2013 International Conference on Power, Energy and Control (ICPEC)

A Literature Review on Optimal Placement of Phasor

Measurement Units
K.K.More H.T. Jadhav
Dept.ofElectrical Engg. Dept. of ElectricalEngg.
Rajarambapu Institute of Technology Rajarambapu Institute of Technology
Islampur, India Islampur, India
Kalpana_778@rediffmail.com htj@ritindia.edu

Abstract—The Phasor Measurement Unit (PMU) is very A proper methodology is required to find the optimum
important tool for monitoring and control of the power system. number of the PMUs which will fully observe the power
PMUs give real time, synchronized measurements of voltages at network. To solve the Optimal PMUs Placement (OPP)
the buses and also current phase values which are incident to problem a number of methods have been employed. These
those buses where these PMUs are located. It is unnecessary and
methods usually classified into conventional methods and
impossible to place PMU at each bus to estimate the states
because the PMUs and communication facilities are very costly. advanced heuristic and modern metaheuristic methods. Linear
It is necessary to determine the minimum number of PMUs for Programming, Nonlinear Programming, Dynamic
entire observability of the power network. The optimal placement Programming are the common optimization methodologies are
of PMUs (OPP) problem solved by various techniques such as proposed to solve this problem. Problems such as difficulties of
mathematical programming, metaheuristic techniques. A obtaining local minima and handling constraints in
literature review on these technologies to solve OPP problem is conventional techniques are overcome by advanced heuristic
proposed in this paper. and modern metaheuristic optimization methodology. These
methodologies are Depth First Search, Minimum Spanning
Keywords-Phasor measurement unit; Conventional
optimization techniques; Metaheuristic optimization techniques.
Tree, Simulated Annealing, Tabu Search, Genetic Algorithms,
Differential Evaluation, Immune Algorithms, Partical Swarm
I. INTRODUCTION Optimization or Ant Colony Optimization [4].
At present due to increased power demand, fast growth of This paper reviews the research work and studies that have
generation, transmission and development in power systems been done in the area of optimal placement of phasor
congested the existing networks and therefore stability margin measurement units (PMUs). Mainly the conventional and
of these networks are decreased. In this situation to make sure recent advanced heuristic and metaheuristic optimization
proper and stable operation of the power system, an accurate techniques are presented in this paper to solve the typical
measurement and system states monitoring is required. This optimal placement of PMUs problem. The formulation of this
was normally done by Supervisory Control and Data problem is described in Section II.The new methods to solve
Acquisition (SCADA) system, where system states estimation the OPP problem are discussed in Sections III and IV. Section
depends on unsynchronized measurement. These V concludes this paper.
measurements have errors such as measurement and telemetry
bias. To overcome these limitations in the SCADA, Wide Area II. OPTIMAL PMU PLACEMENT (OPP) PROBLEM
Monitoring Protection and Control (WAMPAC) system is FORMULATION
used. This system consist Phasor Measurement Units (PMUs) PMU is intelligent to measure the phase values of voltage
as fundamental components which gives synchronized and where it is installed and also measure phase values of branch
real-time voltages and currents phasor measurement. Global currents which are connected to these buses. Figure 1 shows
Positioning System Satellite (GPS) provides reference timing PMUs purely isolated to form a Wide Area Monitoring System
signals to achieve synchronization of sampling voltage and (WAMS).GPS time stamped measurement signals are fed to a
current waveform with respect to this reference time. A PMU Phasor Data Concentrator (PDC) by using PMUs. The PDC
directly measure the voltage Phase of the bus where these collects and sorts the phasor measurements and signal
PMUs are placed and also measure the current phases of a few processor converts data of PMUs into useful information which
or all the branches connected to that bus. In recent years rapidly is visible on Human Machine Interface (HMI).The operator can
increases use of PMUs to improve monitoring, needed to place easily access the critical information of the power system state.
these PMUs on all of the buses for full observability of the Some rules can be used for the placement of PMUs which are
network. It is also impossible to place these units on entire given in [5] like, assigning one voltage measurement at the bus
system buses because PMUs and communication services are where PMU is located, one branch current measurement, one
very costly. Thus determination of the optimal number of voltage and current pseudo measurement.
PMUs and its location for overall observability of the system is
very important.

