CC BY
1
UDC 519.6
DOI: 10.18698/1812-3368-2023-5-4-31
TRUE POWER LOSS DWINDLING AND STABILITY AUGMENTATION BY EXTREME LEARNING MACHINE BASED HYBRID LEPIDOPTERA-LABIDOGNATHA ALGORITHMS AND RHINOTIA HAEMOPTERA BASED HYBRID CANIS AUREUS GIRNEYS OPTIMIZATION ALGORITHM
L. Kanagasabai gklenin@gmail.com
Prasad V. Potluri Siddhartha Institute of Technology, Kanuru, Vijayawada, India
Abstract
In this paper Extreme Learning Machine based Hybrid Lepidoptera-Labidognatha Algorithms and Rhinotia haemoptera based Hybrid Canis aureus and Girneys optimization algorithm has been applied for solving the power loss lessening problem. In Lepidoptera algorithm Location and stage are rationalized in all iteration. The location modernizing procedure is sustained iteratively up until the end norm is satisfied. And in Labidognatha algorithm every Labidognatha in population, subsequent to the capricious walk step, will have diminutive probability to make a decision on not following its current target and bound away from its existing position. A vibration is spread over the web when a Labi-dognatha shifts to a new-fangled location. Every vibration seizes the information of one Labidognatha and other Labidognatha can get the information in receipt of the vibration. At its current location it creates vibration when Labidognatha moves to a novel position. In this paper Extreme Learning Machine based Hybrid Lepidoptera and Labidognatha Algorithms is designed to solve the problem. Then in this paper Rhinotia haemoptera based hybrid Canis aureus and Girneys optimization algorithm is modelled for solving the problem. In Canis aureus optimization algorithm deeds of the Canis aureus are used to formulate the algorithm. Through stalking, sneaking and jumping on prey, it hunts. Canis aureus optimization algorithm algorithm imitates the behaviour of Canis aureus as Discover and Stalk segment. Girneys algorithm imitate the deeds of the Girneys have been imitated to formulate the algorithm. Dominant male run the subgroups on the periphery of the central group and communicates
Keywords
Optimal reactive power, transmission loss, extreme learning machine, Lepidoptera, Labidognatha, Rhinotia haemoptera, Canis aureus, Girneys
messages between the peripheral males and the central. In the projected Rhinotia haemoptera based hybrid Canis aureus and Girneys optimization algorithm Portent Canis aureus will control the quarry expanse by the complete pragmatic from earlobes. This exploit is very alike to the doings of Rhinotia haemoptera drive. Then a modernizing strategy which grounded on the cosine function is used to control the process of the algorithm for evading the local optima. Then Girneys movement are included in the hybridized algorithm. Legitimacy of the Extreme Learning Machine based Hybrid Lepidop-tera-Labidognatha algorithms and Rhinotia haemop-tera based hybrid Canis aureus and Girneys optimization algorithm is substantiated in IEEE 30 bus system (with and devoid of L-index). Actual power loss lessen- Received 26.08.2021 ing is reached. Proportion of actual power loss lessening Accepted 25.10.2021 is augmented © Author(s), 2023
Introduction. Optimal reactive power dispatch is deliberated as one of the significant conditions for safe and pecuniary operation of a system. It is attained by appropriate organization of the structure apparatus used to cope up the power flow with the goal of diminishing the true power losses and progress the voltage outline of the structure. Zhu et al [1] solved the optimal reactive power control using modified interior point method. Quintana et al [2] did reactive-power dispatch by successive quadratic programming. Jan et al [3] did application of the fast Newton — Raphson economic dispatch and reactive power/voltage dispatch by sensitivity factors to optimal power flow. Terra et al [4] did security-constrained reactive power dispatch. Grudinin [5] did reactive power optimization using successive quadratic programming method. Ebeed et al [6] did the optimal reactive power dispatch using marine predators algorithm considering the uncertainties in load and wind-solar generation systems. Sahli et al [7] did reactive power dispatch optimization with voltage profile improvement using an efficient hybrid algorithm. Davoodi et al [8] did a novel fast semi-definite programming-based approach for optimal reactive power dispatch. Bingane et al [9] applied tight-and-cheap conic relaxation for the optimal reactive power dispatch problem. Sahli et al [10] applied hybridized PSO-Tabu exploration for the problem. Mouassa et al [11] applied Ant lion algorithm for solving the problem. Mandal et al [12] solved the problem by using quasi-oppositional teaching. Khazali et al [13] solved the problem by harmony search procedure. Tran et al [14] solved problem by innovative enhanced stochastic fractal search procedure. Polprasert et al [15] solved the
problem by using enhanced pseudo-gradient pursuit particle swarm optimization. Thanh et al [16] solved the problem by an Operative Metaheuristic Procedure. Raghuwanshi et al [17] did class imbalance learning using under bagging based kernelized extreme learning machine. Yu et al [18] had done dual-weighted kernel extreme learning machine for hyperspectral imagery classification. Lv et al [19] did hyperspectral image classification based on multiple reduced kernel extreme learning machine. From Illinois Center [20] for a smarter electric grid (ICSEG) IEEE 30 bus system data obtained. Dai et al
[21] used seeker optimization procedure for solving the problem. Subbaraj et al
