Optimization of Patrol Manpower Allocation Using Goal Programming Approach -A Case Study

K. Venkata Subbaiah et al. Int. Journal of Engineering Research and Applications www.ijera.com
ISSN: 2248-9622, Vol. 5, Issue 10, (Part - 3) October 2015, pp.142-153
www.ijera.com 142 | P a g e
Optimization of Patrol Manpower Allocation Using Goal
Programming Approach -A Case Study
K. Venkata Subbaiah*, G. Daniel Vivek**, Ch. Suresh***
*(Department of Mechanical Engineering, Andhra University College of Engineering (A), Visakhapatnam
530003)
** (Department of Mechanical Engineering, Andhra University College of Engineering (A), Visakhapatnam
530003
***(Department of Mechanical Engineering, Andhra University College of Engineering (A), Visakhapatnam
530003)
ABSTRACT
One of the most difficult tasks of the patrol administrators is allocation of manpower; i.e. determining the. most
effective level of operational manpower for patrol tasks. Typically, administrators resolve the allocation
problem by relying on prior statistical data and by employing subjective analysis. In general, only limited
systematic analyses have been applied to the problem. This thesis presents a non linear goal programming model
for allocating patrolmen to road segments within a patrol region. The model is demonstrated via a case example
of the East section of Visakhapatnam. The results of the model are valuable to the patrol administrator for
considering departmental goals and priority structure, in addition to available historical data, in the assignment
of patrol manpower for a given urban area.
Keywords - Goals, Patrolmen Allocation, Non Linear Goal Programming
I. Introduction
The primary functions of the patrol department
are to enforce laws regulating the operation of
vehicles on unincorporated roadways, to
investigate accidents, assist motorists, provide
emergency services, direct traffic, and provide
limited criminal law enforcement assistance. For the
most part there are only slight variations in these
duties among the various patrol organizations within
the reach.
In order to be successful in their primary task of
traffic law enforcement, the patrol attempts to
obtain voluntary compliance with traffic rules and
regulations by all motorists and pedestrians.
Since many drivers are inclined to violate laws unless
strong deterrents are developed and implemented,
enforcement needs to be intelligently, directly, and
efficiently applied.
II. Literature Review
 Raja Jayaraman et al. (2015) stated that
sustainable development is an important and
strategic priority for global nations which
requires simultaneously satisfying multiple
conflicting objectives involving social,
economic, energy, and environmental
constraints. He proposed a goal programming
model that integrates efficient allocation of
labour resources to achieve sustainability
objectives relating to economic, energy and
environmental goals of the United Arab Emirates
by the year 2030.
 Aleksander Aristovnik, Janko Seljak, Jernej
Mencinger (2014) have discussed the
methodology to allocate the patrol manpower
according to the efficiency scores obtained by
applying it in a given local domain or area. A
three-stage Data Envelopment Analysis (DEA)
technique is presented and then applied to
measure the relative efficiency of police-work-
related data for selected police units at the local
level.
 Taylor B. W. et al. (1985) described a modeling
approach that can be used to deploy state
highway patrol cars to the road segments
comprising a highway patrol district.
Specifically, an integer nonlinear goal
programming model was used and demonstrated
within the context of a hypothetical case
example. The model reflected the fact that the
relationships between unit deployment and
performance measures are often non linear as
additional patrol units are assigned to a road
segment, performance will increase but at a
decreasing rate.
III. Goal Programming
The theory of Goal Programming was well
defined by Nobel Laureate Herbert A. Simon's theory
of "Satisfaction". He stated that today's manager does
not attempt to optimize, instead, the manager
RESEARCH ARTICLE OPEN ACCESS