978-1-4673-6030-2/13/$31.00 ©2013 IEEE

220
 2013 International Conference on Power, Energy and Control (ICPEC)
injections (i.e. zero injections) and (3) PMU measurements,
injections and flows.

III. MATHEMATICAL PROGRAMMING METHODS

A. Integer Linear Programming(ILP)

Integer Programming (IP) is a numerical programming
method [4]. It solves an optimization problem which has
integer design variables. According to the types of objective
function and the constraints, whether they are linear, nonlinear
or quadratic, an integer programming is divided into Integer
Linear Programming (ILP), Integer Nonlinear Programming
(INLP) and Integer Quadratic Programming (IQP)
respectively.
Fig 1.Layout of PMU along with GPS time stamped signals
Reference [7] presents the method with the objective of
minimization of number of PMUs located at the buses that
The PMUs can be placed at planned buses to completely remove critical measurement in the system. Then typical
observe the total network. These located PMUs are measuring conventional measurement integrated by extending the
the voltage phase value of that bus and current phase values of placement problem. In addition, the same formulation may be
the lines which are connected to the same bus. The aim is to applied to find out optimum locations if a preferred level of
completely observe the network with an optimum number of general redundancy is considered. Bad data detection is
PMUs. The problem for n-bus system is formulated and solved improved with improving the redundancy level. For this
by Integer Programming method [6] as given below: purpose placement algorithm which allows measurement
system designs with various degrees of risks under the loss of
min∑ measurements and bad data.

subject to … 1 Reference [8] proposed a new algorithm to determine

Where = binary decision variable vector, optimum number of PMUs for fully observability of the
i = bus number, whose entries are, system under any contingency. The suggested algorithm
=1 if a PMU is located at bus i, operates in four stages; the essential measurement for totally
=0 otherwise. known condition is found out in first stage. In next stage, it
= the cost of the PMU installed at the bus, selects redundant measurement set under any contingencies.
=vector function, having non-zero entry if corresponding Above both stages uses least condition number of normalized
bus voltage is observable and measurement matrix to select the proper measurement
location. Further, minimum redundant measurement set is
1 = vector having all ones entries.
chosen from these measurements using the binary integer
The nonlinear constraint expressions are created
programmining technique. Finally these measurements i.e.
considering the placement and types of available
essential and minimum redundant measurements are arranged
measurements. Assume the phasor value of voltage at the bus
such that locations of PMUs are minimum. For this purpose
where PMU located and values of current phasors along the
heuristic method is used. The IEEE test systems results show
branches which connected to that bus will be easily accessible.
proposed placement method which ensures the system
The other adjacent bus voltages will also be accessible.
observability under the given contingency. Also it provides the
Determine the solution vector which is a set of minimum
reliable measurement to the system.
and satisfy above equation. The constraint function can be
defined with the help of Binary Connectivity Matrix A which
Bei Gou proposed a generalized formulation in [9], for
gives the information about bus connectivity of power
optimal PMU placement which is necessary to complete
network. The elements of matrix A is defined as,
system observability under with and without zero injections
consideration. He proposed PMU placement model of integer
, = 1 if linear programming (ILP) in [10] which is formulated in [9].
= 1 if there is connection between bus m and bus n.
= 0 otherwise. Using an Integer Linear Programming (ILP) structure, a
process for placement of multistage PMU has been proposed
The constraint equations are considered for the three cases: (1) in [11].The zero injection constraints may be modeled as
PMU measurements only, (2) PMU measurements and linear constraints. The optimum placement difficulties have
the different solutions, therefore to choose the optimal