[22] used self-adaptive real coded genetic procedure to solve the problem. Pandya et al [23] applied particle swarm optimization to solve the problem. Hussain et al [24] applied amended particle swarm optimization to solve the problem. Vishnu et al [25] applied an enhanced particle swarm optimization to solve the problem. Omelchenko et al [26-28] did development of a design algorithm for the logistics system of product distribution of the mechanical engineering enterprise, did the work on organization of logistic systems of scientific productions, solved the problems and organizational and technical solutions of processing management problems of material and technical resources in a design-oriented organization. Khunkitti et al [29] solved multi-objective optimal power flow problems based on slime mould algorithm. Diab et al [30] solved multi-objective optimal power flow control of electrical transmission networks using intelligent meta-heuristic optimization techniques. Yet many approaches failed to reach the global optimal solution. In this paper, Extreme Learning Machine based Hybrid Lepidoptera-Labidognatha algorithms (ELMLLA) and Rhinotia haemoptera based Hybrid Canis aureus and Girneys (RCG) optimization algorithm is applied to solve the factual power loss lessening problem. Lepidoptera algorithm is based on natural deeds of Lepidoptera and they are aquatic or semi-aquatic. Adult Lepidoptera is commonly seen adjacent to bodies of aquatic and are omnivorous all through their life, primarily feeding on petty insects. Usually Lepidoptera are regularly seen close to bodies of aquatic and on the ventral side of Male Lepidoptera have an ovulatory structure an additional abdominal segment will be there. Lepidoptera can function as bio indicators their great quantity designates the profusion of prey in ecosystem. Species prosperity of vascular plants has also been positively related with the classes' richness of Lepidoptera in a positive habitat. Labidognatha algorithm imitates the deeds of Labidognatha and they could not go away from the web as the location of the web stand for infeasible solutions. Every Labidognatha in population, subsequent to the capricious walk
step, will have diminutive probability to make a decision on not following its current target and bound away from its existing position. A vibration is spread over the web when a Labidognatha shifts to a new-fangled location. Every vibration seizes the information of one Labidognatha and other Labidognatha can get the information in receipt of the vibration. Vibration Concentration is in the range of [0, . At its current location it creates vibration when Labidognatha moves to a novel position. In Lepidoptera algorithm Location and stage are rationalized in all iterations. Apprising O and AO vectors, is by computing the Euclidean distance and pick N of them. The location modernizing procedure is sustained iteratively up until the end norm is satisfied, and in Labidognatha algorithm every Labidognatha in population, subsequent to the capricious walk step, will have diminutive probability to make a decision on not following its current target and bound away from its existing position. A vibration is spread over the web when a Labidognatha shifts to a new-fangled location. Every vibration seizes the information of one Labidognatha and other Labidognatha can get the information in receipt of the vibration. Vibration Concentration is in the range of [ 0, . At its current location it creates vibration when Labidognatha moves to a novel position. Then in this paper, RCG is proposed for solving the power loss lessening problem. Deeds of the Canis aureus have been modelled to formulate the algorithm. Canis aureus optimization (CO) algorithm imitates the behaviour of Canis aureus into two modes: 1) discover segment; 2) stalk segment. In discover segment the explore behaviour of Canis aureus have four main parameters, which are mentioned as follow: discover memory pools (DMP), search range of the chosen dimension (SRD), produce the dimension to change (PDC) and self location option (SLO). Girneys algorithm (GA) is based on actions of Girneys and it is brown or grey and based on rank in a group, Girneys position themselves. There are two or three oldest and most dominant males which are co-dominant, along with females, their infants, and juveniles in the central male subgroup. They occupying a great diversity of altitudes throughout Central, South, Southeast Asia and have the widest geographic ranges of any nonhuman primate. The farther to the periphery a subgroup is less dominant than the female. Determining the movements, foraging, and other routines are done by the subgroup. Girneys high-ranking individuals show little tolerance and often show relentless aggression towards non-kin. Sagacity of ELMLLA and RCG is confirmed by corroborated in IEEE 30 bus system (with and devoid of L-index). Factual power loss lessening is achieved. Proportion of factual power loss reduction is amplified.
Problem formulation. Power loss minimization is defined by min F (d, e). Subject to
A (d, e ) = 0, B (d, e ) = 0,
d = [ VLGi, ..., VLG^g; QCi,..., QCnc; Ti,..., Tnt ],
e = [PGsiack; VLi,..., VLnl; QGi,..., QGNg; SLi,..., SLnt ].
The fitness function (¿1, F2, F3) is designed for power loss (MW) lessening, voltage deviancy, voltage constancy index (L-index) is defined by
F = Pmin = min
NTL
X Gm
. m
V2 + V2 _ 2ViVj cos 0ij
F2 = min
nLB. , . ,.2 ng\
X |VLfc -VLpired\ +X\QGk -QGlk
1=1 i=1
lim I
F3 = min Lmax, Lmax = max [Lj ], j = 1, ..., Nlb,
and
NPV
Vi
Lj = 1 - E Fjirrr, Fji =-[Yi ] [Y ], Lmax = max
i = 1 V
1 -[Y1 Г Y]V
Vi
Vj
Parity constraints
0 = PGi -PDi - Vi X Vj [Gij cos [0,- -0j ] + Bij sin [0{ - 0j ]],
j e Nb
0 = QGi - QDi - Vi X Vj [Gij sin [0{ - 0j ] + Bij cos [0{ - 0j ]].
j e Nb
Disparity constraints
PGr* * PGslack < PGmk, QG™ < QGi < QG™x, i E Ng, VLm in < VI < VLf ax, i e NL, Tmin < I < Tmax, i e NT, QCmin < QC < QCmax, i e NC, | SI | < SLmax, i e NTL, VGmin < VG, < VGmax, i e Ng. Multi-objective fitness function:
MOF = F + rF + uF3 =
-|2
= f +
NL
2 Ng
Xxv [VLi - VLmin ] + [QGi - QG™ i = 1 i = 1
+ rfF3,
u is dependent variables;
VLmax, VLi > VLfax; . fQGmax, QGt > QGma
VImm = > t ' t t QGmin = I v t > v , v ,
VLmin, VLi < VL™, t I QGmin, QGt < QGtmil
Extreme Learning Machine based Hybrid Lepidoptera-Labidognatha algorithms. In this paper, ELMLLA is designed to solve the problem. Lepidoptera algorithm is based on natural deeds of Lepidoptera and they are aquatic or semi-aquatic. Adult Lepidoptera is commonly seen adjacent to bodies of aquatic and are omnivorous all through their life, primarily feeding on petty insects. Usually, Lepidoptera are regularly seen close to bodies of aquatic and on the ventral side of Male Lepidoptera have an ovulatory structure an additional abdominal segment will be there. Lepidoptera can function as bio indicators their great quantity designates the profusion of prey in ecosystem. Species prosperity of vascular plants has also been positively related with the classes' richness of Lepidoptera in a positive habitat. The performances are methodically modeled as shown below.