(decision maker) tries to satisfy. It is a powerful tool
that draws upon the highly developed and tested
technique of linear programming but at the same
provides a simultaneous solution to a complex system
of competing objectives. A major advantage of goal
programming over other techniques in dealing with
real-world decision problems is that it reflects the
way manager actually makes decision. In other words
goal programming allows the decision maker to
incorporate the environmental, organizational and
managerial considerations into the model through
goal levels and priorities. Goal programming is an
extension of linear programming and provides a
simultaneous solution to a complex system of
competing objectives. This concept was originally
developed by Charnes and Cooper (1961), and its
methodology has been further refined by Ijiri (1965),
Lee (1972) and others. In goal programming, instead
of trying to maximize or minimize the objective
function directly, the deviation between goals and
what can be achieved within the given set of
constraints are to be minimized
The basic approach of goal programming is to
establish a specific numeric goal for each of the
objectives, formulate an objective function for each
objective, and then seek a solution that minimizes the
(weighted) sum of deviations of these objective
functions from their respective goals.
3.1 Problem formulation
The Procedure for formulating a goal
programming problem involves the following crucial
steps:
1.) Determining the Decision Variables.
2.) Formulation of the Objective Function.
3.) Absolute Objectives
4.) Formulation of Achievement function
5.) Priority Structure
3.1.1 Determining the Decision Variables
The decision variables are those factors over
which the decision-maker has control, for this reason
they are sometimes known as the 'control' variables.
3.1.2 Formulation of the objective function
In this context, the goals or objectives are
designated by G. Each objective function in turn must
be expressed as a function of the decision variables
i.e.,
 
i
i i i
i
b
G f x b
b


 

Where  if x is the function of decision variables
associated with the ith
objectives and ib is the
associated right-hand side value and reflects the value
that  if x must satisfy exceed or be less than. In
addition, every objective is accompanied, on the left-
hand side, by a negative and a positive deviation
variable in and ip respectively. The value of ip
for any choice of x, reflects the positive deviation.
Consequently, each objective will be of the following
final form.
  iiii bpnxf  i=1, 2,.....,m.
3.1.3 Absolute Objectives
If for any objective, either ip or in must equal
zero for the solution to be implementable, such an
objective is denoted as an absolute i.e., One that must
be satisfied. This can be accomplished by assigning
top priority p, to those absolute objectives.
3.1.4 Formulation of the Achievement Function
The above discussion for formulating the
objection function is considered and followed by the
formulation of the achievement function. Thus, the
achievement function will then take on the form
shown below:
Minimise
      pngpngpnga k ,..,.........,,, 21
Or alternatively
Minimise  kaaaa ...,........., 21 , Where
 pnga kk , and  pngk , is a linear function
of the deviation variable
3.1.5 Priority Structure
It is assumed that objectives may be arranged
according to ordered sets having been assigned such
priorities that 1 kk pp . These preemptive priorities
do not represent numbers and should never be
replaced with numbers. For solving any Goal
Programming model firstly, it involves achieving
highest priority goal, before any goals of the lower
priority are considered. Once the highest priority goal
is attained to the fullest extent possible, the Goal
Programming model proceeds to find a satisfactory
level to next highest priority goal, and so on.
However, it is not always possible to achieve every
goal to the extent desired by the decision maker. But
the advantage of linear Goal Programming over the
ordinary Linear Programming is that, it seeks, within
the given set of constraints, to minimize the deviation
from the established goals targets, while the ordinary
Linear Programming seeks to minimize or maximize
certain goal directly.
IV. Model Formulation
A Non-linear Goal Programming (NLGP) model
was developed for allocation of patrolmen with
Multiple goals. The goals include, Total patrolmen

allocation, Total budget allocation, Minimum shift
requirement reduction, Estimated accident reduction,
Accident reduction rate in high accident areas. Most
of the road segments in urban areas have high traffic
density due to the limited road area. . All the Road
segments considered are Intra-city road segments. In
this context, the total working hours has been divided
into three shifts and the shifts are given as
First Shift: 7:00 am to 3:00 pm
Second Shift: 3:00 pm to 11:00 pm
Third Shift: 11:00 pm to 7:00 am
4.1 Problem Formulation
The formulation of the model of the problem
involves indentifying the goals and creating a priority
structure to build the model. Hence, the formulation
of the problem includes
1. Total patrolmen goal
2. Total budget goal
3. Minimum shift requirement goal
4. Estimated accident reduction goal
5. Accident rate reduction in high accident prone
road segments goal
6. Priority Structure
4.1.1 Total Patrol Men Goal
For effective functioning of patrolling activities,
it is desired to allocate the patrolmen properly. Since
daily the available patrolmen are limited, it is
desirable to utilize the available patrolmen for all the
road segments in the city to achieve the target. Hence
it is given as,
  PddX
n
i j
ij 

