Yanbu Journal

of
Engineering
and Science
Vol. 1, October 2010 (1431H)
ISSN: 1658-5321 www.yic.edu.sa/yjes

ARTIFICIAL NEURAL NETWORK APPROACH: AN APPLICATION TO HARMONIC

LOAD FLOW FOR RADIAL SYSTEMS

A.Arunagiri1 , Suresh K2, B.Venkatesh3, R.Ramesh Kumar1 and Mustajab Ahmed Khan1

1
EEET Department, Yanbu Industrial College, Yanbu, KSA.
2
Research Scholar, Bharath University, Chennai, India.
3
Faculty of Engineering, Ryerson University, Toronto, Canada.
Contact: [email protected].
ABSTRACT

Radial Distribution Systems (RDS) require special load flow methods to solve power
flow equations owing to their high R/X ratio. Increasing use of power electronic devices
and effect of magnetic saturation cause harmonics in RDS. This paper reports a multi-
layer feed forward ANN with error back propagation learning algorithm for the
calculation of bus voltages and power loss for different harmonic components. The
proposed method is tested upon a 33-bus RDS and the results are reported for various
harmonics. Extensive testing of the proposed ANN based approach indicates its viability
for harmonic load flow assessment for radial systems.

KEYWORDS: Radial Distribution Systems; Harmonic components.

1. INTRODUCTION Many existing methods for distribution

system load flow, fail to obtain a solution in
Analysis of distribution system using power
several instances. Large RDS have
flow is important in the field of power systems.
complicated structure and are subject to
Distribution systems are predominantly
changes in their topology frequently for
characterized by their high R/X ratio and
maintenance, load balancing, network
radial topology. Matrix based iterative
reconfiguration and emergency operations
methods do not lend themselves for radial
under the umbrella of Supervisory Control
distribution systems owing to these
and Data Acquisition (SCADA). SCADA
characteristics. Numerous algorithms have
requires a fast Distribution Load Flow (DLF)
been developed using simple recursive
algorithm that computes the voltage solution
equations [1-3].
very rapidly for online scheduling.
Rapid industrialization has led to increasing
Load flow calculation in harmonic polluted
use of power electronic devices in
radial system with distributed generation has
transmission and distribution systems.
been carried out using abstract data types with
Modern industrial and domestic consumers
complex parameters[4].
use an ever-increasing number of devices that
A multiple-frequency three-phase load-flow
primarily employ power electronics based
with two sub models including the
power-conditioners. Use of AC machines
fundamental power flow (FPF) and harmonic
employing magnetic circuits in the saturation
frequency power-flow (HPF) model has been
region also introduces harmonics in electrical
developed and the standard Fourier analysis
power systems.
Received, April 15, 2010; accepted October 6, 2010
-1-
 Yanbu Journal of Engineering and Science Vol. 1 (2010)

was used to deal with the harmonic loads to recognition phase. Section 4 discusses an
get injection currents [5]. Fuzzy number implementation of BPN for determining the
based methodology for harmonic load-flow bus voltages for various harmonics. Section 5
calculation including uncertainities has been presents the results of sample systems being
applied for interconnected system[6]. studied by the proposed method for different
Artificial neural network approach has been harmonics. Section 6 presents the conclusion.
applied for the radial distribution system
analysis [7].From the above, one may see the
need for an efficient algorithm that reliably
and rapidly solves the power flow equations
for radial distribution systems characterized
by high R/X ratio, radial topology and for
various harmonic loads. Fig. 1. Ladder Network
In this paper an ANN based harmonic load 2. LADDER ITERATIVE TECHNIQUE
flow solution technique for the radial system
has been developed. A database consisting of It is assumed that the ladder network
different load patterns and the corresponding parameters for lines, loads and substation
voltage solution with the power loss is created voltage VS are known. The voltage solution of
for third, fifth and seventh order of harmonics this network can be obtained by repeating the
using ladder iterative technique. The neural forward and backward sweeps iteratively.
network is trained to learn the features of the 2.1 FORWARD SWEEP:
load to estimate the bus voltage, angle and the Compute bus voltages and associated currents
total loss. The trained neural network can be starting from last bus to the first bus.
instantly recalled to give output for an
Ii = (Si Vi ) ⎫
*