The disinterestedness is premeditated by
G =-£o-Oj. (1)
j = 1
Design of amalgamation is defined as
H
Z Aj
A - ^ <2>
Design of unison is described as
H
Z Oj
Ui = ^--O.
H
Collected stirring in the direction of the sustenance sources is articulated as Ct = O+ - O. Drive from entrant is calculated as Et = O~ + O.
Drive in the examine space is signposted by the step vector and it defined
as
AOt +1 = (tGi + hHt + kAi + bUi + fEt) + wAOf. (3)
Magnetism grade and step factor is scientifically carved as
Z„ext = zt + Ptj (Zjzyt) + a (R - 0.50). (4)
Location vectors are premeditated as
Ot +1 = Ot +AOt +1. (5)
Levy flight is encompassed to concentrated the exploration and it scientifically defined as
Ot +1 = Ot + Levy (z )Ot. (6)
Levy flight is a rank of non-Gaussian arbitrary processes and simple power-law formula L (s) ~ |s| 1 ^ , where 0 < P < 2 is an index;
L (s, у, |д) =
fa ■ етр
(s
3/2
if 0 S < œ;
2 (5
0, if 5 < 0, F(k) = exp -a\k|P , 0<p<2.
In the proposed method, the step sizes are produced using Levy distribution to examine the exploration zone and calculated as
55 (t) = 0.0010 5 (t) 5l. (7)
Solutions of modernized equation of the premium entity is defined as
zj (t + 1 ) = Zij (t ) + 5 _ 5 (t )U (0,1 ). (8)
Location and stage are rationalized in all iteration. Apprising O and AO vectors are by computing the Euclidean distance and pick N of them. The location modernizing procedure is sustained iteratively up until the end norm is satisfied:
a. Start
b. Engender the population
c. Steps vectors are initialized
d. Compute the objective function
e. Adjust the nutriment spring and contending factor
f. Calculate and modernize the crucial features
g. (1) h. (2)
i. Streamline nearby Lepidopteraradius
j. Streamline the drive
k. (3)
l. Modernize the location and step vectors
m. (4)
n. (5)
o. Streamline the exploration and location
p. (6)
q. Step sizes are produced by means of Levy distribution
r. (7)
s. Modernize the solution t. (8) u. End if
v. Articulate and control the renewed locations based on the limits of parameters
w. End while x. End
Labidognatha algorithm imitates the deeds of Labidognatha and they could not go away from the web as the location of the web stand for infeasible solutions. Every Labidognatha in population, subsequent to the capricious walk step, will have diminutive probability to make a decision on not following its current target and bound away from its existing position. A vibration is spread over the web when a Labidognatha shifts to a new-fangled location. Every vibration seizes the information of one Labidognatha and other Labidognatha can get the information in receipt of the vibration. Vibration concentration is in the range of [0, +«]. At its current location it creates vibration when Labidognatha moves to a novel position. Concentration value demarcated as
i L, is, t )41/(/Crr 7 (Lma"
V 7 [ 1/(f (Ls)-VCmin H min.
Extreme distance between two points in the examination space indicated as EDmax = X - Xl . Distance between Labidognatha is calculated as ED (La, Lb ) = = La - LbL. Vibration diminution for specific distance premeditated as
id i A id i A ( ED (La, Lb I (La, Lb, t) = I (La, La, t) exp--Ed- I.
V Edmaxra )
Vibration diminution over time is calculated as
I (La (t), La (t), t + 1 ) = I (La, La, t )ta.
Labidognatha on the web shift to novel position in iterations and fitness value is computed as Ls (t +1) = Ls +(Ltar -Ls )(1 -RA R).
Each Labidognatha in population, consequent to the variable walk step, will have miniature possibility to make a decision on not behind its present
target and bound away from its prevailing position is described by
L. - rj
l} -
exp ( Ed (Ls, Ltar ) / Edmax )
Labidognatha produce novel solutions for the succeeding iteration with the mutation procedure, after an assortment procedure is smeared as follows:
Ltrail = Ltarget + ngi ®^get - LT ).
Binary vector elements are produced by
_ J1, if Rij < VC; g^ [ 0 or else.
Succeeding to creation of the trail vector i, the position of Labidognatha i in the following iteration is calculated by:
¿t+1)JLtrail, if f (Lfail )< f (Li);
[ Li otherwise.
The procedure:
a. Start
b. Bounds values are allocated
c. Population of Labidognatha is produced with memory size
d. For each Labidognatha assign the value
e. while stop criteria not met do
f. For each Labidognatha in population do
g. Compute the fitness value
h. Each locations vibration are created
i. End for
j. For each Labidognatha in population do
k. Vibration formed by other Labidognatha is computed
l. Choose the premium vibration
m. Swap the best one to inferior one
n. End if
o. Capricious walk will be applied
p. Capricious number r will be formed from [0, 1]
q. Allocate an arbitrary location
r. End if
s. End
Extreme learning machine (ELM) is pragmatic and learning speed [17-19] of feed-forward neural networks is composed of input, hidden and output layer. Extreme learning machines are feed forward neural networks for classification, regression, clustering, sparse approximation, compression and feature learning with a single layer or multiple layers of hidden nodes, where the parameters of hidden nodes (not just the weights linking inputs to hidden nodes) need not be tuned. These hidden nodes can be arbitrarily allocated and certainly not rationalized (they are arbitrary protuberance but with nonlinear make over), or can be innate from their ancestors devoid of being altered. Almost in all cases, the output weights of hidden nodes are regularly learned in a solo step, which fundamentally amounts to learning a linear model.
For N
(zi, OWi); z = [zn, Zi2, •••, Zidn fW e SRdn, OWi = [OWi, OW-2, •", OWidn ]0W e SRdn,
N
X &mn ((oiZj + ai ) = OWj, j = 1,2, ..., N.
i = 1
Output matrix (QM) Output weight (|3) = OW
QM (Z1, , zl ; Ш1, , ; a1, , ai ) =
mn (ra1z1 + a1) ^ mn (raLz1 + aL)
• • • >
mn (ra1zN + a1) mn (&LzN + aL) p = QM "1OW.
