MODULE-3

TRAFFIC FLOW THEORIES: Speed, flow and density relationships and

concept of LOS from them. , Traffic flow theory - Green Shield theory–
Goodness of fit, correlation and regression analysis (linear only) – Queuing
theory and its applications , Car following theory and relevant problems on
above.

Important Definition
1) Speed(v): Distance traversed by a vehicle in unit time
(mt/sec)
2) Time mean speed(vt): The average of the speeds
measured at one point in space over a period of time
(kmph)
3) Space mean speed (vs): The average of the speeds
measured at an instant of time over a space (kmph)
4) Flow or volume (Q): The number of vehicles passing a
specified point during a stated period of time (veh/hr)
5) Concentration or Density (K): Number of vehicles
present in the stated length of road at an instant
6) Space Headway (s): It is the distance between the
fronts of successive vehicles (m)
7) Time Headway (h): It is the time interval between the
passage of the fronts of successive vehicles at a
specified point (sec)

Relationship between the variables

a) Space mean speed(vs), Flow(Q) and Density(K)
𝑸
𝒗𝒔 = − − − − − (𝟏)
𝑲
But K (Density or Concentration) is the reciprocal of the
space headway(s) (s/1000 when expressed in km)
𝑸 𝑸𝑺
Therefore 𝒗𝒔 = 𝟏 = 𝟏𝟎𝟎𝟎 − − − − − (𝟐)
𝒔
𝟏𝟎𝟎𝟎
„s‟ Space headway is related to „h‟ Time headway
𝒉
Therefore 𝑺 = ∗ 𝒗𝒔 ∗ 𝟏𝟎𝟎𝟎 − − − − − (𝟑)
𝟑𝟔𝟎𝟎

Problem
Speed observations from a radar speed meter have been
taken, giving the speeds of the subsidiary streams
composing the flow along with the volume of traffic of each
subsidiary stream. The readings are as under:
 Volume of
Speed
subsidiary stream
(km/hr)
(Veh/hr)
2-5 1
6-9 4
10-13 0
14-17 7
18-21 20
22-25 44
26-29 80
30-33 82
34-37 79
38-41 49
42-45 36
46-49 26
50-53 9
54-57 10
58-61 3
Calculate ;
(i) Time mean speed
(ii) Space mean speed
(iii) Variance about space mean speed

Solution
The mid point of speed class “vi” and “qi” are tabulated
below, along with qivi and ki = qi/vi
Volume of
Speed Speed mid-
subsidiary stream ("qivi") ("ki=qi/vi")
(km/hr) Point ("vi")
(Veh/hr)"qi"
2-5 1 3.5 3.5 0.29
6-9 4 7.5 30 0.53
10-13 0 11.5 0 0.00
14-17 7 15.5 108.5 0.45
18-21 20 19.5 390 1.03
22-25 44 23.5 1034 1.87
26-29 80 27.5 2200 2.91
30-33 82 31.5 2583 2.60
34-37 79 35.5 2804.5 2.23
38-41 49 39.5 1935.5 1.24
42-45 36 43.5 1566 0.83
46-49 26 47.5 1235 0.55
 50-53 9 51.5 463.5 0.17
54-57 10 55.5 555 0.18
58-61 3 58.5 175.5 0.05
450 15084 14.93

∑qi=Q =450veh/hr
∑viqi = 15084
∑ki= k = 14.93
Time mean Speed
vt = (∑ viqi)/Q = 15084/450 = 33.52 km/hr

Space mean Speed

vs = Q/k = 450/14.93 = 30.1 km/hr

vt = vs + (σs2)/ vs

σs2 = (33.52 - 30.1) 30.1 = 102.94

Problem
Twenty five spot speed observations were taken and
were as under:
50,40,60,54,45,31,72,58,43,52,46,56,43,65,33,69,34,5
1,47,41,62,43,55,40,49
Calculate:
(i) Time mean speed
(ii) Space mean speed
and verify the relation between the two.
Solution:
∑vi = 50+40+60+54+45+31+72+58+43+52+46+56+
43+65+33+69+34+51+47+41+62+43+55+40+49
= 1239
Time mean speed vt = 1239/25 = 49.56 kmph
 ∑(vi - 49.56)2 = 2840.16

