Rules and examples for defining meso network nodes

Meso network nodes must be modeled correctly for Vissim to be able to model conflicts realistically in mesoscopic simulation. The level of correctness has a decisive impact on the result of dynamic assignment in mesoscopic simulation. Follow the rules for defining meso network nodes when you manually define individual meso network nodes in the Network editor. Before starting to model meso network nodes it is necessary to look at the examples and read the descriptions of correct and incorrect definitions for meso network nodes. Follow the instructions for modeling meso network nodes (Modeling meso network nodes).

Rules for defining meso network nodes

• Rule 1: Meso network nodes must be defined everywhere on a link where more than one connector begins or ends.

Exception - PT lay-by-stops: It is not necessary to define meso network nodes at the starting point and the end of a connector, which runs towards a link of a lay-by-stop, as well as at the starting point and the end of a connector, which runs from the lay-by-stop to the original link.

• Rule 2: For each intersection, at least one meso network node must be defined. Depending on the node geometry, several meso network nodes may be defined.

For non-signalized intersections the following applies: All conflict areas must be defined. In order to decide which conflict areas shall lie within a separate meso network node, check the following:

• Where is the vehicle supposed to stop? For all turn conflicts, the vehicle stops at the meso network node. Model the meso network node so that its edge roughly corresponds to a stopping position of the vehicle in reality, e.g. a stop line.
• Are the incoming meso edges used by vehicles with the right of way to reach the meso network node relevant for all turn conflicts in the meso network node? The size and positioning of the node determine which meso edges are perceived as edges with vehicles that have the right of way (Meso conflict relevant and non-relevant edges).
• How long is the travel time on the incoming meso edge used by vehicles with the right of way to reach to meso network node? This travel time should be longer then the meso critical gap of the subordinate flow.

For signalized intersections note that the decisive factors are the stop position and storage capacity. The vehicle always stops at the meso network node. If the real situation cannot be modeled with one meso network node only, model several nodes, e.g. for a channelized right turn.

• Rule 3: On turn meso edges, the following properties must not change:
• the number of lanes
• the link behavior type
• the meso speed, if the meso speed model Link-related is selected (Car following model for mesoscopic simulation)

This means the Defining links of the meso turn must each have the same value (Attributes of meso turns).

Note: Please note the limitations and information that apply for defining meso network nodes (Defining meso network nodes).

Examples of applying the rules for defining meso network nodes

The following examples show how the rules are applied when you model intersections. First, you are shown how the position of a meso network node impacts where at the conflict area a vehicle stops and which edges it perceives as relevant:

• Consequences of correct and incorrect positioning of meso network nodes
• Meso conflict relevant and non-relevant edges

Then, you are given an explanation of how the rules impact the modeling of different types of intersections. For different network objects, the impact of rules on the meso graph structure and on simulation is demonstrated:

• Nodes in areas where the number of lanes changes
• Modeling connectors in meso network nodes
• Modeling a signalized intersection
• Modeling intersections with lane widening
• Modeling intersections with bypass and channelized turn
• Modeling roundabouts
• Modeling reduced speed areas on links
• Modeling signal controllers on links

Consequences of correct and incorrect positioning of meso network nodes

The following example describes the meaning of travel time as a meso critical gap on an edge for a 3-leg intersection with nine turn conflicts. Nine meso network nodes have been manually defined at the nine turn conflicts (1 to 9):

This type of modeling is not recommended if the travel time on the edge leading into the meso network node is shorter than the meso critical gap of the conflict in the meso network node. This leads to incorrect modeling of the conflicts in mesoscopic simulation. It is illustrated in the following figure and explained in the description given below it.

Situation:  The vehicle is coming from below and turns upward left:

The conflicts are not modeled correctly in mesoscopic simulation: Cause: The travel times on some edges leading into the nodes are too short. Effect: Vehicle also stops at wrong positions. The vehicle stops for the conflict at node 9, position a, as it should. The vehicle stops for the conflict at node 6, position b, and only pays attention to the edge between node 5 and 6. The travel time at the edge between nodes 5 and 6 acts as a critical gap, if it is shorter than the meso critical gap entered for the conflict. The same applies for the subsequent nodes: The vehicle stops for the conflict at node 3, position c, and only pays attention to the edge between node 3 and 2. The travel time at the edge between nodes 3 and 2 acts as a critical gap, if it is shorter than the meso critical gap entered for the conflict. The vehicle stops for the conflict at node 1, position d, and only pays attention to the edge between node 1 and 2. The travel time at the edge between nodes 1 and 2 acts as a critical gap, if it is shorter than the meso critical gap entered for the conflict.