Step by step ELM procedure is defined as:
a. Start
b. Input the parameters
c. Test and eraining sets are formed from the common information
d. With orientation to the "training set" of data — calculate the rate of QM
y. (9)
e. Calculate the rate of weight (output)
z. (10)
f. With orientation to the "test set" of data — calculate the rate of QM
aa. (9)
g. Calculate the actual value by p and QM
h. Compute the error rate
(9) (10)
i. Assessment of actual value with probable value
j. Return the error rate
k. End
Extreme Learning Machine based Hybrid Lepidoptera and Labidognatha Algorithms. In Lepidoptera algorithm location and stage are rationalized in all iteration. Apprising O and AO vectors are by computing the Euclidean distance and pick N of them. The location modernizing procedure is sustained iteratively up until the end norm is satisfied. In Labidognatha algorithm every Labidognatha in population, subsequent to the capricious walk step, will have diminutive probability to make a decision on not following its current target and bound away from its existing position. A vibration is spread over the web when a Labidognatha shifts to a new-fangled location. Every vibration seizes the information of one Labidognatha and other Labidognatha can get the information in receipt of the vibration. Vibration concentration is in the range of [0, +«]. At its current location it creates vibration when Labidognatha moves to a novel position. Here ELMLLA is designed to solve the problem. Figure 1 shows the flow chart of ELMLLA:
a. Start
b. Initialize the population in the search space
c. Build the training dataset
d. Based on fitness rate initial population is converted to training dataset
e. With orientation to training dataset ordering (ELM)will be trained
i. Start
ii. Input the parameters
iii. Test and Training sets are formed from the common information
iv. With orientation to the "training set" of data — calculate the rate of QM bb. (9)
v. Calculate the rate of weight (output) cc. (10)
vi. With orientation to the "test set" of data — calculate the rate of QM dd. (9)
vii. Calculate the actual value by p and QM
viii. Compute the error rate
ix. Assessment of actual value with probable value
x. Return the error rate
xi. End
f. Apply Lepidoptera algorithm
Input the NLS value
i. Start
ii. Engender the population
iii. Step vectors are initialized
iv. Compute the objective function
v. Adjust the nutriment spring and contending factor
vi. Calculate and modernize the crucial features
vii. (1)
viii. (2)
ix. Streamline nearby Lepidoptera radius
x. Streamline the drive ee. (3)
xi. Modernize the location and step vectors ff. (4)
gg. (5)
xii. Streamline the exploration and location hh. (6)
xiii. Step sizes are produced by means of levy distribution ii. (7)
xiv. Modernize the solution jj. (8)
xv. End if
xvi. Articulate and control the renewed locations based on the limits of parameters
xvii. End while
xviii. End
g. Apply Labidognatha algorithm along with mutation procedure
i. Start
ii. Bounds values are allocated
iii. Population of Labidognatha is produced with memory size
iv. For each Labidognatha assign the value i. while stop criteria not met do
v. For each Labidognatha in population do
vi. Compute the fitness value
vii. Each locations vibration are created
viii. End for
ix. Streamline the drive
x. Vibration formed by other Labidognatha is computed
xi. Choose the premium vibration
xii. Swap the best one to inferior one
xiii. End if
xiv. Capricious walk will be applied
xv. Capricious number r will be formed from [0, 1]
xvi. Allocate an arbitrary location
xvii. End if
xviii. End
h. Through trained ordering Lepidoptera are categorized
i. Outstanding vibration concentration is observed, which is in the range of [0, +<»]
j. Process ends when sum of valuation surpasses the extreme sum k. Otherwise
l. For successive generation "N" Lepidoptera are selected by selection of plan operative design m. End
Fig. 1. Flow chart of ELMLLA
Rhinotia haemoptera based Hybrid Canis aureus and Girneys optimization algorithm. Deeds of the Canis aureus have been modelled to formulate the algorithm. In discover segment the explore behaviour of Canis aureus have main parameters: DMP; SRD; PDC; SLO.
The method of explore mode as follow:
i. Engender j copies of the existing location of Canis aureus k, where j = = DMP. When the value of SLO is factual, let j = DMP - 1, conserve the current location as one of the candidates.
ii. As per PDC, for each replica, capriciously adjoin or deduct SRD percentage the present values and regenerate the preceding one.
iii. For all candidate points; fitness values have to be computed.
iv. Fitness values for all are not precisely equal, and then compute the selected possibility of each candidate by selecting the probability of each candidate point is 1.
v. chose the point in arbitrarily mode to shift to from the candidate points, and reinstate the location of Canis aureus k:
SLOi SLOmax I , ,
Hi =J-i-ma^. (11)
STO - STO •
Rendering to the objective of the problem Fb = Fmax. In stalk segment Canis aureus desire to sketch the goal and foods. The process of stalk segment can be outlined as follows:
i. Velocity of each dimension is calculated.
ii. Inside the range of maximum limit velocity should be there. If any violation is found, then it has to be brought back to the limit
Vk,d = Vk,d + riCi ( Zbest, d - Zk,d ), (12)
iii. Renovate the position of Canis aureus k:
Zk ,d = Zk ,d + Vk ,d. (13)
In (12), (13) Zbest, d is the location of the Canis aureus; Zk,d is the position of
Canis aureus k; c1 is an acceleration coefficient.