221
 2013 International Conference on Power, Energy and Control (ICPEC)
solutions two indices have been proposed; namely Bus A Binary Integer Programming technique for
Observability Index (BOI) which gives the total number of instantaneous PMUs placement and conventional power flow
PMUs observing a given bus and to get the sum of all BOI for measurements for fault observability in power systems
the system, System Observability Redundancy Index (SORI) proposed in [16]. The formulation is initially based on a non-
is used. linear programming problem is unsolvable therefore initial
formulation results in a binary non-linear integer programming
A unified method is suggested in [12] to find out the problem that is transformed into a corresponding ILP problem
optimum number of PMUs and place the PMUs such that through boolean implications. The transformation provides a
system can be totally observable. The proposed method take high scale of scalability.
care decision strategy for optimum PMU allocation problem
under the conventional measurements with single or multiple Reference [17] represents an optimization model to
PMU loss contingency. Here placement problem is formulated determine optimum number of PMUs in power network. The
as a Binary Integer Linear Programming (BILP) problem for proposed model consists of observability necessities based on
this purpose it consist the branch–bus model of the system. a set of probabilistic criteria. The nonlinear mathematical
expression related with the probability of observability index
Reference [13] proposed Integer Linear Programming with is changed to a linear model by a useful linearization
a fast and novel practical model to determine the optimum technique. This linear model further described as a Mixed
number of PMU. It consist power system contingency Integer Programming (MIP) problem. This proposed
conditions i.e. line outages and measurement losses. The mathematical model is therefore well-suited with the Mixed
proposed model is flexible than existing model because other Integer Programming solution tools and also it tested on large-
situation can be included separately or all together in power scale power systems.
systems. Also, model consist measurement limitations i.e.
communication constraints of power networks. The main Reference [18] consist the combined placement of phasor
importance of this method; it take low time getting the result and power flow measurements for total power system
also it gives local optimal solution for large-scale power observability. The proposed method initially formulated
system. problem as a Non-linear Integer Problem and after that
transformed into an ILP problem by considering
In [14] a new, simple and easy to implement optimum supplementary variables and constraints. Further problem
placement method is presented for observability of the power formulation extended to large-scale systems, for this purpose
network. This suggested method consist a two-stage which is two heuristic methods are proposed. This proposed method is
based on topological observability with numerical solved on standard IEEE and large bus systems. The result
observability analysis for placing PMUs. The first stage gives more efficient solution.
consist an Integer Linear Programming method to make
certain system observable topologically. To check the system Reference [19] presents a Binary Integer Linear
numerically observable a sequential elimination algorithm Programming method for the power system observability with
(SEA) is used in second stage. The results are tested on optimum number of PMU also for maximizing the
standard IEEE bus systems. measurement redundancy under normal condition and also
under a line outage or PMU failure. The optimization problem
A technique is proposed in [15] for placing the PMUs in formulated as Binary Integer Programming problem with
multiple stages over a given time period that ensures complete consisting sets of linear inequalities constraints. The proposed
power system observability even under a branch outage or a method was also consist effect of zero-injection buses in the
PMU failure. The proposed technique is based on integer power system. This method tested on IEEE standard and also
linear programming (ILP) and a multi-criteria decision-making large-scale power systems.
(MCDM) approach. Integer Linear Programming method is
used in first stage to find the optimum number of PMUs for B. Integer Quadratic Programming(IQP)
the complete observability of the system. These PMU Quadratic Programming (QP) is solving the optimization
placements include the contingency cases such as single problem which has quadratic objective function and linear
branch outage and single PMU outage. A multi-criteria constraint. In this technique all design variables take only an
decision-making (MCDM) subsequently developed to integer value [4].
prioritizing the optimal PMU locations. This criteria include The optimal PMU placement method proposed in [20]
three indices, viz. tie-line oscillations observability index makes the system observable topologically. An Integer
(TOI), voltage control area observability index (VOI) and bus Quadratic Programming technique used to optimize the number
voltage observability index (BOI). Finally, select the PMU of PMUs which maintain system fully observable under normal
locations in each stage by ranking the optimal locations using condition and also under any outage of a line or PMU. This
the proposed indices in the MCDM process. method also maximizes the measurement redundancy of the
system. It has been solved on IEEE standard system under
contingency condition. In [21] another integer quadratic