If the vehicle stops at a wrong position and the travel time at the edge leading into the meso network node is very short, the travel time acts as a critical gap. Vissim is then unable to model the conflicts in mesoscopic simulation realistically (as illustrated in the figure above).

If, for instance, no meso network node is defined for node 3 (at top of figure), Vissim does not recognize the conflict there and the conflict is ignored in mesoscopic simulation.

Solution: If for these types of intersections, with short edges between conflicts, only one meso network node is defined, Vissim is able to model conflicts realistically in mesoscopic simulation. With one meso network node only, the left-turning vehicle has only one stop position in all subsequent conflicts. The travel times at the incoming edges are long enough and the vehicle stops at the correct position. This is illustrated in the following figure and explained in the description given below it.

Correct modeling: The travel times at all edges leading into the node at conflict points are long enough. This ensures that the vehicle stops at the correct positions:

• With conflict 11 in the black dot, the vehicle is aware of the edge leading from node 1 into node 2. If the travel time on this edge is longer than the meso critical gap for the conflict, the specified value is used as critical gap, e.g. 3.5 s.
• With conflicts 22 and 33 in the red dots, the vehicle is aware of the incoming edge between nodes 3 and 2. If the travel time on this edge is longer than the meso critical gap for the conflict, the specified value is used as critical gap, e.g. 3.5 s.

Meso conflict relevant and non-relevant edges

This example shows a roundabout (right-hand traffic) for which multiple meso network nodes have been correctly positioned, in the figure on the left. In the figure on the right, only one meso network node has been positioned across the roundabout. The following two figures show the meso edges the vehicle is aware of when it stops at the meso network node:

• Correctly modeled: The modeling in the figure on the left ensures that the vehicle is aware of the relevant meso edge (yellow between the two meso network nodes), leading directly into the correctly positioned meso network node at which the vehicle stops. The correct meso critical gap is used.
• Incorrectly modeled: The modeling in the figure on the right does not allow the vehicle to become aware of the relevant meso node. For conflicts in the meso network node, e.g. the entry of the vehicle into the roundabout, the vehicle is only aware of non-relevant meso edges (the three meso edges highlighted in yellow that lead into the meso network node from the left, top and right). The vehicle cannot become aware of the relevant node as shown in the figure on the left. Thus, it cannot take a correct meso critical gap into account. The vehicle stops at the meso network mode and gives priority to the vehicles coming from the right, top and left, as it is only aware of their meso nodes.

Correct: Vehicle is aware of relevant meso edge (yellow between the two bottom meso network nodes)	Incorrect: Vehicle is only aware of non-relevant meso nodes (yellow)
: Vehicle is coming from below and wants to turn right into roundabout

Nodes in areas where the number of lanes changes

There are different ways to model areas in which the number of lanes changes. These impact dynamic assignment in mesoscopic simulation in different ways. This is illustrated in the following figure and explained in the table listed below it.

Connector connects a double-lane link with a single-lane link:

The vehicle may only change lane at the end of a meso node. This applies for meso nodes generated automatically by Vissim and for modeled meso network nodes (Mesoscopic node-edge model).

Modeling	Situation	Mesoscopic simulation
A	a connector not a modeled meso network node	Vissim automatically generates a meso node at the beginning of the connector. On the double-lane link, vehicles may use the right lane only which is unrealistic.
B	a connector a modeled meso network node	Vissim automatically generates a meso node at the beginning of a connector. In this case, you need not manually define the meso network node. This meso network node is defined manually, so that contrary to A, both lanes may be used and lane changes are possible. On the double-lane link, vehicles may use both lanes. Vehicles may use the left lane up until the modeled meso network node. At the end of this meso network node, all vehicles must change from the left lane to the right lane.
C	two connectors a modeled meso network node	On the double-lane link, vehicles may use both lanes. For dynamic assignment when one of the connectors is closed then only one connector is available for the path search. However for mesoscopic simulation both connectors remain available. Apply an edge closure to one of the edges for dynamic assignment. This way you can avoid parallel edges in dynamic assignment. Parallel edges multiply the number of possible paths significantly.