In the velocity equation an inertia weight is included. First large inertia value to enhance the global exploration and then inertia value will be abridged to move the stages of local search as follows:
W (/) = Ws . (14)
2 i_max
Acceleration coefficient formulation is used for modernizing as follows:
A (/) = A, + t-max-L . (15)
2 ¿max
Rationalized velocity is defined as
Vk,d = W(d) Vk,d + riA(d)(zbest4 -Zk). (16)
Modernizing the equations has been done by position data and movement data. Present and regular information of first and subsequent dimensions of both velocity and location are updated by applying a disremembering factor y:
Zk ,d = 1 [location info + progression info ], (17)
location info =
, (rZk,d+1) + (1 -y)(Zk,d +2 ) , (yZk,d-1) + (1 -y)(Zk,d-2 )
- Zk,d H---1--; (18)
2 2
progression info =
= | (yVk ,d+1) + (1 -y)( Vk, d+2 ) | (yVk ,d-1 ) + (1 ~y)( Vk,d-2 ) (19) , 2 2 ' The procedure:
a. Start
b. Engender N Canis aureus
c. Initialize the location, flag and velocity
d. Calculate the fitness value of the each Canis aureus
e. Preeminent Canis aureus (Zbest) is stored into memory
f. Rendering to Canis aureus flag, apply Canis aureus to the discover segment
g. (11)
h. Number of Canis aureus are chosen and set them into stalk segment
i. (16) j. (13) k. (14) l. (15) m. (16) n. (17) o. (18) p. (19)
q. End condition has to be checked, if satisfied, then end the process r. End
Girneys algorithm (GA) is based on actions of Girneys and it is brown or grey and based on rank in a group, Girneys position themselves. There are two or three oldest and most dominant males which are co-dominant, along with females, their infants, and juveniles in the central male subgroup. They occupying a great diversity of altitudes throughout Central, South, Southeast Asia and have the widest geographic ranges of any non-human primate. The farther to the periphery a subgroup is less dominant than the female. Determining the movements, foraging, and other routines are done by the subgroup. Girneys high-ranking individuals show little tolerance and often show relentless aggression towards non-kin. Deeds of the Girneys has been modelled as follows.
Population size is commenced ziM ^ zi < ziN. Position of each Girneys has been initiated zj = (zji, zj2, —, Zjn ), i = 1,2,..., n.
Drive of Girneys from primary positions to different position is defined as
Azj = ( Azj1, Azj2,..., Azjn ), i = 1,2,..., n,
azí, J «</>:0-s
' \-gT (,-g) = 0.5.
Replicated gradient is premeditated by
e (zj +Azj )-e (zj -Azj) ej = -u-^-,J-, j = 1,2,...,n,
2 yij
e'ij =(e'j1 (Zj ), e'j2 (zj ), , ejn (zj )),
yi = Zij + g (c'ij (yj)), j =1,2, -, n.
Drive is scientifically defined as
z e ( Xj - h, Xij + h ), j = 1,2, „., n, Zj = Xij + a ( Tj - zij),
1 K
Tj =—Zzij, j =1,2,,n. K i = 1
To poise the exploration and exploitation; info distribution is done by
mdi = Zdi +<&di (zdi -Zei), if mdi > zdmax ^ mdi = zdfx,
if mdi < zmax ^ mdi = zd?in.
Procedure: a. Start
b. Initialize the primary population
c. Compute the fitness value
d. Greedy selection approach used to categorise
e. Fresh positions are engendered
kk. OPnew tj = OPtj + 01 (OPkj - XYtj ) + 02 (OPrj - OPtj) +F (OPrj - OPtj)
f. Based on fitness value greedy selection process is applied between current locations and newly produced location- out of that finest one will be designated
g. Probability value for all the cluster associates will be premeditated by using
ll. pt = 0.9 (Ft / Fmax ) + 0.09
h. Modern locations are produced by
mm. OPnew tj = OPtj +01 ( OPkj - XYtj ) + 02 ( OPrj - OPtj) +F ( OPrj - OPtj)
i. Through greedy selection process the position updating will be done j. Updating of subgroup members by
nn. XYnew tj = XYtj +0( TUj - XYtj ) + 0(XYjj - MNkj) k. Adjoin all groups to articulate as a lone cluster l. Amend the location
m. Up until maximum number of iterations has been touched the process has to be followed n. End
Rhinotia haemoptera uses its appendages and olfactory pairs to investigate an anonymous atmosphere to discover regions with the sturdiest aroma of sustenance. In every phase, Rhinotia haemoptera mark the whiff with the appendages and then it chooses the progression of the succeeding footstep. Unusually, the Rhinotia haemoptera does not change indiscriminately in certain pathway nonetheless standstills consequently each footstep and customs the intelligence of whiff to enhanced acknowledgment of the independent course beforehand creating the succeeding statement.
In the projected RCG optimization algorithm Portent Canis aureus will control the quarry expanse by the complete pragmatic from earlobes. This exploit is very alike to the doings of Rhinotia haemoptera drive. Then a modernizing strategy which grounded on the cosine function is used to control the process of the CO algorithm for evading the local optima. Then Girneys movement are included in the hybridized algorithm. The advancement completed in the examination performance of Portent Canis aureus a is defined as
d=RiZ!,
R (Z ,1),
where R is random (-1, 1); Z is space length of search.
Grounded on the sound pragmatic by the earlobes of Portent Canis aureus a on the both left-hand and right sides the search is described as
Pf (n) = Pa (n)-D (n)®D, P£ght (n) = Pa (n) + D (n)®D,
z \ H (n)
D (n) =-v x .
v y ratio of(Hs)
Then the apprising of step is done by Hs (n +1) = p® Hs (n) + 0.01.
With orientation to earshot and examination action, the fresh spot of Portent Canis aureus a is described as
Psew (n ) = Pa{n ) + Hs (n )® D ® s (f (n ))-F (n
With orientation to the sound the Portent Canis aureus a will modernize the position by
(«) =
Pnew (n)F (Psew (n)<Pa(n)); Pa(n )F (P£ew (n )> Pa(n )).
Convergence element s of the cosine function is described as s = 2 cos
О >
к 2
, max iteration ,
v v y y
Figure 2 shows the flow chart of RCG optimization algorithm.
