Timedora

Traffic Flow Fundamentals

A lesson on the core measures of traffic flow, how they connect, and how fundamental diagrams and the Greenshields model describe traffic states.

Measuring Traffic and Vehicle Separation

Measuring traffic and vehicle separation

Traffic is described with three stream-level measures. is distance traveled per unit time, commonly in mi/h or km/h. (qq) counts vehicles passing a point per unit time, often in vehicles per hour per lane (veh/h/lane). (kk) counts vehicles occupying a roadway length at an instant, often in vehicles per mile per lane (veh/mi/lane). Thus, a roadside count during a specified interval is a volume; divide that count by the interval to obtain a rate. , by contrast, describes how many vehicles are on a section of road at one moment.

Two kinds of should not be confused: time-mean is the arithmetic average of speeds measured as vehicles pass a point, while space-mean is the average over a roadway section. They describe from different observation perspectives.

describe the separation between individual vehicles. Time headway (hh) is the elapsed time between successive vehicles passing the same point, measured from the same reference point on each vehicle. Spacing, or space headway (ss), is the distance between corresponding points on successive vehicles. For a reasonably steady stream, their average values connect these vehicle-level observations to stream measures:

q≈1hˉ,k≈1sˉ.q \approx \frac{1}{\bar h}, \qquad k \approx \frac{1}{\bar s}.

Use compatible units. For example, if mean headway is 2 seconds, convert one hour to 3,600 seconds: approximately 3,600/2=1,8003{,}600/2=1{,}800 vehicles pass the point per hour. Likewise, spacing must be expressed in the reciprocal of the desired unit; an average in miles corresponds approximately to vehicles per mile. These are relationships between stream averages, not exact descriptions of every pair of vehicles.

The

The

For steady traffic, is multiplied by space-mean :

q=kus.q = k u_s.

The units provide a useful check. Multiplying veh/mi/lane by mi/h cancels miles and gives veh/h/lane. This also shows why the in this relationship is specifically the space-mean , not simply the arithmetic average of speeds observed as vehicles pass one point.

Worked example. The source example gives a space-mean of 60 mi/h and of 30 veh/mi/lane. Substitute these values and track the units:

q=(30 veh/mi/lane)(60 mi/h)=1,800 veh/h/lane.\begin{aligned} q &= (30\ \text{veh/mi/lane})(60\ \text{mi/h})\\ &= 1{,}800\ \text{veh/h/lane}. \end{aligned}

So, under the steady-traffic relationship, that and section-average correspond to a of 1,800 vehicles per hour per lane.

, Congestion, and the

, congestion, and the

Fundamental diagrams show relationships among , , and in three paired views: –, –, and –. These are not three unrelated traffic systems; they present the same traffic states from different perspectives.

As rises from a very low level, generally increases while vehicles can still travel relatively freely. At the critical , reaches its maximum, called . Beyond that point, can continue to rise while congestion causes both and to fall. By , vehicles are effectively stopped and approaches zero. The precise curves depend on roadway, traffic, and operating conditions, so a fundamental diagram models observed behavior rather than prescribing one universal curve.

Why the peaks at half

The makes a simplifying assumption: decreases linearly as rises. It expresses this as

u=uf(1−kkj),q=ku,u = u_f\left(1-\frac{k}{k_j}\right), \qquad q=ku,

where ufu_f is free- and kjk_j is . Substituting the equation into q=kuq=ku gives

q=ufk(1−kkj)=uf(k−k2kj).q = u_f k\left(1-\frac{k}{k_j}\right) = u_f\left(k-\frac{k^2}{k_j}\right).

This is a curved – relationship. Completing the square makes the location of its maximum visible:

q=uf[kj4−1kj(k−kj2)2].q = u_f\left[\frac{k_j}{4} - \frac{1}{k_j}\left(k-\frac{k_j}{2}\right)^2\right].

For positive free- and , the squared term is smallest when k=kj/2k=k_j/2, so the model's is greatest at half the . In this model, is zero at the low- endpoint and again at , with a maximum between them. This helps explain the basic rise-to-, then decline pattern; it should not be mistaken for a complete description of real traffic, whose relationships are often more complex.