untrained set of inputs without going through ⎪

Ii −1,i = Ii + Ii,i +1 ⎬ for i = 4, 3, 2, 1. (1)
the conventional iterative procedure, and that ⎪
Vi = Vi +1 + Zi,i +1 * Ii -1,i ⎭
saves considerable execution time especially
on a large systems. For i=5, V5 is assumed to be VS in the first
The proposed method makes use of multi- iteration and equals the value computed in
layer feed forward ANN with error back the backward sweep in the subsequent
propagation learning algorithm for the iterations. I5 is computed using (1).
calculation of bus voltages and power loss for 2.2 BACKWARD SWEEP:
different harmonic components. The backward sweep starts from 2nd bus to the
In section 2, a simple ladder network last bus (5th bus). Taking V1=VS. The ith bus
technique is explained for solving the radial voltages are computed as below using current
system power balance equations. Using this values computed in the forward sweep:
technique, a data base providing information Vi = Vi -1 + Z i -1, i * I i -1, i for i = 2,3,4,5. (2)
of the possible real and reactive power
The forward and backward sweeps are
demands for various harmonics at different
continued until the difference between the
buses and their corresponding voltage
specified voltage at source and computed
solution is created. Section 3 briefly
voltage in the forward sweep is within the
introduces the Back Propagation Network, its
tolerance limit.
architecture, training algorithm and

-2-
 Yanbu Journal of Engineering and Science Vol. 1 (2010)

3. BPN ARCHITECTURE
X1 1 1 1 Y1
The most common BPN architecture is
presented in Fig. 2. It is shown to have three
layers, namely, input, hidden and output
layers. Other applications may have several Xi i j k Yi
hidden layers. During training, several sets of
input and their corresponding output vectors
are considered. The training phase is used to
XN N M L YN
determine the weights between the input,
hidden and output layers.
The neurons used in the study use the
sigmoid activation function defined by the Fig. 2. Basic BPN Architecture
following equation:
1. Read the training set and randomly
⎡Neuron⎤ 1.0 initialize the weights. Set iteration index n=1.
⎢ ⎥= -αv
(3) 2. Set training set index p=1.
⎣output ⎦ 1.0 + e
3. Propagate Xp through the network.
where α is the abruptness of the sigmoid 4. Determine the error vector of the pth
function and the v is the total input to the training set
neuron. Ep = Op - Yp where OP is the vector of
Let the vector X represent an input to the expected output.
input layer as shown in the Fig. 2. The net 5. Correct the weights using Newton's
input at the hidden layers is computed by the steepest descent technique.
matrix equation as below: 6. If p < number training sets P, set p =
VH = [WH] X (4) p+1 and go to step 3.
P

∑E
2
where WHji denotes the weight between ith 7. If p
> tolerance ε, increment
input layer node and jth hidden layer node. p =1

The output of the hidden layer nodes are the iteration index n and go to step 2.
given by The above method works well and has been
YH = Φ (VH) (5) well documented. The method requires that,
where Φ is the appropriate activation function. the input and output to be from a continuous
In a similar manner, the total input at the domain. Further, it also requires that the
output layer is given by the following input and output set of vectors are non-
equation: contradictory for a successful training and
VO = [WO] V H (6) operational function.
The output of the output layer node is given The RDS under study consists of 33 buses.
by The substation transformer is connected to
Y = Φ (VO) (7) bus 1 and there is no direct loading at bus1.
The voltage at bus 1 is known and is specified
The steps for well-established training as 1.0 per unit. The resistance and reactance
algorithm based upon Newton's steepest of lines between any two buses and the base
descent technique is given below: load condition is mentioned in table 1.

