323
1 INTRODUCTION
Due to the increasing number of cars in urban areas,
cities must adapt their infrastructure to ensure safety
as well as an acceptable level of capacity. Throughput
is an important factor in the main communication
routes with the highest traffic, especially during rush
hours. Different countries have different rules for
traffic lights at intersections in cities [1]–[5]. For this
purpose, the most common solution is the use of
traffic lights due to their wide implementation
possibilities. Signaling organizes traffic by using
separate traffic phases where specific signal groups,
by the rules, may or may not receive a green light for
a traffic maneuver [6]–[8]. As well as traffic lights
connected to intersections in a sequence where the
distance between them is no longer than 1000 m [9]
allows you to coordinate the continuity of the journey
in the designated direction [10]. Another advantage is
the safety of pedestrians, who, thanks to the
appropriate signal phase, can cross the road without
collision. In this case, we distinguish phases for
pedestrians where the green signal always appears
regardless of whether the pedestrian is present at the
intersection or not - constant-time signaling or
adaptive signaling that will adjust the necessity of the
pedestrian phase from direct notification of such a
need to the controller. This controller can obtain
information, for example, from a special button that
the pedestrian is forced to press to cross or using
automatic detection of pedestrians using various
types of radars and sensors. Depending on the length
of the road to overcome, pedestrians must be given
the appropriate number of seconds of green light so
that the route is safe. For example, in Poland, there are
rules where the pedestrian speed is set at 1.4 m/s [11].
If the pedestrian had to cross only two traffic lanes
with a total width of 7 m, the pedestrian would need 5
seconds to cross, however, in the example of Polish
regulations, this time cannot be shorter than 8
seconds. The eight seconds are divided into two
different beeps, four seconds of a solid green beep and
4 seconds of a flashing beep to indicate that the beep
is ending. If, in justified cases, pedestrians at a given
intersection move slower due to e.g. dysfunction or
disability, the regulations allow reducing the speed of
Analysis of the Capacity of Intersections with Fixed-
time Signalling Depending on the Duration of the Green
Phase for Pedestrians
M
. Ziemska-Osuch & D. Osuch
Gdynia Maritime University, Gdynia, Poland
ABSTRACT: The article aims to analyze the capacity conditions of intersections with constant-time traffic lights
using a road transport micro-modeling tool. The article uses actual data on intersections and then a model was
made in the PTV Vissim program of the existing state and the changed state about the duration of the green
time for pedestrians shortened to the minimum resulting from Polish regulations, the relationship between the
length of the pedestrian path and the speed of pedestrian movement, but not less than 8 seconds. The results
were tested at different intersections depending on the number of entrances as well as different stages of traffic
lights.
http://www.transnav.eu
the
International Journal
on Marine Navigation
and Safety of Sea Transportation
Volume 18
Number 2
June 2024
DOI: 10.12716/1001.18.02.
08
324
movement to 1 m/s[11]. In this case, the time needed
to cross two traffic lanes totaling 7 m will be 7
seconds. Depending on the calculation used, a
pedestrian cannot get less than a certain minimum
according to the law in force in each country. In the
example of Poland, not less than 8 seconds if the
pedestrian distance is less than or equal to 11.2
meters. However, due to the comfort of pedestrians,
the time allocated for the crossing is usually longer
and depends on the length of the traffic phase of
vehicles that have a traffic permit at the same time.
From the point of view of the pedestrian, this is a very
beneficial solution because the time to cross is much
longer and the pedestrian does not have to appear at
the crossing in the first second of the green light to
have a safe crossing. However, from the point of view
of capacity, especially right and left turns at inlets
parallel to the crossing, the shortest possible time for
pedestrians may affect the capacity of the entire
intersection, obtaining better results. In this article, on
the example of different geometry of real intersections
with the real traffic of vehicles, the existing state was
compared where the length of the green light for
pedestrians was consistent with real constant-time
programs and the changed state where the length of
the green light for pedestrians was directly dependent
on the length of the road that the pedestrian must
cover on the crosswalk pedestrians. All calculations
and assumptions are by the Polish law in force on
April 15, 2003 regarding the design of constant-time
traffic light programs [12].