222
 2013 International Conference on Power, Energy and Control (ICPEC)
programming approaches, determining the solution of optimal C. Immune Algorithm (IA)
placement of PMUs problem considering connectivity matrix The Immune Algorithm (IA) is a search strategy which is
which represents the system topology and also formulate the based on genetic algorithm principles and enthused by
minimization problem. protection mechanism [4].
In [25] the application of Immunity Genetic Algorithm
IV. ADVANCED HEURASTIC AND MODERN METAHEURISTIC (IGA) proposed for the optimal placement of Phasor
Measurement Units (PMUs) problem in power network. The
METHODS
topological observability analysis rules are used to abstract the
A. Simulated Annealing(SA) three vaccines. A new rule has been employed for evaluation of
Simulated Annealing (SA) is a genetic probabilistic observability. The anticipation from familial reproduction has
metaheuristic method for the global optimization problem of been studied and also its effects improved the algorithm
placing a good approximation to the global optimum of a given converging speed in case of few individual in each population.
function in a large search space. It is mostly used for discrete The result of the proposed algorithm is compared with other
search space. The target is to find a suitably good solution methods on IEEE standard test system.
rather than the best possible solution in a fixed amount of time. D. Tabu Search (TS)
The proposed SA method in [22] presents technique or Tabu Search (TS) is metaheuristic local search algorithm
optimum placement of PMU depends on incomplete that can be used for solving combinational optimization
observability of the system. Firstly to determine the number of problems by tracking and guiding the search [4].
PMUs such that system completely observable, place the PMU
considering spanning trees of power system graph. Then new In [26] a fast analysis technique proposed for the power
concept of “depth-of-unobservability” is suggested and shows network topological observable. This technique consists
its effects on the quantity of PMUs placements. This method augments incidence matrix which manipulates only integer
also gives the suitable results to PMU placement problem with numbers and Tabu Search (TS) algorithm. The solution of
communication constraint. combinatorial OPP problem by this method has less
computation and higher strength. Hence this method gives fast
B. Genetic Alorithm(GA) and accurate results as compared to conventional observability
Genetic algorithm (GA) is adaptive heuristic search analysis.
algorithm that repeats the process of natural evolution. This E. Differential Evolution (DE)
process is used to generate solutions to optimization and also
search problems [4]. Differential Evolution (DE) is a metaheuristic process
which improves a candidate solution by optimizing a problem
The non-dominated sorting genetic algorithm suggested in iteratively.
[23] solves the optimal placement of PMUs problem with
biobjectives, such as minimization of the number of PMUs and In [27] a multi-objective PMU placement model is
maximization of measurement redundancy. This optimization presented for entire power network observable. For this
method finds the Pareto optimal solutions. The proposed purpose NSDE algorithm is used which is integrated from
algorithm first follow graph theory and simple GA procedure Differential Evolution and Pareto Non-dominated Sorting
to obtain individual solution and then finds most desirable algorithm. The proposed model easily and quickly finds
solution which may choose the decision-maker using Non- number of Pareto-optimal solutions and gives global and
dominated Sorting Genetic Algorithm (NSGA). The main accurate solution with full observability.
benefit of this NSGA is that it gives the best Pareto-optimal
solution instead of one solution, and could provide large search F. Exhaustive Search (ES)
space for multi-objective optimization problems. Exhaustive Search (ES) is an insignificant but very general
problem solving technique. It has all possible candidates for the
Reference [24] a Multi-Objective Biogeography Based solution and verifies whether each candidate satisfies the
Optimization (MO-BBO) algorithm is proposed to find problems statement or not. The search is said to be exhaustive
optimum PMU placement which ensures the power system because the search has surety to generate all accessible states
completely observable. It consist Non-dominated Sorting before it terminates with failure.
Genetic Algorithm (NSGA) and ranking procedure to get the
Pareto optimal solution and also choose best solution using a To observe the system network in minimum number of
fuzzy method. This proposed method gives the well suited PMUs proposed an exhaustive search method in [28]. In some
Pareto optimal solutions than the previous method NSGA-II cases more than one solution sets are presented so to select the
and NSDE. This MO-BBO gives flexibility and it can be preferable one solution this exhaustive binary search method is
applied to more objectives. used due to its exhaustive nature. Thus this method gives the
global minimum solution and also overcome the limitations of
the conventional optimization method. This proposed method

