2023 3rd International Conference on Energy, Power and Electrical Engineering (EPEE)

Reactive Power Optimization of UPFC System Based

on Particle Swarm Optimization
Peng Wu1, Yaping Wang2*, Yuan Hu1, Penghui Zhao1, Ming Gao, Shuai Zhang 1, Yan Li1
1StateGrid Hebei Economic Research Institute, Shijiazhuang, Hebei Province,050011, China
2Hebei Jiaotong Vocational and Technical College Shijiazhuang Hebei Province, 050035, China

*Corresponding author’s e-mail: [email protected]

2023 3rd International Conference on Energy, Power and Electrical Engineering (EPEE) | 979-8-3503-1818-0/23/$31.00 ©2023 IEEE | DOI: 10.1109/EPEE59859.2023.10351799

Abstract. The introduction of new intelligent components in the power grid, this paper analyzes the objective function and
power system produces more harmonics, and UPFC can adjust constraints after adding UPFC, establishes a better optimal
them well. For improving the effectiveness of performance calculation mode for solving UPFC, and optimizes the reactive
parameters of the power system, a reactive power optimization energy system considering UPFC under the background of
method based on particle swarm optimization considering UPFC particle swarm optimization. Finally, an example is simulated
is proposed. Through the paper, the energy parameter with UPFC by IEEE9 nodes, and the effect of particle swarm optimization
is calculated under the way of the power injection model, and the under UPFC is compared and analyzed.
objective function is to minimize the energy loss in the system. On
this basis, reactive power optimization is realized by particle Ⅱ. Basic structure and calculation model of UPFC
swarm optimization. Under the background of the simulation of
the IEEE9 bus, the results of system parameters before and after A. Basic structure of UPFC
using particle swarm optimization in UPFC are analyzed. It is The system theoretical diagram of UPFC in Figure 1 is
found that the proposed method has a good advantage to improve mainly composed of a transformer and two axisymmetric power
the system network loss and optimize the system quality, which converters, which are connected in series and parallel in the line.
has good practical value.
On line ij, two inverters are connected through a filter capacitor
Keywords: UPFC, Power Optimization, Particle Swarm
for circuit topology, and transformers T1 and T2 are
Optimization respectively arranged on opposite sides. Node i voltage is
converted to voltage through a transformer named T1. Through
Ⅰ. Introduction changing the amplitude or phase of the system, the power flow
of line ij can be controlled in real time, and the functions of
UPFC can realize the comprehensive measurement of voltage stability and network loss reduction can be realized [6].
voltage, phase, and impedance parameters of power flow,
which has outstanding advantages in outstanding advantages in Ui Uj
balancing network parameters, reducing losses, and regulating T2 ZL
i j
power flow throughout the entire system [1,2]. With the
continuous innovation and improvement of new technology and
T1
equipment, UPFC technology has been rapidly developed and
applied, and its good control effect on power flow in
engineering has been confirmed. In recent years, UPFC has C1 C2 active power
been applied more and more in the regional power grid, and
Shanghai, Zhejiang, Hubei, et al. have successfully put into Figure 1. The system theoretical diagram of UPFC
operation and achieved good demonstration effects. For better
to ensure the system balance and minimum loss of energy B. Reactive power optimization calculation model
system considering UPFC, the research on reactive power considering UPFC
optimization considering UPFC has significant economic value
[3~5] When the UPFC device is configured in the power network,
.
it can be replaced with a regular voltage and current-generating
Reactive energy optimization is described as the power device, as well as controllable variables of voltage, current, and
system, under the premise of meeting the constraints of safe phase are generated in the power flow. The power principle can
operation, by formulating a given objective function, and under be represented in Figure 2. Furthermore, the energy flow
the premise of meeting the constraints, making the objective calculation mode used for the particle swarm optimization
function value conform to the optimal solution, so as to achieve algorithm can handle parameter optimization problems more
the purpose of reducing system loss and stabilizing voltage. For quickly, as shown in Figure 3.
a long time, many experts and scholars have done many studies
and explorations for this problem, from traditional numerical
analysis to intelligent algorithms, and constantly carried out
optimal solutions. Such as linear programming methods and
neural algorithms, have achieved good research results. Based
on solving the traditional reactive power optimization of the

979-8-3503-1818-0/23/$31.00 ©2023 IEEE 927

Authorized licensed use limited to: Andhra University College of Engineering. Downloaded on December 29,2023 at 06:34:08 UTC from IEEE Xplore. Restrictions apply.
 Ui∠φi Uj∠φj to make the power loss in the entire network reach the required
i T j value, and its result function is:
iz UT∠δT Zij  V −V
2
  Qi − Q(ilim) 
2
(2)
jBc/2 jBc/2 minf = P + V   i (ilim )  + Q i
iN PQ  Vi ,max − Vi ,min 
 
