ASDA / PHOTO

The traffic flow models ASDA and FOTO , which are based on Boris Kerner's three-phase traffic theory , can detect traffic jams on expressways and track them in time and space (Fig. 1).

Fig. 1: Empirical examples of traffic jams shown by the ASDA / FOTO models using traffic data obtained with local measurements on various motorways in Great Britain, Germany and the USA. Representation of traffic jams in place-time diagrams with areas of qualitatively different traffic phases: 1. Wide moving jam (red). 2. Synchronized traffic (yellow). 3. Free traffic (white).

FOTO ( F orecasting o f t raffic o bjects) recognizes and tracks the areas of the synchronized traffic in time and space. ASDA ( A utomatic S tau d ynamik a nalysis: Automatic Tracking of Moving Jams) detects and tracks slow moving traffic. The ASDA / FOTO models are designed for online applications and work online without changing the model parameters under different weather conditions and road infrastructures as well as other traffic flow parameters (truck shares, etc.)

General

Road traffic is either jammed or free. Traffic takes place in time and space, ie it is a temporal-spatial process. Even so, traffic is usually only measured directly in some places, e.g. B. with measuring loops, video devices, vehicle data or data from mobile phones. For efficient traffic control and other intelligent transport systems, a temporal and spatial detection of traffic disruptions is necessary, especially at all locations in the road network where no direct measurements are available.

Figure 2: Temporal and spatial properties of traffic disruptions: (a) Measured data of mean vehicle speeds in time and space. (b) Representation of vehicle speeds in a distance-time diagram. (cf) time courses of the measured values of speed (c, e) and traffic flow (d, f) at two different locations for the traffic disruption of (a), (b); the data in (c, d) and (e, f) were measured at positions 17.1 km (c, d) (downstream of the approach in (a, b)) and at position 16.2 km (e, f) (upstream of the driveway). At 17.1 km the flow of traffic (d) in free and synchronized traffic is greater than at 16.2 km (f) because of the vehicles flowing in via the approach.

First of all, the ASDA / FOTO models recognize the two traffic phases of synchronized traffic (S) and moving traffic jam (J) in measured data in accordance with the measured properties of traffic disruptions. One of these properties that the ASDA / FOTO models use to identify the respective traffic phase is as follows: In phase J, both speed and traffic flow are very low (Fig. 2 (cf)). In contrast to this, the traffic flow in an area of synchronized traffic can be as high as in free traffic (Fig. 2 (d, f)), even if the speed is significantly reduced compared to free traffic (Fig. 2 (c, e) ).

Image 3: ASDA / FOTO models. Designations "jam 1", "jam 2" are two different areas of phase J; "syn" is for phase S areas; "up" and "down" indicate the respective upstream and downstream fronts.

After the detection of the phase S, the Model keeps track of the two fronts of synchronized traffic , over time (Fig. 3). After detection of the phase J the ASDA model pursued the two fronts of the wide moving traffic jams , (Fig. 3). This tracking also takes place between the positions of measuring points if the fronts of the traffic phases could not be measured directly. ${\ displaystyle x_ {down} ^ {(syn)} (t)}$ ${\ displaystyle x_ {up} ^ {(syn)} (t)}$ ${\ displaystyle t}$ ${\ displaystyle x_ {down} ^ {(jam)} (t)}$ ${\ displaystyle x_ {up} ^ {(jam)} (t)}$

In other words: the FOTO and ASDA models predict the positions of the congestion fronts over time, regardless of the local measuring points. ASDA / FOTO can also be used to predict a merging and / or dissolution of one or both traffic phases J and S between individual measuring points.

ASDA / FOTO for local measurements

Approach of cumulative traffic flows in FOTO

While the downstream front of the synchronized traffic is mostly stationary (see Fig. 2 (a, b)), the upstream front, where vehicles decelerate into the synchronized traffic, can move against the direction of travel.

In measured traffic data, the speed of the upstream front of the synchronized traffic depends on the traffic variables within the synchronized traffic and in free traffic further upstream.