The procedure:
o. Start
p. Produce the population
q. Parameters are defined
r. Each search agent's fitness value are calculated
s. Define the pragmatic by the earlobes of Portent Canis aureus
t. while (n < maximum iteration)
u. For each search agent apprise the location
v. End for
(20)
w. Rendering to Canis aureus flag, apply Canis aureus to the Discover segment x. (11)
y. Number of Canis aureus are chosen and set them into stalk segment z. (12) aa. (13) bb. (14) cc. (15) dd. (16) ee. (17) ff. (18) gg. (19)
hh. Modernize parameter ii. (20)
jj. Agent's fitness values are computed
kk. Compute the fitness value
ll. Greedy selection approach used to categorise
mm. Fresh positions are engendered
nn. OPnew ij = OQj + 01 (Ofy - XYj ) + 02 (OPrj - OPj) +F (OPj - OPj) oo. Based on fitness value greedy selection process is applied between current locations and newly produced location — out of that finest one will be designated
pp. Probability value for all the cluster associates will be premeditated by using
qq. pi = 0.9 (Fi / Fmax ) + 0.09
rr. Modern locations are produced by
ss. OPnew ij = OPjj + 01 ( OPkj - XYj ) + 02 ( (OPrj - OPjj) +F ( OPrj - OPjj)
tt. Through greedy selection process the position updating will be done uu. Updating of subgroup members by
vv. XYnewij = XYij +0(TUj -XYij) + 0(XYij -MNkj)
ww. Adjoin all groups to articulate as a lone cluster; xx. Amend the location
yy. Based Rhinotia haemoptera an approach, renovate the way of movement
zz. Update the width
aaa. Reorganization of step length
bbb. Compute the fresh position of Canis aureus
ccc. Streamline the position of Portent Canis aureus
ddd. End for
eee. Update the values
fff. End
Fig. 2. Flow chart of RCG optimization algorithm
Simulation results. With considering voltage constancy, ELMLLA and RCG optimization algorithm is substantiated in IEEE 30 bus system [20]. Appraisal of loss has been done with PSO, amended PSO, enhanced PSO, widespread learning PSO, Adaptive genetic algorithm, Canonical genetic algorithm (GA), enriched genetic algorithm, Hybrid PSO-Tabu search (PSO-TS), Ant lion (ALO), Quasi-oppositional teaching learning based (QO-TLBO), Improved stochastic fractal search optimization algorithm (ISFS), Harmony search (HAS), Improved pseudogradient search particle swarm optimization and Cuckoo search algorithm. Power
loss abridged competently and proportion of the power loss lessening has been enriched. Predominantly voltage constancy enrichment achieved with minimized voltage deviancy. In Table 1 shows the loss appraisal, Table 2 shows the voltage deviancy evaluation and Table 3 gives the L-index assessment. Figure 3 gives graphical appraisal.
Table 1
Assessment of factual power loss lessening
Algorithm Factual power loss, MW Algorithm Factual power loss, MW
Standard PSO-TS [10] 4.5213 Standard PSO [13] 4.9239
Basic TS [10] 4.6862 HAS [13] 4.9059
Standard PSO [10] 4.6862 Standard FS [14] 4.5777
ALO [11] 4.5900 ISFS [14] 4.5142
QO-TLBO [12] 4.5594 Standard FS [16] 4.5275
TLBO [12] 4.5629 ELMLLA 4.4998
Standard GA [13] 4.9408 RCG 4.5001
Table 2
Evaluation of voltage deviation
Algorithm Voltage deviation, PU Algorithm Voltage deviation, PU
Standard PSO-TVIW [15] 0.1038 MPG-PSO [15] 0.0892
Standard PSO-TVAC [15] 0.2064 QO-TLBO [12] 0.0856
Standard PSO-TVAC [15] 0.1354 TLBO [12] 0.0913
Standard PSO-CF [15] 0.1287 Standard FS [14] / Standard FS [16] 0.1220 / 0.0877
PG-PSO [15] 0.1202 ISFS [14] 0.0890
SWT-PSO [15] 0.1614 ELMLLA 0.0830
PGSWT-PSO [15] 0.1539 RCG 0.0838
Table 3
Assessment of voltage constancy
Algorithm Voltage constancy, PU Algorithm Voltage constancy, PU
Standard PSO-TVIW [15] 0.1258 ALO [11] 0.1161
Standard PSO-TVAC [15] 0.1499 / 0.1271 ABC [11] 0.1161
Standard PSO-CF [15] 0.1261 GWO [11] 0.1242
End of the table 3
Algorithm Voltage constancy, PU Algorithm Voltage constancy, PU
PG-PSO [15] 0.1264 BA [11] 0.1252
Standard WT-PSO [15] 0.1488 Basic FS [14] 0.1252
PGSWT-PSO [15] 0.1394 ISFS [14] 0.1245
MPG-PSO [15] 0.1241 Standard FS [16] 0.1007
QO-TLBO [12] 0.1191 ELMLLA 0.1002
TLBO [12] 0.1180 RCG 0.1004
Fig. 3 (beginning). Appraisal of factual power loss (a), voltage deviation (b)
>, 0.16
g 0.14
1 0.12 h
g 0.10
8 0.08
S) 0.06
Jr 0.04
3 0.02
> 0
AW
^vv c
Fig. 3 (ending). Appraisal of voltage constancy (c)
Then projected ELMLLA and RCG optimization algorithm is corroborated in IEEE 30 bus test system deprived of L-index. Loss appraisal is shown in Table 4. Figure 4 gives graphical appraisal between the approaches with orientation to factual power loss.
Table 4
Assessment of true power loss
Algorithm Factual power loss, MW Proportion of lessening in power loss
Base case value [24] 17.5500 0
Amended PSO[24] 16.0700 8.40000
Standard PSO [23] 16.2500 7.40000
Standard EP[21] 16.3800 6.60000
Standard GA [22] 16.0900 8.30000
Basic PSO [25] 17.5246 0.14472
DEPSO [25] 17.52 0.17094
JAYA [25] 17.536 0.07977
ELMLLA 13.84 21.1396
RCG 14.05 19.9430
Table 5 shows the convergence characteristics of ELMLLA and RCG optimization algorithm. Figure 5 shows the graphical representation of the characteristics.
