The interval sensitivity analysis and optimization of the distribution
network parameters considering the load uncertainty
Zhigang Lu
a,
⇑
, Jia Liu
a,b
, Yane Liu
a,c
, Rongrong Ding
a
, Fang Yang
a
a
Key Lab of Power Electronics for Energy Conservation and Motor Drive, Yanshan University, Hebei Province 066004, PR China
b
Qinhuangdao Power Corporation Limited, Qinhuangdao 066004, Hebei Province, PR China
c
State Grid Jibei Changli Power Supply Co. Ltd., Changli Hebei 066600, PR China
article info
Article history:
Received 6 September 2012
Received in revised form 27 July 2014
Accepted 19 August 2014
Available online 18 September 2014
Keywords:
Loss sensitivity
Network loss
Interval analysis
Bacterial colony chemotaxis
Distribution system
abstract
This paper proposes an advanced approach to optimize the line and transformer parameters for the
distribution system. The approach first adopts a novel set of interval sensitivity models, based on which
an interval forward–backward power flow algorithm and an interval analysis method are further
developed. In order to improve computational speed, this approach uses a simplified set models to cal-
culate power system loss interval sensitivity with respect to the admittance of lines and transformers.
An interval objective function is formulated and the optimization problem is solved with a discrete
bacterial colony chemotaxis algorithm. The applicability and effectiveness of the proposed approach is
demonstrated on two real distribution systems.
Ó 2014 Elsevier Ltd. All rights reserved.
Introduction
In recent years, with the considerable increase of the load
demand, some distribution lines and transformers can no longer
meet the actual power delivery requirements, which leads to the
loss increase and distribution system operation. Hence, the identi-
fication of lines and transformers that are responsible for the sys-
tem loss increase becomes necessary in order to establish the
reconstruction strategies.
The subject of power system loss sensitivity is widely addressed
in various areas such as transaction evaluation, voltage stability,
sensitivity analysis and loss minimization [1–6]. Refs. [7–9] present
a set of models to calculate power system loss sensitivities with
respect to power system parameters. These models are derived
from the Tellegen’s theorem and the adjoint network concept.
Ref. [10] adopts the least-square fitting technique to fit the rela-
tionship curves between the line cross-sectional area and invest-
ment, proposes a sensitivity analysis method to optimize line
parameters based on Particle Swarm Algorithm, and formulates
the objective function to minimize the comprehensive investment
cost and system power loss. Ref. [11] uses the economic current
density and load rate to reconstruct lines and transformers and
analyze their economic operation range. Based on the definition
of the conjugate branch current sensitivity with respect to the line
or transformer parameters, the power system loss sensitivity with
respect to the line or transformer parameters is calculated. Accord-
ing to the loss sensitivities, it is possible to determine every com-
ponent’s influence degree to the power system loss, and provide an
easy way for estimating the network loss when the parameters are
changed. Besides, Ref. [12] selects optimization interval using the
lines and transformers economic operation range. In addition, a
particle location library is establish to update particle location.
Ref. [13] improves the adequacy of constructed sensitivity models
and allows to consider a power system reaction on the basis of con-
structing so-called functionally oriented equivalents.
Taking load uncertainties [14] into consideration, the loss sensi-
tivity with respect to certain parameters varies along with time
changes. Therefore, this paper focuses on computing the loss inter-
val sensitivities over a period of time. The meaning of the loss
interval sensitivity is that the value of the sensitivity is an interval
number consisting of all the numbers between a pair of given num-
bers. The two given numbers are referred to as upper and lower
limits. Using the loss interval sensitivities makes it more reason-
able to identify weakness in the distribution system and establish
reconstruction measures in time.
This paper applies the three-phase system modeling technique
[15] including nine transformer connection types (both grounded
and ungrounded connections), lines, switches, and ZIP loads. In
http://dx.doi.org/10.1016/j.ijepes.2014.08.022
0142-0615/Ó 2014 Elsevier Ltd. All rights reserved.
⇑
Corresponding author. Address: School of Electrical Engineering, Yanshan
University, Qinhuangdao, Hebei 066004, PR China.
E-mail addresses: zhglu@ysu.edu.cn (Z. Lu), ljia1987@sina.com (J. Liu),
lyeysu@163.com (Y. Liu), 1250094720@qq.com (R. Ding), faayaa2013@hotmail.
com (F. Yang).
Electrical Power and Energy Systems 64 (2015) 931–936
Contents lists available at ScienceDirect
Electrical Power and Energy Systems
journal homepage: www.elsevier.com/locate/ijepes