2 MICROMODELING IN PTV VISSIM TOOL
The analysis was developed using the transport micro
modeling program PTV VISSIM [8], [13]–[16]. To
reproduce the state of the existing model as accurately
as possible, real data from two intersections in the city
of Gdynia were imported from [8]. Data on traffic
intensity and duration of individual signals in the
signaling program come from the TRISTAR intelligent
traffic control system in Gdynia. The model was made
using five main steps. The first step is to collect input
data for the model, such as traffic volume, vehicle
type structure, or the percentage of turning vehicles.
Step 2 is mapping the road network, adding a generic
vehicle structure, and adding traffic light programs.
Step 3 is directly related to the mapping of the
network because the following elements have been
added to the drawn network: adding traffic lights,
determining the right of way in disputed situations,
implementing pedestrian traffic generators, defining
vehicle routes and vehicle generators, and adding
vehicles and public transport routes. Step 4 is model
calibration with validation, the GEH method was
used here. The last step 5 is to measure parameters
such as travel times, queue lengths, or assessments of
entire intersections. The scheme of creating the model
is shown in Figure No. 1.
The model examined two intersections in the city
of Gdynia, connected by a flyover. The model lasts
two hours, of which the first half hour is for filling the
network with vehicles, then one hour for
measurement, and the last half hour. These
intersections differ from each other in terms of both
traffic volume and geometry and the number of
entrances. The first intersection is a connection
between the Morska Road and the Kwiatkowski
flyover. It is a central island junction with four inlets
and four outlets. Traffic at this intersection is a
combination of urban traffic from high-density
districts with traffic entering the city from the Tri-City
Ring Road. The second intersection is the connection
of Hutnicza Road with the Kwiatkowski Flyover. This
intersection is characterized by increased traffic of
heavy goods vehicles moving to and from Port
facilities or private companies related to container
transport. Junction 2 has only three entrances and
exits. The geometry of Junction 1 is shown in Figure 2.
Figure 3 shows a diagram of Junction 2.
Figure 1. PTV VISSIM - Micromodel framework [8]
Figure 2. Junction 1 - Morska - Kwiatkowskiego
Figure 3. Junction 2 - Hutnicza-Kwiatkowskiego
325
2.1 Data
Data on traffic volume are from the vehicle counting
system as part of the Intelligent Traffic Control
System in Gdynia. These data distinguish the
intensity also by vehicle type. Referring to Figures 2
and 3, where the entrances of two intersections have
been marked, Table 1 shows the traffic volume of
vehicles at the analyzed intersections along with the
percentage share of traffic directions.
Table 1. Traffic volume at the Junction 1 and Junction 2
________________________________________________
A – Intersection; B – Inlet; C – Traffic Flow (0-3600)
D – Direction; E – Volume; F – Traffic Flow (3600-7200)
G – Direction; H – Volume
________________________________________________
A B C D E F G H
________________________________________________
J1 N 980 R 359 868 R 387
S 0 S 0
L 353 L 481
U-turn 0 U-turn 0
E 1117 R 481 1254 R 505
S 444 S 553
L 169 L 170
U-turn 22 U-turn 25
W 1690 R 570 1525 R 529
S 763 S 664
L 356 L 332
U-turn 0 U-turn 0
S 297 R 144 298 R 151
S 0 S 0
L 153 L 147
U-turn 0 U-turn 0
J2 N 312 R 236 395 R 325
S 53 S 41
L 0 L 0
U-turn 0 U-turn 0
W 420 R 210 453 R 317
S 0 S 0
L 210 L 136
U-turn 0 U-turn 0
S 782 R 0 704 R 0
S 52 S 9
L 730 L 695
U-turn 0 U-turn 0
________________________________________________
Figure 4. Signal program - Junction 1
Figure 5. Signal Program - Junction 2
Traffic light programs from the Tristar Intelligent
Traffic Control System have been implemented in the
model. According to the schedule on the signaling
controller, the model was made based on the program
for the morning peak. Figure 4 below shows the
signaling program for Junction 1 and Figure No. 5
shows the signaling program for Junction 2.