A good match with measurement data is achieved if the time dependency of the upstream front of the synchronized traffic in FOTO is determined using a so-called cumulative traffic flow approach:

${\ displaystyle x_ {up} ^ {(syn)} (t) = \ mu {\ frac {1} {n}} \ int _ {t_ {0}} ^ {t} (q_ {S} (t) -q_ {F} (t)) dt, \ \ t \ geq t_ {0} \ qquad \ qquad (1)}$

where and [Vzg / h] correspond to the respective traffic flows upstream and downstream of the upstream front of the synchronized traffic, is a model parameter [m / Vzg] and is the number of lanes. ${\ displaystyle q_ {F}}$ ${\ displaystyle q_ {S}}$ ${\ displaystyle \ mu}$ ${\ displaystyle n}$

Two approaches to congestion tracking in ASDA

There are two basic approaches to congestion tracking with ASDA:

Use of the Stokes shock wave formula
Use a characteristic speed for moving wide traffic jams

Use of the Stokes shock wave formula

The current speed v (jam) of the front of a moving wide traffic jam is determined using Stoke's formula from 1848:

${\ displaystyle v ^ {(jam)} (t) = {\ frac {q_ {2} (t) -q_ {1} (t)} {\ rho _ {2} (t) - \ rho _ {1 } (t)}} \ qquad \ qquad (2)}$ ,

with and as traffic flow and traffic density upstream ( and downstream) of the congestion. In (2) is no relationship, in particular not a fundamental diagram between the traffic flows , and traffic density , assumed: the values are independently determined from the measured data. ${\ displaystyle q_ {1}}$ ${\ displaystyle \ rho _ {1}}$ ${\ displaystyle q_ {2}}$ ${\ displaystyle \ rho _ {2}}$ ${\ displaystyle q_ {1}}$ ${\ displaystyle q_ {2}}$ ${\ displaystyle \ rho _ {1}}$ ${\ displaystyle \ rho _ {2}}$

Use a characteristic speed

If there is no measured data downstream of a moving wide congestion with which to use formula (2) above, use

${\ displaystyle v_ {down} ^ {(jam)} = v_ {g} \ qquad \ qquad (3)}$

using as the characteristic velocity of the downstream congestion front from the Kerner phase property [J] . This means that after a detection of the downstream congestion front of a moving wide congestion at the time , the position results from: ${\ displaystyle v_ {g}}$ ${\ displaystyle t = t_ {1}}$

${\ displaystyle x (t) -x (t_ {1}) = v_ {g} (t-t_ {1}), \ \ t> t_ {1} \ qquad \ qquad (4)}$

The characteristic speed is shown in Fig. 4. Two moving wide traffic jams follow one another through a route section while maintaining the average speed of the downstream traffic jam face. In contrast to the mean speed of the downstream front of a moving wide congestion, the speed of the upstream front depends on the traffic flow and traffic density in the traffic upstream of the congestion. The use of formula (4) can consequently lead to a larger error in the estimation of the mean speed of the upstream front of the congestion.

Figure 4: Measured traffic data to show the characteristic congestion properties of [J]: (a, b) Average speed v in km / h (a) and traffic flow q in [vehicles / h] (b) in time and space. (c, d) Time courses of traffic flow and speed in wide moving traffic jams at two different locations per lane from (a, b).

On the basis of a lot of measured data on German motorways, a value of was determined. Even if this average speed is independent of the traffic variables in front of a moving traffic jam, the value is influenced by the proportion of trucks, weather, driver behavior, etc. As a result, the value for in various data from several years fluctuates in the range . ${\ displaystyle v_ {g} \ approx -15 {\ mbox {km / h}}}$ ${\ displaystyle v_ {g}}$ ${\ displaystyle v_ {g}}$ ${\ displaystyle -12> v_ {g}> - 20 {\ mbox {km / h}}}$

On-line applications of ASDA / FOTO in traffic control centers

ASDA / FOTO is now used permanently in the Hessen traffic control center on 1200 km of autobahns. Since April 2004, measurement data from around 2500 detectors have been analyzed automatically with ASDA / FOTO. The resulting temporal-spatial traffic patterns are shown in a distance-time diagram like Fig. 5. In 2007 the ASDA / FOTO online system was used in North Rhine-Westphalia (NRW). The raw data was transmitted to the WDR , the public broadcasting corporation of the state of North Rhine-Westphalia based in Cologne, which broadcasts traffic information to end customers via the RDS radio channel. The application processes relevant parts of the entire network in NRW with a length of 1900 km using more than 1000 detectors. In addition, ASDA / FOTO has also been used online in Northern Bavaria since 2009.