1
11
3
1
(i)
Where P is the total number of Patrolmen available
for allocation, i= Road Segments 1, 2,.......n and j=
Shits 1, 2, 3.
4.1.2 Total Budget Goal
For immediate maintenance of the patrol vehicle
and fuel charge. The Government, provides some
amount of money to each vehicle daily. Hence it is
given as:
 



n
i j
iji PddXC
1
22
3
1
(ii)
Where P is the maximum budget available. Ci is the
variable cost (₹) per patrol vehicle per day with
respect to the corresponding road segment i
.
4.1.3 Minimum Shift Requirement Goal
All the segments have different traffic density
and accident frequency as mentioned before. Hence,
In order to provide emergency services, a minimum
number of patrolmen required for a.) Road Segments
which are linked together and b.) Individual Road
Segments. The minimum no. of patrolmen required
are estimated according to the traffic density and
accident frequency of each road segment. Hence, the
goal equations are given as follows:
a.) MddXXX
j
jijiji


3
1
33.......)( '''
(iii)
b.) NddX
j
ij 

3
1
44)( (iv)
Where a) corresponds to the goal equation of a
junction or road segments that are linked together and
b) corresponds to the individual road segments. M
and N represent the respective requirement of police
patrolmen for the given junction and individual road
segment.
4.1.4 Estimated Total Accident Reduction Goal
The reduction in accident rate (measured as
accidents per Km travelled) depends on the number
of patrolmen assigned to the road segment during the
shift j. The accident rate for a particular road segment
will be reduced with the proper of allocation
patrolmen. So, the goal equation will be of the form:
nnn
n
i j ij
ij
iij Qdd
X
b
la
L

























 
 
1
3
1
1 (v)
Where, nQ is the estimated reduction in the
number of accidents per Km.
il is the length of the respective road segment i. L is
the total length of all the road segments.
The values of parameters ija and ijb can be
estimated by defining the relationship for two point
estimates i.e., the number of patrolmen per shift and
the number of accidents per Km and solving the
obtained equations simultaneously.
4.1.5 High Accident Prone Road Segments Goal
In addition to achieve the total accident rate
reduction target of all the road segments as
mentioned before, it is required to achieve the
individual accident rate reduction target of each road
segment. The goal equations are of the form
nnn
j ij
ij
ij gdd
X
b
a 








 


3
1
(vi)
where i= 1, 2,....n. and ng is the corresponding
accident reduction rate of road segments.
Hence, to carry out the optimization of
patrolmen allocation, the following goals have been
considered with respective priorities:
1) Estimated Total accident reduction and Accident
reduction rates in high accident prone road
segments. (P1)
2) Total number of Patrolmen and the minimum
shift requirement.(P2)
3) Total Budget expenditure per day (P3)

In this decision making situation, four cases
have been considered accordingly with respect to the
goal considered as above. This allocates the
patrolmen in the given road segments and in their
corresponding shifts. A non linear goal programming
method is adopted to carry out the optimization. The
formulation of the above four cases is given below:
Case 1: All the priorities (P1, P2 & P3) are considered
altogether simultaneously and the achievement
function is given as:
Minimise  


n
i
iddV
1
)(
subject to (i to vi)
Case 2: Only P1 is considered in this case. Hence the
function will be:
Minimise  