2.2 Variants
The assumption of the model was to compare two
variants. Variant 1 is the existing state of intersections
with real data on traffic intensity and vehicle type
structure. The signaling program is also real. Variant
2, on the other hand, is a variant where the duration
of the green light for pedestrians has been reduced to
the minimum specified in the act on the design of
constant-time traffic lights. The general rule is that the
pedestrian speed is 1.4 m/s and then you can calculate
the time needed for the pedestrian to cross by using
the general formula t = s/v [s],where: s – the road that
must be walked by a pedestrian [m], v – walking
speed [m/s], t – the time needed by the pedestrian to
cross the street [s].
According to the rules, however, the total time
cannot be shorter than 8 seconds, it is the sum of four
seconds of a continuous green signal and four seconds
of a flashing signal [11] informing the pedestrian
about the termination of the pedestrian crossing
permit. After calculations, it turns out that the limit of
8 seconds applies in the case where there are three or
fewer lanes of 3.5 m wide each. In the case of four
lanes, the length of the green light is already extended
to 10 seconds. Table 2 presents the data of the
modeled pedestrian crossings: their length (the path
that the pedestrian must take), the duration of the
green signal in variant 1, and the duration of the
green signal in variant 2.
Table 2. Comparison of the length of the green time
________________________________________________
Length Variant 1 Variant 2
________________________________________________
J1_N_P1 10.5 82 8
J1_N_P2 10.5 40 8
J1_E_P2 7 13 8
J1_E_P1 14 54 10
J1_W_P2 10.5 15 8
J1_W_P1 10.5 42 8
J1_S_P1 10.5 38 8
J1_S_P2 7 76 8
J2_S_P1 7 47 8
J2_S_P2 10.5 44 8
J2_N_P1 10.5 47 8
J2_N_P2 7 16 8
J2_W_P1 7 25 8
________________________________________________
2.3 Results
In the PTV VISSIM tool, it is possible to use node
evaluation; you can record data from nodes of
microscopic and mesoscopic simulations in the Vissim
network. Node evaluation is especially used to
determine specific data from intersections without
first having to define all sections manually to
determine the data. One such assessment is the level
of service (LOS). [8] “Level of service (transport
quality): The levels of transport quality A to F for
movements and edges, a density value (vehicle
units/mile/lane). It is based on the result attribute
vehicle delay (average). The current value range of
326
vehicle delay depends on the level of service scheme
type of the signalized or non-signalized nodes. The
LOS in Vissim is comparable to the LOS defined in the
American Highway Capacity Manual of 2010” [20]. To
better illustrate the results, numerical values have
been assigned to specific LOS determinations and
thus LOS A corresponds to 1, LOS B corresponds to 2,
and similarly to the other values. The results of the
simulation showed that by shortening the duration of
the green light for pedestrians without shortening the
length of the entire phase, we will obtain improved
traffic conditions at intersections for vehicles. The
number of pedestrians was assumed to be constant in
both variants and amounts to 1,000 pedestrians per
hour for each pedestrian crossing. In both variants,
pedestrians could only cross during the green light. It
should also be noted that the traffic volume at the
inlets increased because more vehicles could arrive at
the intersection than in the case of variant 1. Despite
the increased traffic volume, the LOS parameter did
not deteriorate anywhere. The only case on the South
inlet of Junction 1 was where both traffic volume and
LOS did not change. In the remaining cases, the
changes were positive, and the obtained results
significantly improved the road traffic conditions, e.g.,
at Junction 2 from level 6 (i.e. F - the worst) there was
an improvement to level 2 (i.e. B). In Figure 6 these
changes are presented in a graph.