Connector connects a single-lane link with a double-lane link:

Modeling	Situation	Mesoscopic simulation
D	a connector not a modeled meso network node	Vissim automatically generates a meso node at the end of the connector. Vehicles can use both lanes of the double-lane link. To use the lane on the right, the vehicle must change lanes. Lane changes are penalized during lane selection. This is why the left lane is preferred.
E	two connectors a modeled meso network node	On the double-lane link, vehicles may use both lanes. As both lanes can be easily reached, no lane change is required and the vehicles are distributed evenly across the lanes, if both lanes are permitted for the vehicle route. Apply an edge closure to one of the edges for dynamic assignment. This way you can avoid parallel edges in dynamic assignment. Parallel edges multiply the number of possible paths significantly.

Modeling connectors in meso network nodes

Rule	Description
1	Meso network nodes must be defined everywhere on a link where more than one connector begins or ends.
	A link leads into a node. Two connectors lead out of the node. The connectors do not have to lie entirely within the node. The connectors must begin within the node.
	Correct:	False:

 

Rule	Description
1	Meso network nodes must be defined everywhere on a link where more than one connector begins or ends.
	The connectors do not have to lie entirely within the node. Left meso node: Two connectors lead into node. The connectors must end within the node. Right meso node: Two connectors lead out of node. The connectors must begin within the node.
	Correct:	False:

 

Rule	Description
1	Meso network nodes must be defined everywhere on a link where more than one connector begins or ends.
	If the transition from a one-lane link to a two-lane link is modeled across two connectors, these must lie entirely within the node.
	Correct:	False:

 

Rule	Description
3	On turn meso edges, the following properties must not change: the number of lanes the link behavior type the meso speed, if the meso speed model Link-related is selected (Car following model for mesoscopic simulation)
	The number of lanes must be the same for each defining link contained within the meso network node. However, the number of lanes of the inbound meso edge and the number of lanes of the outbound meso edge may differ.
	Correct:	False:

Modeling a signalized intersection

Rule	Description
2	For each intersection, at least one meso network node must be defined. Depending on the node geometry, several meso network nodes may be defined.
	A signal head may be defined on links or connectors. Position signal heads within a meso network node. If a signal head is less than 5 m away from the border of the node, the software will not automatically generate another meso node. Vissim then assumes that the vehicle stops at the node border and the signal head belongs to the node. If a signal head is more than 5 m away from the border of a node, Vissim automatically generates an additional meso node. If meso nodes are positioned too close to each other, the edge between them might become so short that, in certain situations, Vissim cannot not model the driving behavior realistically. Vissim only accounts for the vehicles at the edge leading into the node, not any other nodes further downstream. In the figure at the bottom right this means: If the signal heads are 10 m from the meso network node of the intersection, Vissim automatically generates a meso node at the signal heads. The edge between the two nodes is then 10 m long. The travel time of a vehicle driving at 10 m/s on this edge is 1 s. This second acts as a critical gap for the vehicle approaching from the right and turning left, regardless of the actual meso critical gap defined, as the vehicle cannot tell whether, beyond the meso node, there is a vehicle approaching from the left that it must yield to. A critical gap of 1 s does not give the vehicle enough time to yield. Solutions: a) Position the signal heads within the meso network node or b) reduce the distance between signal heads and meso network nodes to below 5 m or c) ensure that the length of the edge leading into the node is long enough to create a travel time on the edge that is longer than the meso critical gap of the turn conflict in the node.
	Correct: The signal heads are positioned within the meso network node or at a maximum of 5 m from it	Not recommended: The signal heads are positioned at a distance of more than 5 m from the meso network node

Modeling intersections with lane widening

Rule	Description
2	For each intersection, at least one meso network node must be defined. Depending on the node geometry, several meso network nodes may be defined.
	Lane widening must not lie within the node. When creating a meso graph, Vissim automatically generates a meso node of the type Other where the lane widening begins (Attributes of meso nodes).
	Correct:	False:

Modeling intersections with bypass and channelized turn

Rule	Description
2	For each intersection, at least one meso network node must be defined. Depending on the node geometry, several meso network nodes may be defined.
	According to Rule 1 meso network nodes must be placed at the branchings where the bypass begins and ends. These nodes must have the attribute Use for mesoscopic simulation. In the figure on right, the vehicle approaching from the left and turning to the right (downwards), already waits at the beginning of the right turn lane, at the large node that represents the entire intersection. This is not recommended. In the figure on the left, the vehicle waits at the end of the right turn lane, at the small meso network node modeled for this purpose. In addition, the intersection itself must lie within a meso network node.
	Correct:	Not recommended:

Modeling roundabouts

Note: The following tips refer to the modeling of simple roundabouts, e.g. those with a single lane, with no or only one bypass and few entries and exits. To model more complex roundabouts or roundabouts whose conflicts cannot be modeled correctly in mesoscopic simulation, define sections and perform a hybrid simulation (Using hybrid simulation).