Figure 5: Temporal-spatial traffic pattern in ASDA / FOTO: Distance-time diagram with vehicle trajectories 1–4 and respective travel times. Measured values from detectors as input variables for ASDA / FOTO on the A5-Nord in Hesse on June 14, 2006.

Mean traffic flow characteristics and travel times

In addition to the temporal and spatial detection of traffic disruptions (Fig. 1 and Fig. 5), ASDA / FOTO offers further average parameters for the traffic phases S and J. This allows travel times to be estimated on route sections or along a vehicle journey (see examples in Vehicle trajectories 1–4 in Fig. 5).

ASDA / FOTO for data from moving measurements

ASDA / FOTO can also be used to reconstruct traffic patterns based on data from moving sources (e.g. vehicles, mobile devices). First, ASDA / FOTO recognizes the state and phase transitions along the journey of a vehicle: each of these transition points is connected to a front that separates two traffic phases. With these crossing points located, ASDA / FOTO tracks the areas of synchronized traffic and moving wide congestion. In this tracking, ASDA / FOTO models use the measured properties of phases J and S, which were described above (see Fig. 2 and Fig. 4).

literature

BS Kerner: Introduction to Modern Traffic Flow Theory and Control: The Long Road to Three-Phase Traffic Theory. Springer, Berlin / New York 2009.
BS Kerner: The Physics of Traffic. Springer, Berlin / New York 2004.

credentials

BS Kerner, P. Konhäuser: Structure and parameters of clusters in traffic flow. In: Physical Review E. Vol. 50, 1994, p. 54.

BS Kerner, H. Rehborn: Experimental features and characteristics of traffic jams. In: Physical Review E. .. Vol. 53, 1996, p. 1297.

BS Kerner, H. Rehborn: Experimental properties of complexity in traffic flow. In: Physical Review E. Vol. 53, 1996, p. R4257.

BS Kerner, H. Kirschfink, H. Rehborn: Automatic traffic jam tracking on motorways. In: Road traffic technology. No. 9, 1997, pp. 430-438.

BS Kerner, H. Rehborn: Measurements of the traffic flow: Characteristic properties of traffic jams on motorways. In: International Transport. 5/1998, pp. 196-203.

BS Kerner, H. Rehborn, M. Aleksić, A. Haug, R. Lange: Tracking and prediction of traffic disruptions on motorways with "ASDA" and "FOTO" in online operation in the traffic control center in Rüsselsheim. In: Road traffic technology. No. 10, 2000, pp. 521-527.

BS Kerner, H. Rehborn, M. Aleksić, A. Haug: Methods for Tracing and Forecasting of Congested Traffic Patterns on Highways. In: Traffic Engineering and Control. 09/2001, pp. 282-287.

BS Kerner, H. Rehborn, M. Aleksić, A. Haug, R. Lange: Online Automatic tracing and forecasting of traffic patterns with models “ASDA” and “FOTO”. In: Traffic Engineering and Control. 11/2001, pp. 345-350.

BS Kerner, H. Rehborn, M. Aleksić, A. Haug: Recognition and Tracing of Spatial-Temporal Congested Traffic Patterns on Freeways. In: Transportation Research C. 12, 2004, pp. 369-400.

J. Palmer, H. Rehborn: ASDA / FOTO based on Kerner's Three-Phase Traffic Theory in North-Rhine Westfalia (in German). In: Road traffic technology. No. 8, 2007, pp. 463-470.

J. Palmer, H. Rehborn, L. Mbekeani: Traffic Congestion Interpretation Based on Kerner's Three-Phase Traffic Theory in USA. In: Proceedings 15th World Congress on ITS. New York 2008.