N g  Qi ,max − Qi ,min 

where P is actual losses in the power grid;  is the penalty

iq it
factor of voltage and imaginary power. NPQ and NG are the total
Figure 2. Algorithm energy schematic diagram number of PQ nodes and the total number of generator nodes
respectively.
US B. Restraint condition
Upq 1) Equality restraint
φi The equation constraint mainly considers that all points of
Ui Uz
the system should also keep power conservation after
φj I Uzq considering UPFC. For the power flow calculation process of
power systems containing UPFC devices, the impact on voltage
and power should be considered, and the power equation should
Uj be derived to obtain:
n
Figure 3. Energy flow calculation mode for particle swarm optimization
algorithm Pl = U l U o (cos loGlo + sin  lo Blo ) + Pl (l n)
o =1

Each power system parameter can be decomposed into n

multiple components and can achieve a balanced state in all Ql = U l U o (Glo sin lo − Blo cos  lo ) + Ql (ln ) (3)
aspects, similar to traditional power flow optimization research, o =1
as shown in Figure 2, only adding parameters for the injection
n
stage. In Figure 3, the energy flow calculation mode for the
particle swarm optimization algorithm is shown. The Pm = U m U o (cos loGlo + sin  lo Blo ) + Pm (lo )
parameters of injection energy are optimized under the need of o =1
the scene. For facilitating the optimal reactive energy model, n
such scene can be obtained through calculation. Qm = U m U o (cos loGlo + sin  lo Blo ) + Qm (lo )
Pi (in) = U j UT [bij sin( j + T ) − gij cos( j - dT )] + 2U iU（ 2
T g ij + Bc / 2）cos(ji - dT )+U T g ij
o =1

1
Qi (in) = -UiUT [ gij sin(T − i )+( Bc + bij ) cos(T − i )] − I qUi (1) where Pl and Pm are actual power loss for nodes l , m . Ql
2
and Qm are as imaginary power of nodes l , m . Based on the
Pj (each) = UtoU je [ gij cos(T −  j )-sin(T −  j ) bij ] premise that each position in the power grid system conforms
to the above equation, it is possible to study the best system
Q j (each ) = U jUT [sin( −  ) g + cos( −  ) b] parameter performance and obtain good economic benefits in
this situation.
where P(ieach ) and P( jeach ) both are the active energy injected 2) Inequality constraints
into node, Q(i each) and Q( jeach ) both are reactive power Inequality constraints are premise for limiting variables to
maintain a reasonable range. Based on the main variables in the
injected into node, and Ui , U j
UT are voltage
as well as power grid, key parameters are selected as limiting variables.
As long as the actual values of the variables are within the upper
amplitudes of transformers and t respectively. each and  each and lower limits that meet the requirements, it is considered that
are phases for transformers at nodes i , j and t. b and g the stability of the power grid meets the expectations. Based on
this, performance parameters are worth selecting to ensure the
correspond to the conductivity and admittance of the line i-j. Bc most effective implementation:
is electrical susceptance between line and ground.
VLi  VL  VLa
Ⅲ. Reactive power optimization model under UPFC
QLi  QL  QLa
A. Result function
For ensuring sufficient optimum performance parameters Vli  V  Vla (4)
under the premise of power grid stability and realizing the goal
of saving double carbon resources, state optimization of power
throughout the entire system usually focuses on the economy.
Cli  C  CLa
Therefore, this paper establishes an optimal power flow model

928

Authorized licensed use limited to: Andhra University College of Engineering. Downloaded on December 29,2023 at 06:34:08 UTC from IEEE Xplore. Restrictions apply.
where a and i are upper and lower limits of the the switching capacity of the compensation of imaginary power
corresponding parameters each. After installing the UPFC, the device and control variables set by UPFC. The iterative formula
UPFC will introduce new voltage, current and phase control is:
variables. The constraints are as follows.
vmn (h + 1) = yvmn (h) + q1r1 ( pmn (h) − xmn (h) ) + q2 r2 ( pdn (h) − xmn (h) ) .
UTmin  UT  UTmax The location update formula is:
0  (5) xmn (h + 1) = xmn (h) + vmn (h + 1) .
I qmin  I q  I qmax B. Example study
After calculation, an 11 node power network can better
Parameters UT ,  T and Iq in the formula are shown in conduct simulation research, and the topology diagram of a
regional power grid is constructed in the following way. As
Figure 2. shown in Figure 4, the system is a circular power grid, with two
Ⅳ. Reactive power optimization based on PSO algorithm connected to each other. The UPFC is located near the 8th side
of the bus and has a stable configuration. The transformer
A. PSO principle adopts 220 kV national grid universal equipment, which meets
The ultimate goal of PSO is the position vector of particles the system operation requirements and has a capacity of 240
when the objective function value is the optimal solution, that MVA. Firstly, by setting the basic parameter values, the entire
is, the control variables of each bit [7]. When UPFC is regional power grid is kept in a stable state. Then, simulation
introduced into the line, the UPFC link is added to the reactive analysis is used to continuously adjust the variables while
power optimization model besides the traditional regulating meeting the interval range, to reduce the power loss of the line
variables, that is, the vector consisting of the terminal voltage and control the error within a controllable range, achieving
in the generator, the variable ratio of the adjustable transformer, high-quality operation of the power grid.