∑1/vi = 1/50+1/40+1/60+1/54+1/45+1/31+1/72+1/58+
1/43+1/52+1/46+1/56+1/43+1/65+1/33+1/69+1/34+1/
51+1/47+1/41+1/62+1/43+1/55+1/40+1/49
= 0.529

Space mean speed vs = 25/0.529 = 47.26 kmph

Variance s2 = 2840.16/(25-1) = 118.34

The relationship between vt and vs is:

𝑠2
𝑣𝑡 = 𝑣𝑠 +
𝑣𝑠

118.34
𝑣𝑡 = 47.26 +
47.26

𝑣𝑡 = 49.76 kmph

This may be compared with the value of Time mean speed

49.56 kmph obtained earlier.

The Queueing theory

Elements of queueing theory
1) Arrival pattern: The arrival pattern can be
i) Regular – with a rectangular distribution
ii) Random – with a Poissonian or normal distribution
2) Service facility
i) Single channel
ii) Multiple channel with one queue for each channel

iii) Single queue feeding into multiple channels

iv) Number of queues greater than one but less than

the number of channels

v) Multiple channels in series

 vi) Combination of multiple service channels, some in
series and some in parallel
3) The que discipline characteristics can be
i) Join a single queue and wait
ii) The queue is of fixed capacity and further
costumers are turned away
iii) Not to join the queue at all
iv) Join the queue and leave because not served
before a certain time
v) In a system of multiple queues, to change from
one que to another
4) The order of service can be
i) First in first out
ii) Last in first out
iii) No regular system or priority
iv) A system of priority
5) Service time characteristics
i) Rectangular distribution
ii) Random, exponential, Poissonian distribution
6) Training and familiarity determine whether the service
is affected by the number of costumers and whether
there is any difference between different operators
7) The service discipline can mean serving one costumer
at a time or serving many customers at a time

Assumptions made in simple Queueing approach

a) The system is in a steady state and has settled down
𝐀𝐯𝐞𝐫𝐚𝐠𝐞 𝐫𝐚𝐭𝐞 𝐨𝐟 𝐚𝐫𝐫𝐢𝐯𝐚𝐥 ƛ
𝐓𝐫𝐚𝐟𝐟𝐢𝐜 𝐢𝐧𝐭𝐞𝐧𝐬𝐢𝐭𝐲 𝛒 =
(𝐀𝐯𝐞𝐫𝐚𝐠𝐞 𝐫𝒂𝒕𝒆 𝒐𝒇 𝒔𝒆𝒓𝒗𝒊𝒄𝒆(µ)
 𝝀
𝑻𝒉𝒖𝒔 𝝆 = <1
µ
b) The number of costumers is „discrete‟
c) The population of potential customers is infinite
d) The arrivals are random in nature and Poissonian
distribution applies
e) There will be no simultaneous arrivals
f) There will be single service channel, separate
modification of the approach are available for
multiple channel
g) The queue is single and is of infinite capacity
h) The order of service is first in first out
i) There is a single follow on service discipline
j) The service time varies and follows an exponential
distribution with a mean rate „µ‟