n
mi
iddV )(
function will be:
Minimise  

2
)(




i
m
ki
iddV
function will be:
Minimise 
 2)( ddV
The above Non- linear Goal Programming
(NLGP) models (case 1, case 2, case 3 and case 4)
are solved by using Lingo 9.0 Solver.
V. Case Study
The considered region for the case study which
is the East Section of Visakhapatnam which is
categorized in two sub zones. They are namely, East-
I zone and East-II zone respectively. The East-I zone
is under the jurisdiction of the III Town Police
Station and East-II zone is under the jurisdiction of II
Town Police station respectively. The roads and
Junctions considered are those with high traffic
density and all are intra road segments . These details
can be viewed from the map enclosed in Fig 5.1, Fig
5.2 and Fig 5.3.
5.1 Problem Formulation
Firstly, the number of patrolmen to be allocated
is considered to be the decision variable which is
denoted by ijX , such that it gives the number of
patrolmen assigned to the road segment i within the
corresponding shift j.
Fig 5.1 Visakhapatnam East Section Limits

Fig 5.2 Considered Road Segments
Fig 5.3 Considered Junctions
Secondly, the Goals have been formulated as per
the discussion in the section of model formulation.
These goals have been considered as constraints to
which the objective function will be subject to.
5.1.1 Total Patrol Men Goal
This goal has been formulated as per the data
given in table 5.1:
Table 5.1 Police Statistics of II Town and III
Town Police Stations as on 18-07-2015
Based on the above Statistic obtained the total
number of patrolmen (PC) required is 90. Hence the
goal is formulated as
  90
14
1
11
3
1


i j
ij ddX

5.1.2 Total Budget Goal
Based on the data collected from the authorities,
the traffic department expends ₹1100 per day to
patrol the considered section of Visakhapatnam. The
total length of all the considered road segments is
29.16 km and the variable cost per km is ₹5.00,
based on which the corresponding cost per km table
has been formulated as below as shown in table 5.2:
Table 5.2 Variable cost for each Road Segment
The Hence the corresponding goal equation is
given as:
 



14
1
22
3
1
1100
i j
iji ddXC
5.1.3 Minimum Shift Requirement Goal
All the segments have different traffic density
and accident frequency as mentioned before. Hence,
In order to provide emergency services, a minimum
number of patrolmen is assigned as shown in the
table above. It is required for a) Road segments
which are linked together or Junctions and b) Each
road segment. Based on this corresponding data has
been collected and formulated as shown in Table 5.3.
Table 5.3 Minimum Shift Requirement
Based on the above formulated table the
following goal equations have been formulated as
discussed in earlier section, for both Junctions and
individual road segments:
Junctions
1)
3
1 2 12 3 3
1
( ) 19j j j
j
X X X d d 

    
2)
3
2 3 13 4 4
1
( ) 21j j j
j
X X X d d 

    
3)
3
1 3 4 5 5
1
( ) 13j j j
j
X X X d d 

    
4)
3
4 5 7 6 6
1
( ) 14j j j
j
X X X d d 

    
5)  
3
9 5 7 7
1
14j j
j
X X d d 

   
6)
3
5 14 8 8
1
( ) 12j j
j
X X d d 

   
7)  
3
12 11 9 9
1
14j j
j
X X d d 

   
8)
3
8 9 10 10
1
( ) 12j j
j
X X d d 

   
9)
3
6 7 8 11 11
1
( ) 13j j j
j
X X X d d 

    
10)
3
6 9 10 12 12
1
( ) 20j j j
j
X X X d d 

    
Individual Road Segments
1)
3
1 13 13
1
( ) 4j
j
X d d 

  
2)
3
2 14 14
1
( ) 4j
j
X d d 

  
3)
3
3 15 15
1
( ) 4j
j
X d d 

  
4)
3
4 16 16
1
( ) 4j
j
X d d 

  
5)
3
5 17 17
1
( ) 7j
j
X d d 

  
6)
3
6 18 18
1
( ) 7j
j
X d d 

  
7)
3
7 19 19
1
( ) 3j
j
X d d 

  
8)
3
8 20 20
1
( ) 5j
j
X d d 

  
9)
3
9 21 21
1
( ) 7j
j
X d d 

  
10)
3
10 22 22
1
( ) 7j
j
X d d 

  
11)
3
11 23 23
1
( ) 5j
j
X d d 

  
12)
3
12 24 24
1
( ) 9j
j
X d d 

  
13)
3
13 25 25
1
( ) 11j
j
X d d 

  
14)
3
14 26 26
1
( ) 6j
j
X d d 

  

