R. Jamal, J. Zhang, B. Men et al.

Energy Reports 9 (2023) 4157–4173

• Unlike the most of presented paper in literature that solved 2.1.2. Voltage deviations reduction
the ORPD at deterministic state, this paper solved the ORPD The second objective function aims to improve the voltage
with and without considering the stochastic nature of re- profile, directly related to reducing voltage deviations in electrical
newable energy resources and the load demand. networks. It is calculated as the summation of the voltage devi-
• Solving the stochastic ORPD with inclusion renewable en- ations of load buses from the reference voltage, which ordinarily
ergy resources (RERs) including solar and wind-hydro pow- equals to 1 p.u. Thus, the second objective function (F2 ) can be
ers generation systems. expressed as follows:
• Proposing a novel modified Artificial Hummingbird Algo- Nlb
rithm (MAHA) based on the levy flight and the distance ∑ ⏐( )⏐
F2 = VD = ⏐ Vi − Vref ⏐ (7)
bandwidth motions for solving the deterministic and stoch-
i=1
astic ORPD.
• Application of the MCS and SBR along with the normal, where Nlb refers to the number of connected load buses, Vi repre-
lognormal, Weibull, and Gumbel PDFs for representing the sents the ith load bus voltages while Vref represents the reference
uncertainties of the system. voltage.
• A statistical comparison is presented for verifying the ef-
fectiveness of the proposed algorithm compared to other 2.1.3. Voltage stability enhancement
well-known optimizers. The third objective function aims to enhance the voltage sta-
bility taken as the hazardous phenomena cause instability in
The paper is organized as follows: the mathematical formulation
the power networks. The voltage fluctuation can harm or cause
of ORPD with objectives functions is presented in Section 2;
voltage collapse suddenly or steadily in the power networks
formulation-related uncertainties of loading, wind speed, and
that should be required to improve to avoid voltage collapse.
solar irradiance with river flow rates are presented in Section 3;
So, to enhance the voltage stability objective (F3 ), the following
methodology of the novel modified AHA algorithm is introduced
expression is counted as follows:
in Section 4; results and the discussions are discussed in Sec-
tion 5; while the last part of this paper is conclusion given in ∀ j = 1, 2, . . . , Nbus
( ( ))
F3 = min(Lmax ) = min max Lj (8)
Section 6.
NG
∑ Vi
Lj = |1 − Fji |∀ j = 1, 2, . . . , NL (9)
2. Problem formulation of ORPD Vj
i=1

The optimal reactive power dispatch (ORPD) solution aims where, Lj is the stability index at jth bus. Fji = −[Y1 ]−1 [Y2 ]. Y1
to identify the best control parameter sets to enhance the sys- and Y2 are the systems YBUS sub-matrices.
tem performance with satisfying the operational constraints. The
mathematical formulation of the ORPD problem is expressed by 2.2. Constraints
using the following equations:

Minimize F (o, w ) (1) 2.2.1. Equality constraints

Subject to NB
∑
⏐Vj ⏐ Gij cos δij + Bij sin δij
⏐ ⏐( )
ek (o, w) = 0k = 1, 2, . . . , p (2) PGi − PLi = |Vi | (10)
j=1
in (o, w) ≤ 0n = 1, 2, . . . , m (3)
NB
∑
where ek and in represent the equality and inequality constraints, ⏐Vj ⏐ Gij sin δij − Bij cos δij
⏐ ⏐( )
QGi − QLi = |Vi | (11)
w is a vector that represents the control parameters, including j=1
the voltages of the generators (VG ), reactive powers of capacitors
(QC ) as well as transformer taps (Tp ). o is a vector that represents
2.2.2. Inequality constraints
the dependent variables which include slack power bus (Pb ), load
buses voltages (VL ), power flow of transmission lines (ST ) and
the reactive power of generators (QG ). The vectors w and o are ,k ≤ PG,k ≤ PG,k k = 1, 2, . . . , NG
PGmin max
(12)
represented as follows:
QGmin
,k ≤ QG,k ≤ QGmax
,k k = 1, 2, . . . , NG (13)
w = [QC , Tp , VG ] (4)
,k ≤ VG,k ≤ VG,k k = 1, 2, . . . , NG
VGmin max
(14)
o = [ST , QG , VL , Pb ] (5)
,n ≤ Tp,n ≤ Tp,n n = 1, 2, . . . , Np
Tpmin max
(15)
2.1. Objective functions
Qcmin
,n ≤ Qc ,n ≤ Qcmax
,n n = 1, 2, . . . , Nc (16)
2.1.1. The power losses reduction
,n n = 1, 2, . . . , NT
ST ,n ≤ STmin (17)
The first objective function aims to minimize the active power
losses, which can be represented as follows: VLmin
,n ≤ VL,n ≤ VLmax
,n n = 1, 2, . . . , Nq (18)
nL
∑ where, max and min superscripts are denoted for the upper
F1 = PLoss = gij (Vi2 + Vj2 − 2Vi Vj cosδij ) (6) and the lower limits of power generators, reactive power, volt-
i=1
age magnitude, transformer taps, and shunt VAR compensators,
where gij indicated the transmission line inductance, nL repre- whereas NG , Nc , NT , Nq and NL , represents the number of genera-
sents the number of transmission lines, whereas Vi and Vj are the tors, capacitors, transformers, load buses, and transmission lines.
magnitudes of the voltages at buses i and j. The system constraints can be considered using the weight sum
4159