 

Rule	Description
2	For each intersection, at least one meso network node must be defined. Depending on the node geometry, several meso network nodes may be defined.
	Each branching and thus each entry and exit must lie within a meso network node. The connectors do not have to lie entirely within the node. If there is a bypass, meso network nodes must be placed at the branchings where the bypass begins and ends. These nodes must have the attribute Use for mesoscopic simulation.
	Correct:	False:

Model meso network nodes for a roundabout depending on the distance between the exit and the next entry and according to approach A or approach B. The is illustrated in the figures and their descriptions below:

Description of approach A
If the distance between the exit and next entry downstream is large enough, define a meso network node for the exit and another one for the entry. This is the case in the following situations: Condition 1: The travel time on the roundabout between exit x and the next entry downstream y is equal to or larger than the meso critical gap for the conflict in y: txy > tCG(y). If this condition is not met, but the following one is, you can still use approach A to model the roundabout: Condition 2 (for right-hand traffic): The travel time on the roundabout between exit x and the next entry downstream y is equal to or larger than the travel time on the lane between upstream entry b and the next downstream exit x: tbx < txy. For the conflict in y, the critical gap is the travel time on the roundabout between exit x and the next entry downstream y.
Correct: txy > tCG(y) or tbx < txy

When you export a network from Visum and import it into Vissim via ANM import, Vissim automatically generates meso network nodes based on approach A. These nodes do not require any subsequent editing (Generated network objects from the ANM import). The table lists different speeds to illustrate the minimum distance between exit x and the next downstream entry y with a critical cap of 3.5 s, in order for approach A to meet condition 1:

Veh speed on roundabout lane		min. distance [m] x-y to meet condition 1
m/s	km/h
1	3.6	3.5
2	7.2	7.0
3	10.8	10.5
5	18	17.5
7	25.2	24.5
10	36	35.0
14	50.4	49

Description of approach B
If the distance between the exit and next entry downstream is not large enough, define a common meso network node for both the exit and entry. This is the case, when the two following situations happen at the same time: The travel time on the roundabout lane between entry b and the next downstream exit x is larger than the travel time between exit x and the next exit downstream y: tbx > txy and the travel time on the roundabout lane between exit x and the next entry downstream y is smaller than the meso critical gap for the conflict in y: txy < tCG(y)
Correct: tbx > txy and txy < tCG(y)

The two following figures show wrong approaches to define meso network nodes: These approaches produce incorrect results when used to model conflicts in mesoscopic simulation:

Incorrect approach 1: The distance between the entry and the next exit downstream is not large enough. As a result, too many conflicts arise at each of the nodes: 4 meso turn edges: from roundabout from entry into roundabout into exit 6 meso turn conflicts Solution: If the entry and the next exit downstream are very close to each other, use approach A, even if this results in short edges between the meso network nodes. In that case, at each of the two meso network nodes, there will be only one merging or branching conflict. At the branching conflict, the short edge leading out of the meso network node does not pose a problem.

Figure below Incorrect approach 2: Only a single meso network node has been defined for all conflicts in the roundabout. Effect: Conflicts are not modeled realistically, vehicles stop at wrong positions and for conflicts, the time gap is based on non-relevant edges.

Modeling reduced speed areas on links

Description
Reduced speed areas on links are only taken into account when the vehicle-based meso speed model is used on links. In the top figure, the speed distribution specified for the reduced speed area has an impact on the entire meso node for vehicles heading from west (left) to east (right) on the right lane. If you want the reduced speed area to have only a local impact, you need to insert a meso network node (Defining nodes). The reduced speed area must lie entirely within the node. The meso edge within the node is a turn meso edge that is impacted by the speed distribution of the reduced speed area. Desired speed decisions are treated the same way in mesoscopic simulation.

Modeling signal controllers on links

Description
You do not have to create nodes for signal heads on links. In this case, a meso node of the type Other is automatically generated on the link (Attributes of meso nodes). When doing so, Vissim automatically generates two meso edges.