5.1.4 Estimated Total Accident Reduction Goal
The formulation for this goal can be referred
with (iv), as discussed in the model formulation
section. The corresponding data is taken from the
accident statistics of 2013-2014 which is formulated
as shown in table 5.4
Table 5.4 Accident Statistics for the years 2013-
2014
Hence proceeding further using the above data,
and evaluating the parametric values of ija and ijb ,
we obtain the their corresponding values as shown in
table 5.5
Table 5.5 Parameter values of ija and ijb
By substituting these values in the given
model equation (v), we obtain our goal equation for
this goal equation as:
11 12 13 21 22 23
31 32 33 41 43
51 52 53 61 62 6
9.82 2.47 9.76 3.31 2.22 2.22
176.23 0.6 0.9
1.67 1.65 6.52 5.74 7.7
0.9 0.26
1.18 0.36 1.76 3.16 1.58 1.54
4.5 3.8
1
29.16
X X X X X X
X X X X X
X X X X X X
   
        
   
    
       
   
  
      
  3
72 81 82 83
91 92 93 101 103
111 112 113 121 122 123
6.66 19.25 3.44 1.13
0.3 1.3
2.23 3.44 1.66 0.44 0.46
3.5 4.5
1.86 7.35 2.5 10.53 18.2 18.2
1.6 2.2
3.24
3.8
X X X X
X X X X X
X X X X X X
 
 
 
   
      
   
   
       
   
   
        
   


131 132 133 141 142
27 27
5.2 1.53 8 2
6.45
X X X X X
d d 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
           
    
  
5.1.5 Accident Rate Reduction in High Accident
Prone Segments Goal
Similarly, as per the discussed model
formulation (vi), for this goal and using the data
given above, the corresponding parameters are
substituted and the goal equations are obtained as
follows:
1)
28 28
11 12 13
9.82 2.47 9.76
22.58 18d d
X X X
 
     
2)
29 29
21 22 23
3.31 2.22 2.22
18.89 16.992d d
X X X
 
     
3) 30 30
31 32 33
1.67 1.65 6.52
21 16.992d d
X X X
 
     
4) 31 31
41 43
5.74 7.7
28.86 13.842d d
X X
 
    
5) 32 32
51 52 53
1.18 0.36 1.76
7.48 7.6d d
X X X
 
     
6) 33 33
61 62 63
3.16 1.58 1.54
4.97 5.92d d
X X X
 
     
7) 34 34
72
6.66
16.65 9d d
X
 
   
8) 35 35
81 82 83
19.25 3.44 1.13
15.51 4.84d d
X X X
 
     
9) 36 36
91 92 93
2.23 3.44 1.66
8.12 5.14d d
X X X
 
     
10) 37 37
101 103
0.44 0.46
5.3 5d d
X X
 
    
11) 38 38
111 112 113
1.86 7.35 2.5
9.435 11.25d d
X X X
 
     
12) 39 39
121 122 123
10.53 18.2 1.82
8.78 15.54d d
X X X
 
     

13) 40 40
131 132 133
3.24 5.27 1.53
4.66 3.546d d
X X X
 
     
14) 41 41
141 142
8 2
4 8.1d d
X X
 
    
A non linear goal programming method is
adopted to carry out the optimization. The
formulation of the above four cases is given below:
Case 1: All the priorities (P1, P2 & P3) are considered
altogether simultaneously and the achievement
function is given as:
Minimise  
41
1
( ) i
i
V d d

 
subject to all the constraints in 5.1.1 to 5.1.5.
function will be:
Minimize  
41
27
( ) i
i
V d d

 
subject to all the constraints in 5.1.1 to 5.1.5
function will be:
Minimise
 
26
1
2
( ) i
i
i
V d d


 