Relationship between traffic intensity, mean rate of arrival

and mean rate of service
𝝀
𝝆 = < 1-------(1)
µ
Where ′𝝆′ 𝒊𝒔 𝒕𝒓𝒂𝒇𝒇𝒊𝒄 𝒊𝒏𝒕𝒆𝒏𝒔𝒊𝒕𝒚
′𝝀′ 𝒊𝒔 𝒎𝒆𝒂𝒏 𝒓𝒂𝒕𝒆 𝒐𝒇 𝒂𝒓𝒓𝒊𝒗𝒂𝒍 𝒗𝒆𝒉/𝒖𝒏𝒊𝒕𝒕𝒊𝒎𝒆
′µ′ 𝒊𝒔 𝒎𝒆𝒂𝒏 𝒓𝒂𝒕𝒆 𝒐𝒇 𝒔𝒆𝒓𝒗𝒊𝒄𝒆 (𝒗𝒆𝒉/𝒖𝒏𝒊𝒕𝒕𝒊𝒎𝒆)

𝝆
𝒏= − − − − − (𝟐)
𝟏−𝝆
Where
′𝒏′ 𝒊𝒔 𝒆𝒙𝒑𝒆𝒄𝒕𝒆𝒅 𝒏𝒖𝒎𝒃𝒆𝒓 𝒐𝒇 𝒄𝒖𝒔𝒕𝒐𝒎𝒆𝒓𝒔 𝒊𝒏 𝒕𝒉𝒆 𝒔𝒚𝒔𝒕𝒆𝒎
𝑨𝒗𝒆𝒓𝒂𝒈𝒆 𝒏𝒖𝒎𝒃𝒆𝒓 𝒐𝒇 𝒗𝒆𝒉𝒊𝒄𝒍𝒆𝒔 𝒊𝒏 𝒕𝒉𝒆 𝒒𝒖𝒆𝒖𝒆

𝝆𝟐
𝒒= − − − − − (𝟑)
𝟏−𝝆
Where „𝒒’ is average length of queue
𝑵𝒖𝒎𝒃𝒆𝒓 𝒐𝒇 𝒗𝒆𝒉𝒊𝒄𝒍𝒆𝒔 𝒊𝒏 𝒕𝒉𝒆 𝒒𝒖𝒆𝒖𝒆

𝟏
𝒅= − − − − − (𝟒)
µ 𝟏−𝝆
Where „𝒅’ is average time spent by the vehicle in the system (sec)

𝝆
𝝎= − − − − − −(𝟓)
µ 𝟏−𝝆
Where ′𝝎′ is average time spent by the vehicle in the queue (sec)

Applications of Queueing Theory

Queueing theory can be applied to determine
i) The average (mean) and distribution of the queue
length, the waiting time, the time spent in the
queue and the number of vehicles in the queue.
ii) The probability of
(a) There being „n‟ units in the system
(b) There being more than „n‟ units in the system
(c) The waiting time being greater than „t‟
(d) The waiting time being between „t‟ and „t+δt‟
(e) A customer having spent time „t‟ in the system
(f) The service channel being idle

Limitations of queueing theory

Queueing theory cannot be applied if the que is
i. Not in a steady state and if it has not “settled down”
ii. In a “peaking situation” or “transient situation”
𝝀
That is 𝝆 = ≮ 𝟏
µ
iii. Not infinite
iv. Having simultaneous arrivals
v. Multiple and having restricted capacity
 vi. Having order of service other than first in first out
vii. A multiple follow on service discipline
viii. Having a fixed service time

Problem
A toll booth at the entrance to a bridge can handle
120veh/hr, the time to process a vehicle being
exponentially distributed. The flow is 90veh/hr with a
Poissonian arrival pattern. Determine
(i) The average number of vehicles in the system
(ii) The length of the queue
(iii) The average time spent by the vehicle in the
system
(iv) The average time spent by the vehicle in the queue