25
Fig. 4. Appraisal of factual power loss ( ), and proportion of lessening in power loss ( )
Table 5
Convergence characteristics
Algorithm Factual power loss with / without L-index, MW Time with / without L-index, s Number of iterations with / without L-index
ELMLLA 4.4998 / 13.84 20.90 / 18.99 33 / 29
RCG 4.5001 / 14.05 19.49 / 18.12 31 / 27
ELMLLA
RCG
Fig. 5. Convergence characteristics:
factual power loss, MW: with (■) or without (■) L-index; time, s: with (■) or without (■) L-index; number of iterations with (■) or (■) without L-index
Conclusion. Extreme Learning Machine based Hybrid Lepidoptera-Labidognatha algorithms and RCG optimization algorithm abridged the factual power loss dexterously. Extreme Learning Machine based Hybrid Lepidoptera-Labidognatha algorithms and RCG substantiated in IEEE 30-bus test system with and devoid of voltage constancy. In Lepidoptera algorithm location and stage are rationalized in all iteration. Apprising O and AO vectors are by computing the Euclidean distance and pick N of them. The location modernizing procedure is sustained iteratively up until the end norm is satisfied. In Labidognatha algorithm every Labidognatha in population, subsequent to the capricious walk step, will have diminutive probability to make a decision on not following its current target and bound away from its existing position. A vibration is spread over the web when a Labidognatha shifts to a new-fangled location. Every vibration seizes the information of one Labidognatha and other Labidognatha can get the information in receipt of the vibration. Vibration Concentration is in the range of [0, +«]. At its current location it creates vibration when Labidognatha moves to a novel position. Then in this paper RGG splendidly solved the power loss lessening problem. Canis aureus optimization algorithm imitates the behaviour of Canis aureus into discover and stalk segments modes. In discover segment the explore behaviour of Canis aureus have main parameters: DMP; SRD; PDC; SLO. Girneys algorithm is based on actions of Girneys and it is brown or grey and based on rank in a group, Girneys position themselves. Determining the movements, foraging, and other routines are done by the subgroup. Girneys high-ranking individuals show little tolerance and often show relentless aggression towards non-kin. In the projected Rhinotia haemoptera based hybrid Canis aureus and Girneys optimization algorithm (RCG) Portent Canis aureus will control the quarry expanse by the complete pragmatic from earlobes. This exploit is very alike to the doings of Rhinotia haemoptera drive. Then a modernizing strategy which grounded on the cosine function is used to control the process of the CO algorithm for evading the local optima. Then Girneys movement are included in the hybridized algorithm. Extreme Learning Machine based Hybrid Lepidoptera-Labidognatha algorithms and RCG optimization algorithm com-mendably reduced the power loss and proportion of factual power loss lessening has been upgraded. Convergence characteristics show the better performance of the proposed ELMLLA and RCG algorithms. Assessment of power loss has been done with other customary reported algorithms.
REFERENCES
[1] Zhu J.Z., Xiong X.F. Optimal reactive power control using modified interior point method. Electr. Power Syst. Res., 2003, vol. 66, iss. 2, pp. 187-192.
DOI: https://doi.org/10.1016/S0378-7796(03)00078-6
[2] Quintana V.H., Santos-Nieto M. Reactive-power dispatch by successive quadratic programming. IEEE Trans. Energy Convers., 1989, vol. 4, iss. 3, pp. 425-435.
DOI: https://doi.org/10.1109/60.43245
[3] Jan R.-M., Chen N. Application of the fast Newton — Raphson economic dispatch and reactive power/voltage dispatch by sensitivity factors to optimal power flow. IEEE Trans. Energy Convers., 1995, vol. 10, iss. 2, pp. 293-301.
DOI: https://doi.org/10.1109/60.391895
[4] Terra L.D.B., Short M.J. Security-constrained reactive power dispatch. IEEE Trans. Power Syst., 1991, vol. 6, iss. 1, pp. 109-117. DOI: https://doi.org/10.1109/59.131053
[5] Grudinin N. Reactive power optimization using successive quadratic programming method. IEEE Trans. Power Syst., 1998, vol. 13, iss. 4, pp. 1219-1225.
DOI: https://doi.org/10.1109/59.736232
[6] Ebeed M., Alhejji A., Kamel S., et al. Solving the optimal reactive power dispatch using marine predators algorithm considering the uncertainties in load and wind-solar generation systems. Energies, 2020, vol. 13, iss. 17, art. 4316.
DOI: https://doi.org/10.3390/en13174316
[7] Sahli Z., Hamouda A., Bekrar A., et al. Reactive power dispatch optimization with voltage profile improvement using an efficient hybrid algorithm. Energies, 2018, vol. 11, iss. 8, art. 2134. DOI: https://doi.org/10.3390/en11082134
[8] Davoodi E., Babaei E., Mohammadi-Ivatloo B., et al. A novel fast semidefinite pro-gramming-based approach for optimal reactive power dispatch. IEEE Trans. Industr. Inform., 2020, vol. 16, iss. 1, pp. 288-298.
DOI: https://doi.org/10.1109/TII.2019.2918143
[9] Bingane C., Anjos M.F., Le Digabel S. Tight-and-cheap conic relaxation for the optimal reactive power dispatch problem. IEEE Trans. Power Syst., 2019, vol. 34, iss. 6, pp. 4684-4693. DOI: https://doi.org/10.1109/TPWRS.2019.2912889
[10] Sahli Z., Hamouda A., Bekrar A., et al. Hybrid PSO-tabu search for the optimal reactive power dispatch problem. IECON, 2014, pp. 3536-3542.
DOI: https://doi.org/10.1109/IECON.2014.7049024
[11] Mouassa S., Bouktir T., Salhi A. Ant lion optimizer for solving optimal reactive power dispatch problem in power systems. Eng. Sci. Technol. an Int. J., 2017, vol. 20, iss. 3, pp. 885-895. DOI: https://doi.org/10.1016/j.jestch.2017.03.006
[12] Mandal B., Roy P.K. Optimal reactive power dispatch using quasi-oppositional teaching learning based optimization. Int. J. Electr. Power Energy Syst., 2013, vol. 53, pp. 123-134. DOI: https://doi.org/10.1016/j.ijepes.2013.04.011
[13] Khazali H., Kalantar M. Optimal reactive power dispatch based on harmony search algorithm. Int. J. Electr. Power Energy Syst., 2011, vol. 33, iss. 3, pp. 684-692.
DOI: https://doi.org/10.1016Zj.ijepes.2010.11.018
[14] Tran H.V., Pham T.V., Pham L.H., et al. Finding optimal reactive power dispatch solutions by using a novel improved stochastic fractal search optimization algorithm. TELKOMNIKA, 2019, vol. 17, no. 5, pp. 2517-2526.