-3-
 Yanbu Journal of Engineering and Science Vol. 1 (2010)

4. IMPLEMENTATION OF BPN TO load pattern for the particular order of

DETERMINE HARMONIC LOAD harmonics.
FLOW SOLUTION

The input vector for the BPN is the real and V1,1 V2,1 V60,1
reactive power loads for different harmonics
V1,2 V2,2 V60,2
at various buses of the power system. The
V1,3 V2,3 V60,3
resistance of the different lines remains the
same for different harmonics where as the . . .

reactance changes according to the order of . . .

harmonics. Load flow solution for different . . .
load patterns is obtained using ladder iterative
V1,32 V2,32 V60,32
technique with the relevant impedance
Train Targe = δ1,1 δ2,1 δ60,1
component for the third, fifth and seventh
order harmonics. δ1,2 δ2,2 δ60,2

Sixty sets of loads were created by the δ1,3 δ2,3 δ60,3

following scheme: . . .
(a) Varying both the real and reactive . . .
power loads simultaneously at all the load . . .
buses of the radial system.
δ1,32 δ2,32 δ60,32
(b) Varying both the real and reactive
PL1 PL2 PL60
power loads simultaneously at a single load
bus of the radial system. (9)
(c) Varying only the real power load at a
single load bus of the radial system. Where Vi,j and δi,j represents the voltage and
(d) Varying only the reactive power load corresponding angle solution at the jth bus of
at a single load bus of the radial system. the ith load pattern for the particular order of
Equation (8) and (9) represents the train harmonics. PLi represents the total loss for the
input and train target matrix for a particular ith load pattern for the particular order of
order of harmonics. harmonics calculated from the ladder iterative
technique.
P1,1 P2,1 P60,1
For the multi layer feed forward ANN, tan-
P1,2 P2,2 P60,2
sigmoid transfer function (TANSIG) is used
P1,3 P2,3 P60,3
as activation function. For the considered 33
. . .
. . .
bus system, 64 input layer nodes ( 32+32,
P1,32 P2,32 P60,32 for real and reactive powers at each bus, there
Train Input = Q1,1 Q2,1 Q60,1 is no direct load connected at bus1 ) and 65
Q1,2 Q2,2 Q60,2 output layer nodes are used. ( 32+32+1, for
Q1,3 Q2,3 Q60,3 voltage magnitude and angle at each bus. The
. . . voltage magnitude and angle is specified at the
. . . substation, the last node represents the power
Q1,32 Q2,32 Q60,32
loss for the particular load condition).
(8) After successful training of the ANN it should
Where Pi,j and Qi,j represents the real and able to produce the bus voltage magnitude
reactive power demands at the jth bus of the ith with angle and the total power loss for any of
-4-
 Yan
nbu Journal of En
ngineering and Scieence Vol. 1 (20100)

the un
ntrained in
nput load pattern with
w
minimu
um time and
d maximum accuracy.

5. RESULTS
S AND DISC
CUSSION

A 33 bu us radial disttribution sysstem Fig. 3 was

tested using
u the proposed method.
m T
The
power flow
f equatioons were soolved using the
ladder iterative techhnique explaained in secttion
2.In ordder to achievve a broad reepresentation of
the pow wer system in the Bacck Propagattion
Networrk, approxim mately sixtyy input-outtput
vector pairs
p were generated for f each of the
harmon nics for the considered 33-bus systtem.
The BPN B was trained in i MATLA AB®
environment and the t trained result for th hird
harmon nics is shown n in Fig. 4. Thereafter, the
BPN is ready for use. The reesults from the
conventtional harm monic load d flow soluttion
and froom the traained ANN N for different
harmon nics are showwn in Tablee 2.The metthod
seems to work welll and is fou und to be very
v
efficientt and fast. The
T executioon time to reeach
the volttage solutionn from the trained
t ANN N is
approxim mately one third of the execution time
t Fig.3. 33-bus radial
r distributtion system
of the conventional
c l method. The
T bus voltaages 0
P
Performance is 4.9775e
e-005, Goal is 5e-005
10
and thee power losss from BP PN for the test
inputs for the diffferent orderr of harmonics -1
10