Solution
𝟗𝟎 𝟏
𝝀= = 𝒗𝒆𝒉/𝒔𝒆𝒄 Mean rate of arrival
𝟑𝟔𝟎𝟎 𝟒𝟎
𝟏𝟐𝟎 𝟏
µ= = 𝒗𝒆𝒉/𝒔𝒆𝒄 Mean rate of service
𝟑𝟔𝟎𝟎 𝟑𝟎
𝟏
𝝀 𝟒𝟎 𝟑
𝝆= = 𝟏 = = 𝟎. 𝟕𝟓Traffic intensity
µ 𝟒
𝟑𝟎
a) The average number of vehicles in the system
𝝆 𝟎. 𝟕𝟓
𝒏= = = 𝟑 𝒏𝒖𝒎𝒃𝒆𝒓𝒔
𝟏 − 𝝆 𝟏 − 𝟎. 𝟕𝟓
b) The length of queue (Number of vehicles in the
queue)
𝝆𝟐 𝟎. 𝟕𝟓𝟐
𝒒= = = 𝟐. 𝟐𝟓 𝒏𝒖𝒎𝒃𝒆𝒓𝒔
𝟏 − 𝝆 𝟏 − 𝟎. 𝟕𝟓
c) The average time spent by the vehicle in the system
 𝟏 𝟏
𝒅= = 𝟏
= 𝟏𝟐𝟎 𝒔𝒆𝒄𝒐𝒏𝒅𝒔
µ 𝟏−𝝆 𝟏 − 𝟎. 𝟕𝟓
𝟑𝟎

d) The average time spent by the vehicle in the queue

𝝆 𝟎. 𝟕𝟓
𝝎= = 𝟏
= 𝟗𝟎 𝒔𝒆𝒄𝒐𝒏𝒅𝒔
µ 𝟏−𝝆 𝟏 − 𝟎. 𝟕𝟓
𝟑𝟎

Linear Regression and Correlation

The technique of predicting the values of one variable
(called the dependent variable) from measurements of the
other (called the independent variable) is called Regression
analysis. If the relationship between the independent and
the dependent variable is linear, the analysis is known as
“Linear Regression”.

If the independent variables are two or more in number,

the analysis is known as multiple linear regression
If the dependent variable is denoted by „Y‟ and the
independent variable by „X‟ the linear relationship between
the two can be of the following form

Y = a + bX + c

Where „a‟ and „b‟ are constants and „c‟ is the term denoting
the random error. The random error is very small if the
two variables are closely related
Thus
Y = a + bX
Problem
The speed and concentration of vehicles in a traffic stream
were observed and the following data were obtained

Concentration veh/km 5 10 15 20 25 30 35 40 45 50
Speed kmph 72 68 61 52 47 39 32 27 20 13

Find the regression equation for determining the speed

from concentration
Solution
Let the regression equation be in the form of
Y = a + bX
Where Y is speed in kmph
X is concentration veh/km
„a‟ and „b‟ are constants
Sl Concentration Speed (kmph)
Xi*Yi Xi 2
No (veh/km) Xi Yi
1 5 72 360 25
2 10 68 680 100
3 15 61 915 225
4 20 52 1040 400
5 25 47 1175 625
6 30 39 1170 900
7 35 32 1120 1225
8 40 27 1080 1600
9 45 20 900 2025
10 50 13 650 2500
Total 275 431 9090 9625

𝟐𝟕𝟓
𝒙= = 𝟐𝟕. 𝟓
𝟏𝟎
 𝟒𝟑𝟏
𝒚= = 𝟒𝟑. 𝟏
𝟏𝟎
Using the equation
∑𝒙𝒊 𝒚𝒊 − 𝒏𝒙𝒚
𝒃=
∑𝒙𝒊 𝟐 − 𝒏𝒙𝟐

𝟗𝟎𝟗𝟎 − 𝟏𝟎 ∗ 𝟐𝟕. 𝟓 ∗ 𝟒𝟑. 𝟏 𝟐𝟕𝟔𝟐

= = − = −𝟏. 𝟑𝟒
𝟗𝟔𝟐𝟓 − 𝟏𝟎 ∗ 𝟐𝟕. 𝟓𝟐 𝟐𝟎𝟔𝟐. 𝟓

Using the equation

a=Y-b*X
= 43.1 – (-1.34)(27.5) = 79.95

The Regression equation

Y = 79.95 – 1.34 * X
80

70

60
Speed (kmph) Yi

50

40
Speed (kmph) Yi
30 Linear (Speed (kmph) Yi)

20

10
y = -1.3394x + 79.933

0
0 5 10 15 20 25 30 35 40 45 50 55

Concentration (veh/km)