Figure 4. 11 node power network

Based on above primary conditions, we update power introduction of particle swarm optimization and UPFC device
parameter calculation values through continuous iteration, and can effectively reduce the active power loss of power grid.
the initial active power loss of energy without considering
UPFC is 0.0512 p.u. The PSO algorithm used to optimize has a Table 1. Comparison of optimization results
population size of 30 and it can ensure good processing speed.
Program
Table 1 is the simulation result of the particle swarm PSO UPFC PSO-UPFC
optimization algorithm which is executed alone, considering Active power
the influence of UPFC and UPFC. From the point of view of 0.0501 0.0507 0.0489
loss
active power loss, compared with the initial conditions, the Minimum
0.0952 0.0980 0.0976
active power loss of the particle swarm optimization algorithm voltage
is 0. 0512 p.u., the reactive power optimization considering
UPFC is reduced by 12%, the reactive power optimization Ⅴ. Conclusion
considering UPFC is reduced by 8.5%, and Solved performance Through the study of solving traditional imginary power
parameters is reduced by 14.3%. These results show that the optimization of electric grid, this paper calculates the power

929

Authorized licensed use limited to: Andhra University College of Engineering. Downloaded on December 29,2023 at 06:34:08 UTC from IEEE Xplore. Restrictions apply.
flow with UPFC based on the power injection model, analyzes
the objective function and constraints after adding UPFC,
optimizes the power system structure based on continuously
iterating parameter variables, and on this basis, uses particle
swarm optimization to realize imginary power optimization.
Finally, by the IEEE11-node analysis, the results of grid
parameters before and after using particle swarm optimization
in UPFC are analyzed. It is found that the method of the study
can reduce the system loss and prompt the voltage quality, but
there is still a problem that the solution time of the model is
slow or unable to converge. The power flow optimization can
be further studied by introducing other intelligent algorithms
and analytic hierarchy process, and the results are worthy of
comparison with the proposed method.
Acknowledgments
This study contributes to the construction of a strong new
energy province, and we appreciate the help of the relevant
work team.
References
[1] Qi Tao, Li Yijie, Li Zicheng, Wang Shuo, Jiang Jie. Optimization of reactive
power in distribution networks [J]. Electrical Switch, 2022,60 (05): 16-
20.
[2] Qi Yanan, Liu Baoguo, Wei Daining, Yang Wenxiao. Research on the
calculation method of reactive power compensation capacity for
photovoltaic power plants [J]. Coal Engineering, 2023,55 (02): 24-28.
[3] Liu Guojing, Qi Wanchun, Huang Junhui, Gong Hui, Bai Bin, Chen Zhe.
Research Summary of Unified Power Flow Controller [J]. Power System
and its acta automatica sinica, 2018,30(12):78-86.
[4] Ma Xiaowei, Dong Shaopeng, Sun Jianying, Xue Chen. Automatic
configuration method for reactive power compensation of distributed
nodes in DC converter stations based on particle swarm optimization
algorithm [J]. Manufacturing Automation, 2022,44 (08): 177-180.
[5] Luo Hairong, Cai Jianhui, Zhang Shuang, Gao Bo, Li Yongliang.
Evaluation of voltage stability margin based on three-phase transmission
and distribution joint power flow simulation [J]. Ningxia Electric Power,
2022 (06): 1-7
[6] Wang Yi, Gu Qiaogen, Zhang Xiaoyu, Wen Jifeng, Chen Jianmin. Analysis
of turn-to-turn fault characteristics of UPFC series transformer [J]. Power
System Automation, 2018,42(24):98-104.
[7] Xia Peng, Xu Ning, Zhou Yonggang. Research on multi-objective reactive
power optimization of distribution networks considering static voltage
stability [J]. Information Technology, 2023,47 (03): 64 69. DOI:
10.13274/j.cnki.hdzj.2023.03.012

930

Authorized licensed use limited to: Andhra University College of Engineering. Downloaded on December 29,2023 at 06:34:08 UTC from IEEE Xplore. Restrictions apply.