function will be:
Minimise 
 2)( ddV
The above Non- linear Goal Programming
(NLGP) models (case 1, case 2, case 3and case 4) are
solved by using Lingo 9.0 Solver.
VI. Result and Discussions
Based on the formulated model for the case
study and previous discussions, the NLGP models
were executed in Lingo 9.0 Solver to obtain the
corresponding results for the allocation of patrolmen
for the considered four cases individually. This
model uses an integer non linear goal programming
approach to solve the need for reducing traffic
violations and accidents by optimum allocation of
patrolmen in various road segments in different
shifts.
This type or model can be utilized to the help
management of traffic police department in
Visakhapatnam city, in order to reduce manual effort
and to achieve optimal allocation of existing
patrolmen depending upon the given goal constraints
and priorities. The obtained results have tabulated
accordingly to formulate individual tables which
have been discussed in the following section. The
formulated results have been categorized into two
types of tables:
 Optimal Allocation
 Goal Fulfillments
6.1 Optimal Allocation
This analysis gives the number of patrolmen
allocated under consideration of each case. The value
obtained are integers obtained from solving the
NLGP models as discussed in the previous sections
of the Priority Structure of the Case Study Problem
formulation. These results are formulated as shown in
the table 6.1
Table 6.1 Optimal Allocation of Patrolmen for all
the considered four cases
The results in the table 6.1 are the possible
allocations of' patrolmen to the road segments with
respect to the three shifts obtained after solving the
NLGP model for all the considered four cases.
Similarly, results are also obtained pertaining to the
goals considered. These results have been formulated
as Goal Fulfillment tables. These tables give analysis
of the goal fulfillments for each of the four cases
considered individually.

6.2 Goal Fulfillments
The Goal Fulfillment analysis tables are those
tables which give the analysis of how far the
considered goals for the given model have been
fulfilled in percentiles. These tables have been further
described as below:
 Firstly, There are four goal fulfillment analysis
tables for the four considered cases each
followed by detailed discussion of the results
that are described in the individual tables.
 Each individual table consists of the goals
considered, their aspiration and solution values,
under and over deviations, and goal fulfillment
percentages respectively.
 The following goals have been described in the
tables:
 Total Patrolmen Goal
 Total Budget Goal
 Minimum Shift Requirement Goal which
consists of allocation for both Road Segments
and Junctions.
 Estimated Total accident Reduction Goal
(given as "Est. Tot. Acc. Red." in the goal
fulfillment tables)
 High Accident Prone Road Segments
Deviations consists of under and over deviations
which are denoted by nd 
and nd 
respectively.
These are obtained from the difference between the
aspiration and the solution values. Correspondingly,
the goal fulfillment percentages are calculated for
each goal individually using the obtained under or
over deviations respectively.
6.2.1 Goal Fulfillment Analysis for Case 1
Table 6.2 Goal Fulfillment table for case 1
The aspiration , solution values, deviations and
corresponding goal fulfillment percentages of the
goals considered under case 1 can be seen from the
above formulated table 6.2. In case 1 the priorities P1,
P2, and P3 are considered altogether simultaneously.
It can be seen that all the Minimum Shift
Requirement goals are achieved, but there are some
overachievements observed in case of Accident
reduction goals for high accident prone individual
road segments.

above formulated table 6.3.In case 2, only P1 is
considered individually. Unlike to that in case 1, the
case 2 analysis shows that the although there are
certain over deviations in Accident Reduction goal,
the goal achievements are better since we considered
the accident reduction priority. But in case of the
Minimum Shift Requirement goal, few goals have
been overachieved for certain road segments and
junctions.
above formulated table 6.4. In case 3, only P2 is
considered individually. Here it can be observed that
under Minimum Shift Requirement goal priority
consideration, all the corresponding goals of the
considered priority are achieved. But in case of
Accident reduction goal, the analysis shows some
certain over deviations for few road segments.