Level of Service (LOS)

Level of service is a qualitative measure describing the
operational conditions within a traffic stream and their
perception by material and/or passenger.
The following are the factors which might be considered in
evaluating the Level of Service.
1) Speed and travel time, including the operating speed
and overall travel time
2) Traffic interruptions or restrictions, number of stops
per kilometer, delays involved and the speed change
necessary y to maintain pace in the traffic stream.
3) Freedom to maneuver to maintain the desired
operating speeds
4) Driving comfort and convenience reflecting the
roadway and traffic conditions
5) Economy with due considerations of operating cost of
the vehicle

The operating conditions for the six levels of service

selected by the manual are given below.Level-A
representing the highest and Level-F representing the
Lowest.

Level of Service-A: Free flow, with low volumes of traffic

Level of Service-B: Stable flow, with operating speeds

beginning to be restricted. Drivers still have reasonable
freedom
Level of Service-C: Stable flow, but speed and
maneuverability are closely controlled by higher volumes,
relatively satisfactory operating speed, suitable for urban
areas
Level of Service-D: Approaches unstable flow, with
tolerable operating speed, fluctuation in volume and
temporary restriction to flow may cause substantial drop in
operating speed, drivers have little freedom, comfort and
convenience are low

Level of Service-E: Lower operating speed, volume at or

near to capacity of the highway, Flow is unstable and there
may be stoppages of momentary duration

Level of Service-F: Forced flow operating at lower speeds.

Speeds are reduced substantially and stoppages may occur
for short and long durations

LEVEL OF SERVICE (LOS) DEFINED

The LOS of Mid-Block Section
Factors affecting Capacity and Level of Service
 Road way factors
 Traffic Factors

Roadway Factors: Roadway factors pertaining to

restrictive physical features of a road such as
a) Lane width
b) Lateral clearance; Lateral obstruction such as
retaining walls, abutments, signposts, lightpoles and
parked cars located closer than 1.83m from the edge
of traffic lane reduces capacity
c) Shoulders; badly maintained shoulders reduces
capacity of the roadway
d) Surface condition; deteriorated and poorly
maintained pavement adversely affect LOS
 e) Grades; Grades adversely affect the speed of
vehicles, especially trucks and thus influence the
capacity

Traffic Factors:
a) Mixed traffic condition; different class of vehicles
with different operating speed occupying the same
lane reduces the average speed of vehicles.
b) Power of vehicle
c) Dimension of vehicle
d) Acceleration and Deceleration of vehicle

Problem
The following spot speed observations were obtained
(kmph):
50,40,60,54,45,31,72,58,43,52,46,56,43,65,33,69,34,51,
47,41,62,43,55,40 and 49. Determine (i) Time mean
speed (ii) Space mean speed (iii) Verify the relationship
between the two.
Solution
(i) Time mean speed (vt)
∑vi=50+40+60+54+45+31+72+58+43+52+46+5
6+43+65+33+69+34+51+47+41+62+43+55+40
+49 =1239
Therefore vt = ∑vi/n = 1239/25 = 49.56 kmph

∑(vi - vt)2 = ∑(vi – 49.56)2 = 2972.41

(ii) Space mean speed (vs)

∑(1/vi) = (1/50)+ (1/40)+ (1/60)+ (1/54)+
(1/45)+ (1/31)+ (1/72)+ (1/58)+ (1/43)+
(1/52)+ (1/46)+ (1/56)+ (1/43)+ (1/65)+
 (1/33)+ (1/69)+ (1/34)+ (1/51)+ (1/47)+
(1/41)+ (1/62)+ (1/43)+ (1/55)+ (1/40)+ (1/49)
= 0.529