Figure 6. Junction 1 LOS comparison
Table 3. Level of service comparison on Junction 1 and
Junction 2
________________________________________________
Inlet Traffic Level of Inlet Traffic Level of
volume service at volume service at
of right right turn of right right turn
turn turn
[veh/h] [-] [veh/h] [-]
________________________________________________
J1 Variant 1 J1 Variant 2
________________________________________________
W 291 6 W 583 4
N 333 6 N 463 4
S 172 4 S 172 4
E 306 6 E 531 5
________________________________________________
J2 Variant 1 J2 Variant 2
________________________________________________
N 146 6 W 338 2
W 258 5 N 266 2
________________________________________________
Table 3 presents the detailed results of the
improvement of traffic conditions at the analyzed
intersections and shows changes in the traffic
intensity at the entrances. The improvement of traffic
conditions on right-turning roads has the greatest
impact on the analyzed intersections due to the
immediate vicinity of pedestrian crossings with
simultaneously allowed traffic. It is these places that
are critical when, despite the green light, drivers are
unable to leave the intersection due to pedestrian
traffic.
3 CONCLUSIONS
Although significantly better traffic results have been
obtained by reducing the duration of the green light
for pedestrians in fixed-time schemes, the changes
will not necessarily be readily implemented by local
road authorities. Due to the Smart Cities trend, or
restoring cities to people, shortening the pedestrian
crossing time would not be politically welcome.
However, the results obtained in this article show that
this solution brings significant improvements in traffic
conditions. Therefore, they should be used in special
situations where the traffic volume is so high that
even a short program change could be beneficial. It
should also be remembered that not everywhere such
changes can be introduced, in the case of places with a
high volume of traffic of people with limited mobility,
shortening the duration of the green light for
pedestrians is even discouraged due to the slower
speed of pedestrians. Currently, cities equipped with
intelligent traffic control systems can implement the
presented solution after additionally adding an
appropriate algorithm that calculates the forecast
profits in improving traffic conditions to the
deterioration of pedestrian comfort, which shortens
the crossing time.
ACKNOWLEDGEMENTS
This research was funded by Gdynia Maritime University—
WN/2023/PZ/10 and WN/2023/PZ/06 projects.
REFERENCES
[1] S. Lämmer and D. Helbing, “Self-Control of Traffic
Lights and Vehicle Flows in Urban Road Networks,” J.
Stat. Mech. Theory Exp., vol. 2008, 2008.
[2] S.-B. Cools, C. Gershenson, and B. D’Hooghe, “Self-
Organizing Traffic Lights: A Realistic Simulation,” in
Advances in Applied Self-organizing Systems, M.
Prokopenko, Ed. London: Springer London, 2006, pp.
41–50.
[3] M. Tubaishat, Y. Shang, and H. Shi, “Adaptive Traffic
Light Control with Wireless Sensor Networks,” 2007 4th
Annu. IEEE Consum. Commun. Netw. Conf. CCNC
2007, pp. 187–191, 2007.
[4] B. Toledo, V. Muñoz, J. Rogan, C. Tenreiro, and J.
Valdivia, “Modeling traffic through a sequence of traffic
lights,” Phys. Rev. E. Stat. Nonlin. Soft Matter Phys., vol.
70, no. 1, p. 16107, 2004.
[5] H. Kim, Y.-J. Han, and J.-S. Park, “Impacts of Special
Traffic Lights on Deep Learning Based Traffic Light
327
Recognition Systems,” J. Korean Inst. Commun. Inf. Sci.,
vol. 46, pp. 526–531, 2021.
