerging multi-core computer processor techniques are offering unprecedented available parallel computing power, on most of laptops and desktops currently available in the market. To exploit this paradigm change in computing, we will require a new software architecture and algorithm design so as to facilitate the most efficient use of emergent parallel hardware.
Integrated signal timing optimization (to be added)
Integrated OD demand estimation through path flow estimator (to be added)

The latest software release can be downloaded at our Github website. The source code can be downloaded at https://github.com/asu-trans-ai-lab/DLSIM. Table 1 illustrates the contents of different folders at Github https://github.com/asu-trans-ai-lab/DLSIM.

Table 1. contents of folders at Github.

Github Folder Name	Contents
Src	source code of DLSIM
test	a simple working dataset for console application DLSIMand visualizer of NeXTA.
Doc	user’s guide and other documentations for DTALite
Data	sample datasets for DLSIM: 1. two_corridor 2. Braess’s_paradox 3. three_corridor 4. Sious_Falls 5. Chicago_sketch 6. Tempe ASU network

1.2. System Architecture

The software architecture of DTALite aims to integrate many rich modeling and visualization capabilities into an open-source traffic assignment model suitable for practical everyday use within the context of an entire large-scale metropolitan area network. Using a modularized design, the open-source suite of** simulation engine + visualization interface** can also serve future needs by enabling transportation researchers and software developers to continue to build upon and expand its range of capabilities. The streamlined data flow from static traffic assignment models can allow state DOTs and regional MPOs to rapidly apply the advanced STA/DTA methodology, and further examine the effectiveness of traffic mobility, reliability and safety improvement strategies, individually and in combination, for a large-scale regional network, a subarea or a corridor.

Figure 1.1 Software System Architecture

The components and different modules in the system are listed as following:

a. Network Data includes two essential files, node.csv and link.csv for the macroccopic network representation.

b. OD Demand Meta Database includes the setting.csv as the configuration file that describes information such as agent type, demand period, demand file list, which help users to represent the OD demand information for different user types at specific demand periods.

**c. Traffic Assignment Module **includes the key steps of the assignment, including the BPR Volume Delay Function, Shortest Path Tree Generation, and Flow Assignment, which generates the path flow and link flow according to the UE principle.

d. NEXTA: Visualization Interface Module is able to visualize the network and the output of traffic assignment, including Static Link Performance and Agent Trajectory.

e. Space-Time Simulation Module utilizes the path flow output of Traffic Assignment Module to perform Space-Time Simulation, while the underlying traffic flow models in the Space-Time Simulation Module are Point Queue (PQ) and Spatial Queue (SQ). A simplified kinematic wave (KW) model can be also used in an advanced mode, similar to DTALite.

f. Capacity Management aims to manage the static and time-dependent link capacity input for Space-Time Simulation, such as signal timing plans and multi-modal service plans.

g. Simulation Output Module covers the output file of Space-Time Simulation Module, including Dynamic Link Performance and Agent Trajectory in terms of link_performance.csv and agent.csv, which can be visualized in NeXTA.

Regarding parameters in settings.csv, Table 2 illustrates the differences between two key steps of Static Traffic Assignment and DTA + space-time simulation.

Table 2. The differences between Static Traffic Assignment and DTA+ space-time simulation

	Static Traffic Assignment	Dynamic Traffic Assignment + space-time simulation
Assignment_mode in settings.csv	UE	DTA
Travel time evaluation	BPR function with volume/capacity ratio (soft capacity constraints)	Space-time network based simulation with tight capacity constraints
Demand input	OD demand	OD demand or agent based
Output (1): link performance	Static Link performance VOC, volume, delay	Dynamic Link performance Queue, delay at time t
Output (2): path/agent data	Path flow for OD and k-paths, based on path pool based gradient projection methods	Individual agent trajectory with path sequence and time sequence

1.3. 5 steps of performing traffic analysis using CSV files

The specific instruction for the use of NeXTA and DLSIMis as follows:

Step 0: [Download and locate the project folder] Download and unzip the release software package from github. Locate DLSIM file folder with node.csv, link csv, demand.csv and settings.csv. Typically, copy DTALite.exe and NeXTA.exe in the same folder for easy access.

Step 1: [Check input files in Excel] Open a file explorer, view or edit input files of node, link and demand csv files, in Excel or any text editor. Review and change the configuration in settings.csv in Excel.

Step 2: [Visualize and validate network in NeXTA] Click “NeXTA”—“File”—“Open Traffic Network Project” to choose the node.csv file in your network data set. Check the network connectivity through a simple path calculation by selecting one OD pair.