Therefore vs = n/∑(1/vi) = 25/0.529 = 47.26 kmph

Variance s2 = (∑(vi – 49.56)2)/(n-1)

= 2972.41/(25-1) = 123.85
(iii) Relationship between vt and vs

Vt = vs + (s2/ vs) = 47.26 + (123.85/47.26)

= 49.88 kmph
Comparable with 49.56 kmph obtained

Fundamental Diagrams of Traffic Flow

Free Flow Speed

Kj Jamming
Concentration

Mean Free Speed

Qmax Capacity

Kj Jamming
Concentration
Linear Relationship between Speed and
Concentration
When speed is zero the concentration is maximum and this
corresponds to „Kj‟, the jamming concentration, also when
the concentration is zero, speed is maximum and is the
free speed „vs‟. the exact shape of the curve joining these
two points depends on a number of factors. A straight line
relationship yields the parabolic shape of the curve,
relating flow and concentration. It has been indicated in
the figure that linear relationship between speed and
concentration is one of the possibilities.
𝒗𝒔𝒇
𝒗𝒔 = 𝒗𝒔𝒇 − 𝒌
𝒌𝒋

Green Shield’s Theory

Green Shield assumed a linear relationship between speed
and density
(i) Speed and Density
𝒗𝒇
𝒗 = 𝒗𝒇 − 𝒌−−−−− 𝟏
𝒌𝒋
Where
„v‟ and „k‟ are mean speed and density
„kj‟ is jamming density
„vf‟ is free speed
(ii) Relationship between Flow and Density
𝒒 = 𝒌𝒗 − − − − − − 𝟐
Substitute (1) in (2)
𝒗𝒇
𝒒 = 𝒌 𝒗𝒇 − 𝒌
𝒌𝒋
𝒗𝒇 𝟐
𝒒 = 𝒌𝒗𝒇 − 𝒌 −−−−−− 𝟑
𝒌𝒋
 𝒒
We know 𝒗 = 𝒌
Substituting in (1)
𝒒 𝒗𝒇
= 𝒗𝒇 − 𝒌
𝒌 𝒌𝒋
Therefore
𝒗𝒇 𝟐
𝒒 = 𝒌𝒗𝒇 − 𝒌 −−−−−− 𝟒
𝒌𝒋
This relationship between flow and density is
parabolic in shape.
(iii) Relationship between Speed and Flow
𝒒
Substituting 𝒌 = 𝒗 in eqn (1)
𝒗𝒇 𝒒
𝒗 = 𝒗𝒇 −
𝒌𝒋 𝒗
𝒌𝒋 𝟐
𝒒 = 𝒗𝒌𝒋 − 𝒗 − − − − − −𝟓
𝒗𝒇
This relationship between Speed and Flow is
Parabolic in shape.

To find density at maximum flow differentiate equation (4)

and equate it to zero
𝒅𝒒
=𝟎
𝒅𝒌

𝒗𝒇
𝒗𝒇 − ∗ 𝟐𝒌 = 𝟎
𝒌𝒋
Denoting the density corresponding to maximum flow as
„k0‟
𝒌𝒋
𝒌𝟎 =
𝟐
Therefore
𝟐
𝒌𝒋 𝒗𝒇 𝒌𝒋
𝒒𝒎𝒂𝒙 = 𝒗𝒇 −
𝟐 𝒌𝒋 𝟐

𝒗𝒇 𝒌𝒋
𝒒𝒎𝒂𝒙 =
𝟒

𝒗𝒇 𝒌𝒋
𝒗𝟎 = 𝒗𝒇 −
𝒌𝒋 𝟐

𝒗𝒇
𝒗𝟎 =
𝟐
Therefore speed at maximum flow is half of the free speed