[6] Z. Li, M. Chitturi, D. Zheng, A. Bill, and D. Noyce,
“Modeling reservation-based autonomous intersection
control in VISSIM,” Transp. Res. Rec., vol. 2381, no. 1,
pp. 81–90, 2013.
[7] T.-W. Yeh, H.-Y. Lin, and C.-C. Chang, “Traffic Light and
Arrow Signal Recognition Based on a Unified Network,”
Appl. Sci., vol. 11, p. 8066, 2021.
[8] M. Ziemska-Osuch and D. Osuch, “Modeling the
Assessment of Intersections with Traffic Lights and the
Significance Level of the Number of Pedestrians in
Microsimulation Models Based on the PTV Vissim
Tool,” Sustainability, vol. 14, no. 14, p. 8945, 2022.
[9] S. Gaca, W. Suchorzewski, and M. Tracz, Inżynieria
Ruchu Drogowego Teoria i praktyka, 1st ed. Warszawa:
Wydawnictwa Komunikacji i Łączności, 2014.
[10] M. Ziemska, “The idea of using intelligent transport
systems and coordination of traffic lights,” Pr. Wydz.
Nawig. Akad. Morskiej w Gdyni, vol. 29, pp. 107–112,
2014.
[11] Dziennik Ustaw Poz 2311 Załącznik nr2,
“SZCZEGÓŁOWE WARUNKI TECHNICZNE DLA
ZNAKÓW DROGOWYCH POZIOMYCH I WARUNKI
ICH UMIESZCZANIA NA DROGACH.” 2015.
[12] ROZPORZĄDZENIE MINISTRA INFRASTRUKTURY
z dnia 3 lipca 2003 r. w sprawie szczegółowych
warunków technicznych dla znaków i sygnałów
drogowych oraz urządzeń bezpieczeństwa ruchu
drogowego i warunków ich umieszczania na drogach
(Dz. U. z dnia 23 grudnia 2003 r.)
[13] J. Oskarbski and L. Gumińska, “The application of
microscopic models in the study of pedestrian traffic,” in
MATEC Web of Conferences, 2018.
[14] D. Gunarathne, N. Amarasingha, and V.
Wickramasighe, “Traffic Signal Controller Optimization
Through VISSIM to Minimize Traffic Congestion, CO
and NOx Emissions, and Fuel Consumption,” Sci. Eng.
Technol., vol. 3, 2023.
[15] M. Fellendorf and P. Vortisch, “Microscopic traffic flow
simulator VISSIM,” in International Series in Operations
Research and Management Science, 2010.
[16] Y. Zheng, Q. Shao, Y. Zhang, and S. Zhang, “Simulation
on Two Types of Improved Displaced Left-Turn
Intersections Based on VISSIM,” 2021, pp. 213–221.
[17] H. Farah et al., “Modeling Automated Driving in
Microscopic Traffic Simulations for Traffic Performance
Evaluations: Aspects to Consider and State of the
Practice,” IEEE Trans. Intell. Transp. Syst., vol. PP, pp.
1–17, 2022.
[18] W. Walerjanczyk, “Automatyzacja procesu weryfikacji
mikroskopowego modelu ruchu drogowego dla
aglomeracji poznańskiej,” Logistyka, vol. 4, 2014.
[19] A. C. Dey, S. Roy and M. A. Uddin. (2018);
CALIBRATION AND VALIDATION OF VISSIM
MODEL OF AN INTERSECTION WITH MODIFIED
DRIVING BEHAVIOR PARAMETERS. Int. J. of Adv.
Res. 6 (Dec). 107-112] (ISSN 2320-5407).
[20] “PTV Vissim and PTV Viswalk Help,” PTV Planung
Transport Verkehr GmbH, Karlsruhe, Germany, 2022.
[Online]. Available: https://cgi.ptvgroup.com/vision-
help/VISSIM_2022_ENG/Content/11_Auswertungen/Au
sw_a_ausfuehren.htm. [Accessed: 17-Jun-2022].