Step 3: [Run DLSIM as a Windows console application] Click on the executable of “DLSIM.exe” from a file explorer or run it from Windows command window, to perform traffic assignment and simulation. The output of this Windows console applications is displayed in screen and log file DLSIM_log.txt.

Step 4: [Check output files in Excel] After the completion of DLSIM, users can view the output link performance and agent files in Excel.

Step 5: [Visualize output files in NeXTA] For static traffic assignment, NeXTA is able to display view link travel time, speed and volume, as well as path display in the agent dialog. For dynamic assignment and simulation, one can use NeXTA to view time-dependent queue and density.

Getting Started from NeXTA graphical user interface and running DTALite

Step 1: Download and locate the project folder, check CSV network files.

Locate the project folder of “6_Tempe ASU network”.

Check input files in Excel

node.csv

link.csv

Step 2: Visualize and validate network in NeXTA using shortest path finding

First, select the node layer in the left-hand-side GIS panel, we can use the mouse to select node 44, and node 86. Alternatively, one can use a keyboard shortcut of Control+f to search those nodes.

Go the path GIS layer, right click to check and confirm if this path is connected.

Alternatively, one can use a keyboard shortcut of Control+f to specify the origin and destination for the path.

Step 3. Run DLSIM as a Windows console application from File Explorer

The user now can check output files in Excel for the following two files:

link performance.csv

agent.csv

All DLSIM data files are in CSV format. The files for node, link and zone layers have geometric fields for importing from and exporting to GIS software.

Toy Examples for Computing Static User Equilibrium

3.1 Two-corridor example

This example uses a simple case with a single origin-to-destination pair and two paths p=1 for the primary path, p=2 for the alternative path, see in Figure 3.1 As each path has two links, path 1 has a free-flow travel time of 20 minutes, and path 2 has a free-flow travel time of 30 minutes.

Figure 3.1 illustrative example of two-corridor network

For a given OD demand of 7,000 on this network, we can use the User Equilibrium method to perform traffic assignment. A graphic-based solution process can be described by Figure 3.2. As the path flow changes, the travel time on the two paths reaches the same equilibrium point, which satisfied the requirement of User Equilibrium. User equilibrium solution is reached when the freeway flow is 5400, and arterial flow as 7000-5400=1600, and this leads to the same travel time of 30 min on both routes.

Figure 3.2 illustration of Equilibrium with X axis as freeway path flow.

The detailed parameters are in Table 3.1.

Table 3.1 parameters

Parameters	Value
Freeway flow travel time (min): Freeway:	20
Freeway flow travel time (min): Arterial:	30
Capacity (vehicles / hour): Freeway:	4000
Capacity (vehicles / hour):Arterial:	3000
Demand	7000
BPR alpha	0.15
BPR beta	4

The travel time function is

Freeway_TT = FFTT[1 + 0.15(v/c)4]

Arterial _TT= FFTT[1 + 0.15((demand-v)/c)4]

where:

TT = link travel time

FFTT= free-flow travel time of link

v = link flow

c = link capacity

3.2 Detailed data structure description

Generic network files used for DLSIMinclude files for three layers: physical layer, service layer and demand layer.

Table 3.1 File list for DLSIM

File type	Index: file name	Description
Input for physical layer	1a: node.csv	Define nodes in the network.
	1b.: link.csv	Define links in the network with essential attributes for assignment.
Input for demand layer	2: demand.csv	Define the demand of passengers on each OD pair, which could be extracted by demand_file_list.csv.
Input configuration file	3: settings.csv	Define basic setting for the Network, it contains five sections.
	Section of assignment	Set the number of iteration and the mode of assignment.
	Section of agent_type	Define attributes of each type of agent, including VOT (unit: dollar per hour) and PCE.
	Section of link_type	Define types of links in the network
	Section of demand_period	Define demand period, which could be extracted by demand_file_list
	Section of demand_file_list	Define demand type, period, and format type.
Input for service layer	4: timing.csv	Define space-time arcs for service based on the physical link with time window, time interval and travel time.
Output file	5a: link_performance.csv	Show the performance of each link, including the travel time, volume, and resource balance.
	5b: agent.csv	Show the results of the assignment, including the volume, toll, travel time and distance of each path of each agent, as well as the link sequence and time sequence.

The related files used in DTALite are listed below.

（1）Prepare input data

node.csv

Table 3.2 node.csv

Provide feedback

Saved searches

Use saved searches to filter your results more quickly