223
 2013 International Conference on Power, Energy and Control (ICPEC)
applied on various IEEE standard systems and also on 298-bus [13] F.Aminifar. Khodaei, M. Fotuhi-Firuzabad and M.Shahidehpour
“Contingency-Constrained PMU Placement in Power Networks” IEEE
system with conventional measurement. Trans. Power Syst., vol. 25,no.1, Feb. 2010.
G. Particle Swarm Optimization (PSO) [14] R.Sodhi, S.C. Srivastava, S.N. Singh“Optimal PMU placement method
for complete topological and numericalobservability of power
Particle swarm optimization (PSO) is a population based system”Electric Power Systems Research 80, pp.1154–1159, April
stochastic optimization technique. .2010.
[15] R.Sodhi S.C. Srivastava S.N. Singh, “Multi-criteria decision-making
approach for multistage optimal placement of phasor measurement
In [29] a binary particle swarm optimization (BPSO) units” IET Gene.Trans.Distrib.vol.5, iss.2, pp.181-190, 2011.
technique is proposed to find the optimum number of PMUs [16] R.Kavasseri, S.K.Srinivasan“Joint placement of phasor and conventional
for total observability and to maximize the measurement power flow measurements for fault observability of power systems” IET
redundancy at the buses in the power systems. A topology Gene.Trans.Distrib.vol.5, iss.10, pp.1019-1024, 2011.
[17] F.A.Khodaei, M. Fotuhi-Firuzabad and M.Shahidehpour “Probabilistic
based algorithm is used with the conventional measurement to Multistage PMU Placement in Electric Power Systems”IEEE Trans.
observe the power network. The proposed method verifies on Power Del, vol. 26, no. 2, April. 2011.
the IEEE standard bus systems. [18] R.Kavasseri and S.K.Srinivasan, “Joint Placement of Phasor and Power
Flow Measurements for Observability of Power Systems” IEEE Trans.
Power Syst, vol. 26, no.4, pp.1929-1936, Nov. 2011.
[19] A.Enshaee, R.A.Hooshmand, F.H.Fesharaki “A new method for optimal
placement of phasor measurement units to maintain full network
V. CONCLUSIONS observability under various contingencies” Electric Power Systems
Research 89 ,pp. 1– 10, March 2012
The various methods are used to obtain solutions on the [20] S.Chakrabarti, E. Kyriakides and D. G. Eliades, “Placement of
optimum location of PMUs. The correct methodology is Synchronized Measurements for Power System Observability” IEEE
required to find the optimal placement of PMUs in power Transactions on power delivery, vol. 24, no. 1, Jan. 2009.
network. The conventional and metaheuristic optimization [21] S.Chakrabarti, E. Kyriakides, and M. Albu, "Uncertainty in power
system state variables obtained through synchronized
techniques are presented in this paper. The review of measurements,"IEEE Trans. Instrumentation and Measurement, vol. 58,
optimization techniques suggested in this paper to solve the pp.2452-2458, Jan. 2009.
Optimal Placement Problem (OPP) will be useful for the [22] R.F.Nuqui and A. G. Phadke, “Phasor Measurement Unit Placement
researchers. Techniques forComplete and Incomplete Observability” IEEE Trans. on
power delivery, vol. 20, 2381-2388, October 2005.
VI. REFERENCES [23] B.Milosevicand M. Begovic“Nondominated Sorting Genetic Algorithm
for Optimal Phasor Measurement Placement” IEEE Trans. Power