above formulated table 6.5. In case 4, only P3 is
considered individually. Here it can be observed that
under the consideration Total Budget Goal Priority,
Both the Minimum Shift Requirement and the
Accident Reduction goals show that certain goals
have been achieved while other show considerable
over deviations.
It can been throughout the goal fulfillment
analysis for all the considered cases 1,2, 3 and 4 the
Total Patrolmen goal, Total Budget goal and the
Estimated Total Accident Reduction Goal have been
achieved in each considered case.
VII. Conclusions
The problem of assigning a limited number of
patrolmen to an area in order to achieve maximum
effectiveness has received very little attention in the
law enforcement and management science literature.
Previous practices of accident prediction, control
chart analysis, and contact prediction for assigning
patrolmen or police to various road segments or
streets have been based on subjective judgments of
accident and crime frequency and traffic density
statistics. Since these methods are highly dependent
on trial and error, they have experienced limited
success.
With due time, the advent of computerized
techniques came to the rescue by providing improved
statistical and analytical models for solving tedious
problems. Of them all, mathematical goal
programming has the potential to provide better
solutions and insight to many such problems. Of
which, our problem of allocation of patrolmen in the
given traffic domain has been solved accordingly
using an NLGP model.
In an urban city like the East Section of
Visakhapatnam with incremental growth in the
population and changes in lifestyles, demand for
more number of vehicles is increasing drastically.
Heavy traffic on the roads has led to increased traffic
rules violations and accidents, causing travel
inconvenience and chaos within the city. It is the
responsibility of the traffic police department to
allocate the patrolmen in different shifts in various
road segments of the city for reducing traffic rules
violations, disorders and accidents.
As observed from the tables and analysis, it can
be said that there are certain areas where goals like
the minimum shift requirement and high accident
prone road segments have been overachieved. These
overachievements explain that there is high traffic
density existing in those Junctions (or road segments)
hence requires more allocation of patrolmen in such
areas.
The methodology demonstrated how the model
can be employed for an effective allocation of
patrolmen. The analysis of the case study under
various priorities has revealed the fulfillment of
different goals. Although it is not possible to obtain a
guaranteed optimal solution, this model demonstrates
that a satisfactory solution can be achieved. This
model has been demonstrated as a sample to show
that it can also be extended to a state wide
implementation of optimized patrol manpower
allocation.
In this study the goal programming approach is
utilized because it allows the optimization of multiple
objectives while permitting an explicit consideration
of the existing decision environment. The goal
programming model points out that some objectives
cannot be achieved under the desired conditions.
Therefore, trade-offs must occur due to
administrative policies or limited resources and with
regard to further modifications in traffic system,
policies and increase in resources, demand and
population it is felt that there is a lot of scope for
future work in this area.
References
[1] Aleksander Aristovnik, Janko Seljak
and Jernej Mencinger (March 2014),