J. Palmer, H. Rehborn: Reconstruction of congested traffic patterns using traffic state detection in autonomous vehicles based on Kerner's three-phase traffic theory. In: Proceedings of. 16th World Congress on ITS .. Stockholm., 2009

H. Rehborn, SL Klenov: Traffic Prediction of Congested Patterns. In: R. Meyers (Ed.): Encyclopedia of Complexity and Systems Science. Springer, New York 2009, pp. 9500-9536.

Boris S. Kerner, H. Rehborn, SL Klenov, J. Palmer, M. Prinn: Method for traffic state detection in a vehicle. German Patent publication DE 10 2008 003 039 A1., 2009.

Individual evidence

↑ Boris S. Kerner, H. Kirschfink, H. Rehborn: Method for the automatic monitoring of traffic including the analysis of back-up dynamics. German patent DE19647127C2, USA patent: US 5861820 (Filed: 1996).
↑ Boris S. Kerner, H. Rehborn: Traffic surveillance method and vehicle flow control in a road network. German patent disclosure DE19835979A1, USA patent: US 6587779B1 (Filed: 1998)
↑ Boris S. Kerner, M. Aleksić, U. Denneler: Method and device for traffic condition monitoring German patent DE19944077C1 (Filed: 1999).
↑ Boris S. Kerner: Method for monitoring the condition of traffic for a traffic network comprising effective narrow points. German patent publication DE19944075A1; USA patent: US 6813555B1; Japan: JP 2002117481 (Filed: 1999)]
↑ Boris S. Kerner: German Patent DE10036789A1; Method for determining the traffic state in a traffic network with effective bottlenecks, USA patent: US 6522970B2 (Filed: 2000)
↑ George G. Stokes: On a difficulty in the theory of sound. In: Philosophical Magazine. 33, 1848, 2009, pp. 349-356.
↑ BS Kerner, H. Rehborn, J. Palmer, SL Klenov: Using probe vehicle to generate jam warning messages, Traffic Engineering and Control. Vol 52, No 3, 2011, pp. 141-148.
^ J. Palmer, H. Rehborn, BS Kerner: ASDA and FOTO Models based on Probe Vehicle Data. In: Traffic Engineering and Control. Vol 52 No 4, 2011, pp. 183-191.

[KernerPatent1996-1] Boris S. Kerner, H. Kirschfink, H. Rehborn: Method for the automatic monitoring of traffic including the analysis of back-up dynamics. German patent DE19647127C2, USA patent: US 5861820 (Filed: 1996).

[KernerPatent1998-2] Boris S. Kerner, H. Rehborn: Traffic surveillance method and vehicle flow control in a road network. German patent disclosure DE19835979A1, USA patent: US 6587779B1 (Filed: 1998)

[KernerPatent1999a-3] Boris S. Kerner, M. Aleksić, U. Denneler: Method and device for traffic condition monitoring German patent DE19944077C1 (Filed: 1999).

[KernerPatent1999b-4] Boris S. Kerner: Method for monitoring the condition of traffic for a traffic network comprising effective narrow points. German patent publication DE19944075A1; USA patent: US 6813555B1; Japan: JP 2002117481 (Filed: 1999)]

[KernerPatent2000-5] Boris S. Kerner: German Patent DE10036789A1; Method for determining the traffic state in a traffic network with effective bottlenecks, USA patent: US 6522970B2 (Filed: 2000)

[Stokes1848-6] George G. Stokes: On a difficulty in the theory of sound. In: Philosophical Magazine. 33, 1848, 2009, pp. 349-356.

[Tec1_2011-7] BS Kerner, H. Rehborn, J. Palmer, SL Klenov: Using probe vehicle to generate jam warning messages, Traffic Engineering and Control. Vol 52, No 3, 2011, pp. 141-148.

[Tec2_2011-8] J. Palmer, H. Rehborn, BS Kerner: ASDA and FOTO Models based on Probe Vehicle Data. In: Traffic Engineering and Control. Vol 52 No 4, 2011, pp. 183-191.