[1] J. S. Thorp, A. G. Phadke, and K. J. Karimi, "Real time voltage-phasor Systems, vol. 18, pp. 69-75, February 2003.
measurements for static state estimation," IEEE Trans. PowerApparatus [24] K.Jamuna, K.S. Swarup, “Multi-objective biogeography based
and Systems, vol. 104, pp. 3098–3106, Nov. 1985. optimization for optimal PMU placement” Applied Soft computing 12,
[2] A.G.Phadke, J. S. Thorp, and K. J. Karimi, "State estimation with phasor pp.1503–1510, February 2012.
measurements," IEEE Trans. Power Systems, vol.1, pp. 233–241, Feb. [25] F.Aminifar, C.Lucas, A. Khodaei, and M. Fotuhi-Firuzabad “Optimal
1986. Placement of Phasor Measurement Units Using Immunity Genetic
[3] A.G.Phadke and J. S. Thorp Synchronized Phasor Measurements and Algorithm” IEEE Transactions on Power Delivery, vol. 24, pp.1014-
Their Applications. New York: Springer, 2008. 1020, July 2009.
[4] N.M.Manousakis, G.N.Korres and P.S.Georgilakis,“Optimal Placement [26] J.Peng, Y. Sun, and H. F. Wang, "Optimal PMU placement for full
of Phasor MeasurementUnits: A Literature Review”IEEE network observability using Tabu search algorithm,"
conference,2011 InternationalJournal of Electrical Power & Energy Systems, vol. 28, pp.
[5] T.L.Baldwin, L.Mili,M.B.Boisenand R.Adapa ,“power system 223-231, May 2006.
observability with minimal phasor measurement placement," IEEETrans. [27] C.Peng, H. Sun, and J. Guo, "Multi-objective optimal PMU placement
Power Systems, vol. 8, pp. 707-715, May 1993. using a non-dominated sorting differential evolution algorithm",
[6] B.Xu and A.Abur, “Observability analysis and measurement placement International Journal of Electrical Power & Energy Systems, vol. 32,pp.
for system with PMUs,” in Proc. IEEE Power System Conf. Expo., vol. 886-892, Oct. 2010.
2, pp.943–946, Oct. 2004. [28] S.Chakrabarti and E. Kyriakides,“Optimal Placement of Phasor
[7] J.Chen and A. Abur, "Placement of PMUs to enable bad data detection Measurement Units for Power System Observability” IEEE Trans. on
in state estimation," IEEE Trans. Power Syst, vol. 21, pp. 1608-1615, Power Systems, vol. 23,pp.1433-1440,August 2008.
Nov. 2006. [29] A.Ahmadi,Y. Alinejad-Beromi, M. Moradi,“Optimal PMU placement
[8] C.Rakpenthai,S.Premrudeepreechacharn,S.UatrongjitandN.R.Watson,“A for power system observability using binary particle swarm optimization
n Optimal PMU Placement Method Against Measurement Loss and and considering measurement redundancy” Expert Systems with
Branch Outage”IEEE Trans .Power Del,vol. 22,no.1,pp.101- Applications 38,pp.7263-7269,2011.
107,Jan.2007
[9] B.Gou, "Generalized integer linear programming formulation for
optimal PMU placement", IEEE Trans. Power Syst, vol. 23, no.3, pp.
1099-1104, Aug. 2008.
[10] B.Gou, "Optimal placement of PMUs by integer linear programming,"
IEEE Trans. Power Syst, vol. 23, no. 3, pp. 1525-1526, Sep. 2008.
[11] D.Dua, S. Dambhare, R. K. Gajbhiye, and S. A. Soman, "Optimal
multistage scheduling of PMU placement: An ILP approach,"
IEEETrans. Power Del, vol. 23,no.4, pp. 1812-1820, Oct. 2008.
[12] N.H.Abbasy and H.M Ismail “A Unified Approach for the Optimal
PMU Location for Power System State Estimation” IEEE Trans. Power
Syst., vol. 24, no.2,pp.806-813,May. 2009.

224