"Performance of measurement of police
forces at the local level: A non-parametric
mathematical programming approach",
Expert Systems with Applications, Volume
41, Issue 4, Part 2, Pages 1647–1653.
[2] Algadhi .S.A.H and M.K. Hassan (1994), "A
Traffic Patrol Vehicles Allocation Model
For Riyadh", Scientific Journal of
Economics and Commerce, Volume 1.
[3] B.P. Hughes, S. Newstead, A. Anund, C.C.
Shu, T. Falkmer (2015), " A review of
models relevant to road safety", Accident
Analysis & Prevention, Volume 74, Pages
250-270.
[4] Birge J.R and S.M. Pollock (1989),
"Modeling rural police patrol", The Journal
of Operational Research Society, Volume
40, Issue no.1, Pages 41-54.
[5] Brooke Allen Saladin (1980), "Methodology
for allocation of police patrol vehicles",
Journal of Operations Management,
Volume2, Issue no.4, Pages 239-249.
[6] Gary W. Cordner and Kathryn E.
Scarborough (2010), Introduction to Police
Administration Police Administration,
Volume 7, Issue no. 1, Pages 3- 22.
[7] George Campbell (2015), "Measuring
Security Program Performance", Measuring
and Communicating Security's Value,
Chapter 3, Pages 117-196.
[8] Goghrod Hamid and Jean-Marc Martel
(2003) "Vehicle Park Management through
Goal Programming Model", European
Journal of Operation Research, Volume 41,
Issue 1, Pages 225-231.
[9] Kwak N.K and Michael B.L (2004), "Police
patrol beat design — allocation of effort and
evaluation of expected performance", A
Journal of Decision Sciences, Volume 15,
Issue no.3, Pages 421-433.
[10] Li Li, Zhongbo Jiang, Ning Duan, Weishan
Dong, Ke Hu, Wei Sun (2012), " An
Approach to Optimize Police Patrol
Activities Based on the Spatial Pattern of
Crime Hotspots",Service Science,
Management and Engineering, Volume 8,
Pages 141-163.
[11] Nathan Eagle, Kate Greene (2014), "
Engineering and Policy: Optimizing
Resource Allocation", Reality Mining,
Volume 3, Issue no. 6, Pages 99-108.
[12] Parisa Shahnazari-Shahrezaei, Reza
Tavakkoli-Moghaddam, Hamed
Kazemipoor (2013),"Solving multi-
objective multi-skilled manpower
scheduling model by a fuzzy goal
programming approach", Applied
Mathematical Modeling, Volume 37, Issue
7, Pages 5424-5443.
[13] Philip E. Taylor and Stephen J .Huxley
(1989), "A break from tradition for the San
Francisco police patrol officer scheduling
using an Optimization based decision
support system", Interfaces, Volume 19,
[14] Raja Jayaraman, Davide La Torre, Tufail
Malik, Yanthe E. Pearson (2015), "Optimal
Work Force Allocation for Energy,
Economic and Environmental Sustainability
in the United Arab Emirates: A Goal
Programming Approach", Energy Procedia,
Volume 75, Pages 2999-3006.
[15] Raktim Pal, Indranil Bose (2009), "An
optimization based approach for deployment
of roadway incident response vehicles with
reliability constraints", European Journal of
Operational Research, Volume 198, Issue 2,
Pages 452-463.
[16] Sang M. Lee, Lori Sharp Franz and A.
James Wynne (1979), "Optimizing Stateb
Patrol Manpower Allocation", The Journal
of the Operational Research Society, Vol.
30, Issue no.10, Pages 885-896.
[17] Sharma, Dinesh K., Ghosh D. and Gaur A.
(2007), ―Lexicographic Goal Programming
Model for Police Patrol Cars Deployment in
Metropolitan Cities,‖ International
Journal of Information and Management
Sciences, Vol. 18, Issue no. 2, Pages 173-
188.
[18] Sujeet Kumar Singh, Shiv Prasad Yadav
(2015), "Modeling and optimization of multi
objective non-linear programming problem
in intuitionistic fuzzy environment", Applied
Mathematical Modeling, Volume 39, Issue
16, Pages 4617–4629.
[19] Tamiz. M, S.K. Mirrazavi and D.F. Jones
(1999), "Extensions of pareto efficiency
analysis to Integer Goal Programming",
Omega, Volume 27, Issue no.2, Pages 179-
188.
[20] Taylor. B. W, Moore Li, Clayton E.R, Davis
K.R and Rakes T.R (1985), "An Integer
Non- Linear Goal Programming model for
the deployment of state highway patrol
units", International Journal of Information
and Management Sciences, Volume 31,
[21] Yingyan Lou, Yafeng Yin, Siriphong
Lawphongpanich (2011)," Freeway service
patrol deployment planning for incident
management and congestion mitigation"
Transportation Research, Volume 19, Issue
2, Pages 283-295.

Optimization of Patrol Manpower Allocation Using Goal Programming Approach -A Case Study

More Related Content

What's hot (18)

Viewers also liked (19)

Similar to Optimization of Patrol Manpower Allocation Using Goal Programming Approach -A Case Study (20)

Recently uploaded (20)

Optimization of Patrol Manpower Allocation Using Goal Programming Approach -A Case Study