DOI: http://doi.org/10.12928/telkomnika.v17i5.10767
[15] Polprasert J., Ongsakul W., Dieu V.N. Optimal reactive power dispatch using improved pseudo-gradient search particle swarm optimization. Electr. Power Compon. Syst., 2016, vol. 44, iss. 5, pp. 518-532.
DOI: https://doi.org/10.1080/15325008.2015.1112449
[16] Duong T.L., Duong M.Q., Phan V.-D., et al. Optimal reactive power flow for large-scale power systems using an effective metaheuristic algorithm. J. Electr. Comput. Eng., 2020, vol. 2020, art. 6382507. DOI: https://doi.org/10.1155/2020/6382507
[17] Raghuwanshi B.S., Shukla S. Class imbalance learning using UnderBagging based kernelized extreme learning machine. Neurocomputing, 2019, vol. 329, pp. 172-187. DOI: https://doi.org/10.1016/j.neucom.2018.10.056
[18] Yu X., Feng Y., Gao Y., et al. Dual-weighted kernel extreme learning machine for hyperspectral imagery classification. Remote Sens., 2021, vol. 13, iss. 3, art. 508.
DOI: https://doi.org/10.3390/rs13030508
[19] Lv F., Han M. Hyperspectral image classification based on multiple reduced kernel extreme learning machine. Int. J. Mach. Learn. Cybern., 2019, vol. 10, no. 6, pp. 33973405. DOI: https://doi.org/10.1007/s13042-019-00926-5
[20] Illinois Center for a Smarter Electric Grid (ICSEG). Available at: https://icseg.iti.illinois.edu (accessed: 06.08.2023).
[21] Dai C., Chen W., Zhu Y., et al. Seeker optimization algorithm for optimal reactive power dispatch. IEEE Trans. Power Syst., 2009, vol. 24, iss. 3, pp. 1218-1231.
DOI: https://doi.org/10.1109/TPWRS.2009.2021226
[22] Subbaraj P., Rajnarayan P.N. Optimal reactive power dispatch using self-adaptive real coded genetic algorithm. Electr. Pow. Syst. Res., 2009, vol. 79, iss. 2, pp. 374-381. DOI: https://doi.org/10.1016Aj.epsr.2008.07.008
[23] Pandya S., Roy R. Particle swarm optimization based optimal reactive power dispatch. Proc. ICECCT, 2015. DOI: https://doi.org/10.1109/ICECCT.2015.7225981
[24] Hussain A.N., Abdullah A.A., Neda O.M. Modified particle swarm optimization for solution of reactive power dispatch. Res. J. Appl. Sci. Eng. Technol., 2018, vol. 15, no. 8, pp. 316-327. DOI: http://dx.doi.org/10.19026/rjaset.15.5917
[25] Vishnu M., Kumar T.K.S. An improved solution for reactive power dispatch problem using diversity-enhanced particle swarm optimization. Energies, 2020, vol. 13, iss. 11, art. 2862. DOI: https://doi.org/10.3390/en13112862
[26] Omelchenko I.N., Lyakhovich D.G., Aleksandrov A.A., et al. Development of a design algorithm for the logistics system of product distribution of the mechanical engineering enterprise. Herald of the Bauman Moscow State Technical University, Series Mechanical Engineering, 2020, no. 3 (132), pp. 62-69.
DOI: https://doi.org/10.18698/0236-3941-2020-3-62-69
[27] Omelchenko I.N., Zakharov M.N., Lyakhovich D.G., et al. [Organization of logistic systems of scientific productions: scientific research work of the master's student and evaluation of its results]. Sistemy upravleniyapolnym zhiznennym tsiklom vysokotekhno-logichnoy produktsii v mashinostroenii: novye istochniki rosta. Mater. III vseros. nauch.-prakt. konf. [Organisation of Logistics Systems for Knowledge-Intensive Industries: Master's Student Research Work and Evaluation. Proc. III Russ. Sci.-Pract. Conf.]. Moscow, Pervoe ekonomicheskoe izdatelstvo Publ., 2020, pp. 252-256 (in Russ.). DOI: https://doi.org/10.18334/9785912923258.252-256
[28] Omelchenko I.N., Lyakhovich D.G., Aleksandrov A.A., et al. [Problems and organizational and technical solutions of processing management problems of material and technical resources in a design-oriented organization]. Sistemy upravleniya polnym zhiznennym tsiklom vysokotekhnologichnoy produktsii v mashinostroenii: novye istochniki rosta. Mater. III vseros. nauch.-prakt. konf. [Management systems for the full life cycle of high-tech products in mechanical engineering: new sources of growth. Proc. III All-Russ. Sci. Pract. Conf.]. Moscow, Pervoe ekonomicheskoe izdatelstvo Publ., 2020, pp. 257-260 (in Russ.). DOI: https://doi.org/10.18334/9785912923258.257-260
[29] Khunkitti S., Siritaratiwat A., Premrudeepreechacharn S. Multi-objective optimal power flow problems based on slime mould algorithm. Sustainability, 2021, vol. 13, iss. 13, art. 7448. DOI: https://doi.org/10.3390/su13137448
[30] Diab H., Abdelsalam M., Abdelbary A. A multi-objective optimal power flow control of electrical transmission networks using intelligent meta-heuristic optimization techniques. Sustainability, 2021, vol. 13, iss. 9, art. 4979.
DOI: https://doi.org/10.3390/su13094979
Kanagasabai Lenin — Dr. Sc. (Full), Professor, Department of Electrical and Electronics Engineering, Prasad V. Potluri Siddhartha Institute of Technology (Kanuru, Vijayawada, Andhra Pradesh, 520007 India).
Please cite this article as:
Kanagasabai L. True power loss dwindling and stability augmentation by Extreme Learning Machine based Hybrid Lepidoptera-Labidognatha Algorithms and Rhinotia haemoptera based Hybrid Canis Aureus Girneys Optimization algorithm. Herald of the Bauman Moscow State Technical University, Series Natural Sciences, 2023, no. 5 (110), pp. 4-31. DOI: https://doi.org/10.18698/1812-3368-2023-5-4-31
