Timedora

Transportation Networks, Traffic Flow, and Performance

A lesson on how transportation networks represent movement, how travel demand becomes traffic flow, and how to choose and interpret measures of system performance.

Modeling Networks and Choosing Routes

Modeling networks and choosing routes

A transportation system includes the people and goods being moved, the facilities and vehicles that carry them, and the services and operations that connect trips. A simplifies this larger system into facilities and connections. It can represent roads, transit, rail, walking and cycling facilities, or multiple modes together.

A common model uses three elements:

  • Nodes mark trip origins and destinations, or places where travel changes direction or facility, such as intersections and transit stops.

  • Links connect nodes—for example, a road segment, rail track, or walkway. A can have attributes that matter for a particular analysis, including length, direction, lanes, speed, capacity, and travel cost.

  • Connectors represent access between parts of a model, such as the connection from a traffic-analysis-zone centroid to nearby streets, or a walk between a bus stop and a rail station.

The model can also represent features that affect travel, including signals, stop controls, lane restrictions, bridges, transfer points, and schedules. It need not include every real-world detail: it should encode the features relevant to the question being studied.

From connectivity to route choice

Connectivity describes which network elements are linked and whether movement between places is possible. Places that are geographically close may still be poorly connected—for instance, a river or a missing sidewalk may separate them. Connections may also be directed: a one-way street allows movement in one direction, not the reverse. They can depend on mode as well; a available to cars may not be available to pedestrians or buses.

A is an ordered sequence of connected links between an origin and destination. A route is a selected for a trip. Travelers may consider travel time, distance, cost, convenience, and when choosing among possible routes. Therefore, the shortest-distance route is not necessarily the fastest: a longer road with fewer stops may take less time than a shorter route through congested streets.

One way to compare routes is to add the travel costs of their links. If the assumed cost is expected travel time, add the times along each route; the route with the lowest total is the fastest under those assumptions. This conclusion depends on the assigned costs. Actual travel times can vary with time of day and traffic, and travelers may change routes in response to congestion or information.

From to Traffic

From to traffic

describes trips: where they start and end, when they occur, which mode is used, and sometimes why they are made. An origin–destination (O–D) pair identifies a trip’s starting area and destination area. An O–D table summarizes how many trips occur between area pairs during a stated period. Demand might be counted as people, vehicles, transit passengers, or freight, so always identify its unit as well as the time period and areas covered.

Demand is not the same as . Demand describes trips between places; after travelers choose routes, some of those trips may use the same network . The resulting movement on that is its . Because route choices determine which links trips use, demand between areas and on a particular answer different questions.

The amount and timing of demand relative to a ’s capacity influence congestion. A capacity constraint, incident, or work zone can create queues and . Travelers may respond by changing their departure time, mode, or route.

Reading , , and speed together

Three quantities describe traffic on a roadway:

  • (
    qq) is the number of vehicles or people passing a point per unit time, such as vehicles per hour.

  • (
    kk) is the number occupying a unit length of roadway at a particular time, such as vehicles per mile per lane.

  • Space-mean speed (
    vv) is average travel speed across a roadway segment.

When the measurements and units are consistent, they are related by

q=k×v.q = k \times v.

For example, a lane carrying 900 vehicles per hour at an average speed of 30 miles per hour has a corresponding of

k=qv=900 vehicles/hour30 miles/hour=30 vehicles/mile.k = \frac{q}{v} = \frac{900\ \text{vehicles/hour}}{30\ \text{miles/hour}} = 30\ \text{vehicles/mile}.

The hours cancel, leaving vehicles per mile. This relationship also shows why adding vehicles to a roadway does not always raise : at high , speed may fall sharply, and may eventually decline.

Measuring Transportation Performance

Measuring transportation performance

A performance measure describes one aspect of how a facility or network operates. Choose measures to match the question. A speed measured on a particular may help diagnose a bottleneck; a person-based measure may better describe access to a service. Be clear about the unit, geographic scope, and comparison conditions so that the result can be interpreted correctly.

Travel time is the time taken between two points or along a route. Speed is distance divided by travel time. is extra travel time relative to a stated reference, such as free- travel. Because the reference matters, “” has no single value unless the comparison condition is specified.

Other measures describe different aspects of operation. counts people or vehicles served or passing a point during a period. Queue length measures the people or vehicles waiting, either by count or by distance. VMT and VHT are total vehicle-miles and vehicle-hours traveled across a network during a period. These measures are not interchangeable: for example, describes traffic served over time, while a queue describes those waiting.

Worked example: speed and

Suppose a 6-mile trip takes 12 minutes. Convert the travel time to hours before calculating speed:

12 minutes=1260 hours=0.2 hours.12\ \text{minutes} = \frac{12}{60}\ \text{hours} = 0.2\ \text{hours}.

Then divide distance by time:

Speed=6 miles0.2 hours=30 mph.\text{Speed} = \frac{6\ \text{miles}}{0.2\ \text{hours}} = 30\ \text{mph}.

If the same trip takes 18 minutes in peak traffic, its average speed is

6 miles18/60 hours=20 mph.\frac{6\ \text{miles}}{18/60\ \text{hours}} = 20\ \text{mph}.

If free- travel time is 10 minutes, peak-period relative to free is

18−10=8 minutes.18 - 10 = 8\ \text{minutes}.

The reference is part of the meaning of that result; it is not simply another name for the trip’s total travel time.

Average time versus

Travel time and speed describe mobility, while compares travel with a chosen reference. describes the consistency and predictability of travel times, which an average alone can hide. For example, two routes may each average 30 minutes, while one usually takes 28–32 minutes and the other ranges from 20–50 minutes. Their averages are the same, but the first route is more predictable.

A planning time index commonly compares a high-percentile travel time—often the 95th percentile—with free- travel time. The exact measure should fit the analysis purpose and available data. In general, state the measure’s units, geographic scope, and comparison condition, and select it to answer the question at hand.

  • : A point in a network model.

  • : The starting area and destination area for a trip.