PTV VISSIM Training
Dynamic Routes
[Link]
[Link]
DYNAMIC ASSIGNMENT
Principle of Microscopic Dynamic Assignment
no assessments costs from (k-1)th simulation
no paths all previously found paths
1st Simulation kth
2 nd Simulation
Simulation
Tk,
distance as assessment
all vehicles on shortest path
first set of costs new set of costs
first set of paths some additional paths
[Link] I Page 2
DYNAMIC ASSIGNMENT
[Link] I Page 3
DYNAMIC ASSIGNMENT
Calculations of Costs and Paths (Simulation period: 3600s, Evaluation
interval: 900s)
1st Iteration 2nd Iteration
path search 0s path search
(shortest distance) read costs of 1st iteration, 1st interval
store path costs
store path costs 900s path search
read costs of 1st iteration, 2nd interval
store path costs
store path costs 1800s path search
read costs of 1st iteration, 3rd interval
store path costs
store path costs 2700s path search
read costs of 1st iteration, 4th interval
store path costs 3600s store path costs
[Link] I Page 4
Parking Lot
Parking Lots: the point where the vehicle actually appear or leave the road
network.
Parking Lot types
Zone connector (no slow down or capacity)
Abstract parking lot (vehicles slow down, capacity inc)
Real parking lot (stop upon entry)
Max. One parking lot between two nodes
Place parking lots between two nodes
Wrong: Correct:
Wrong: Correct:
[Link] I Page 5
Nodes and Edges
Nodes: parts of the network where paths could diverge (as a junction).
Edges: the basic building blokcs of the routes, i.e. a route is sequence of
edges (as a link).
Nodes
Network abstraction:
nodes (black polygons)
Edges
edges (link sequence between
two nodes or within a node) Wrong: Correct:
edges inside nodes
Edges between nodes
Each decision and merging point of a Wrong: Correct :
possible route should be placed within
a node
[Link] I Page 6
OD Matices
Trip Matrix (FMA file) : create using VISUM, Text Editor
Time Interval
Same as the Simulation Time
Factor
To scale the matrix globally
Zone Number
To zone Numbers of all zones that are
used in the matrix
From zone
[Link] I Page 7
Modelling Elements
Connector closures: Prevent certain vehicle classes to use
certain parts of the network
Additional links and connectors to place parking lots
Nodes: abstract definition of junctions
Parking lots: for all origins and destinations of traffic
(also zone connectors)
Dynamic Assignment process parameters
Additional evaluations (e.g. paths, convergence)
[Link] I Page 8
Dynamic Assignment Parameters
Multiple matrices can be defined such as multiple time
intervals, different traffic composition, etc
Cost and Path files store necessary Dynamic Assignment
information for next iteration
Check Edges Box – this function checks whether the
network has changed since you created the input path or
cost file. If network has changed it will give you a
message telling you so.
One of the most common causes of error in dynamic
assignment is when you use an existing cost or path file
with a network that has been changed since the cost/path
file was created.
You should always have this box selected - otherwise,
always ensure that the existing path file has been deleted
or renamed.
[Link] I Page 9
Dynamic Assignment Parameters
Scale total volume to avoid Initial congestion
Evaluation interval defines how costs and paths are
evaluated
Activate to write files during Dynamic
Assignment process
Deactivate when Dynamic Assignment process
is complete
Store costs: if checked, VISSIM will writes a new cost
file
Search new paths: if checked VISSIM calculates new
shortest paths through the network
Store paths: if checked VISSIM stores new paths and
their volumes in the path file
[Link] I Page 10
Dynamic Assignment Parameters
Kirchhoff exponent defines polarisation of path
distribution
Logit function defines parking lot choice model
Correction of overlapping paths affects vehicle
distribution on paths which are partially identical
Avoid long detour, paths with long detours will not
be used for vehicle distribution,
Definition of convergence criteria of Dynamic
Assignment
Travel time on paths: computes the change of
TT on every paths compared to its TT in the
previous run. If change for all paths < the user-
defines factor, convergence for this criterion is
detected.
Travel time of edges
Volume on edges
[Link] I Page 11
Dynamic Assignment Parameters
Correction of overlapping paths
Example: 3 paths with identical length,
2 of them partial identical (overlapping)
Without correction With correction
1/3 ~1/3
2/3 < 2/3
1/3 ~1/3
A 1/3
B A > 1/3
B
1/3 ~1/4
2/3 > 1/2
A 1/3
1/3 B A < 1/2
~1/4 B
Vehicle distribution identical Vehicle distribution depends
regardless of length of on length of overlapping
overlapping section section
[Link] I Page 12
Urban Road Network Model
[Link] I Page 13
Convergence of Path Serach
400
350
300
new paths
250
200
150
100
50
0
1 11 21 31 41 51 61 71 81
Iteration
40000
75 Zones 35000
30000
paths found
25000
20000
15000
10000
5000
0
1 11 21 31 41 51 61 71 81
Iteration
[Link] I Page 14
Road Closure
Connectors (turning movements can
be closed to specific vehicle classes
[Link] I Page 15
Convergence Evaluation
Convergence output file to check if
the Dynamic Assignment process is
complete or not
[Link] I Page 16
PRACTICE CODING DYNAMIC ASSIGNMENT
[Link] I Page 17
[Link] I Page 18