are com mpared with h ladder iterrative techniique

Training-Blue Goal-Black

-2
solutionn and is listed in table 2.. 10

For the minor chan nges in the network from

fr
-3
10
the far end of the source, willl not affect the
results very much. However if the sysstem -4
10
topologgy changed fromf the sennding end side,
s
the prroposed ap pproach will w not work
w -5
10
0 20 40 60 80 100 120 140 160
satisfacttorily, since the consid dered system
m is 162 Epoochs

radial. Effectiveness
E s of the prooposed metthod Fig.4.
F BPN Trraining Resultss for third harm
monics
for thee system toopology ch hanges can be
considerred for the future
f work.. 6. CONCLU
USIONS
As longg as if thee ANN is trrained with the This paper preesents a well defined approach
a
sufficiennt data ( it may be reaal or simulaated) to determine
d th
he harmonic load flow solution
the outccome of the ANN will be the expeccted of a radial diistribution system for various
outcom me. ord
der of harmoonics. Since collecting data
d from

-5-
 Yanbu Journal of Engineering and Science Vol. 1 (2010)

the real system with harmonic sources for a Electronics Express, Vol. 3, No. 14, pp.353-360,
large system is a difficult task, a 33 bus radial 2006
system is considered for analysis. Several load [8] W. F. Tinny and C. E. Hart, Power flow solution
of the Newton Method, IEEE Transactions PAS,
sets were considered and their solution was
Vol. PAS-86, No.11, Nov. 1967.
assessed using the conventional method of
[9] B. Stott and O. Alsac , “Fast Decoupled Load
ladder network technique. Then using these Flow” IEEE Transactions on Power Apparatus
sets of input and output vector pairs, the Back and Systems, Vol.PAS-93,pp.859-869, May/June
Propagation Network is trained. Thereafter, 1974.
the BPN is ready for use wherein, given a [10] B. Stott, ‘ Review of Load Flow Calculation
harmonic load, it gives out the voltage Methods’, Proceedings of the IEEE, Vol. 62, No.
7, July 1984
solution with minimum time and maximum
accuracy.

REFERENCES

[1] Kersting, W. H., and Mendive, D. L., 1976, "An

application of ladder network theory to the
solution of three phase radial load flow problem",
IEEE PES winter meeting, New Work, January.
[2] Kersting, W. H., 1984, "A method to design and
operation of distribution system", IEEE
Transactions on PAS-103, pp. 1945-1952.
[3] Stevens, R. A., Rizy, D. T. and Puruker, S. L.,
1986, "Performance of conventional power flow
routines for real time distribution automation
applications", Proceedings of 18Th southeastern
symposium on systems theory, (IEEE computer
society), pp. 196-200.
[4] Ciprian Bud, Bogdan Tomoiaga, Mercia
Chindris.”The Load Flow Calculation in
Harmonic Polluted Radial Electric Networks
with Distributed Generation” 9th International
conference on Electrical power quality and
Utilisation, Barcelona, 9-11 Oct 2007.
[5] Whei-Min Lin, Tung-Sheng Zhan, Ming-Tong
Tsay,” Multiple-frequency three-phase load flow
for harmonic analysis” IEEE transactions on
Power Systems, Vol 19 Issue 2, PP 897-904, May
2004.
[6] A. A. Romero, H .C. Zini, G. Ratta and R. Dib,”
A fuzzy number based methodology for
harmonic load-flow calculation, considering
uncertainties” Lat. Am. appl.
res. vol.38 no.3 Bahía Blanca July/Sept. 2008
[7] A.Arunagiri,B.Venkatesh, K.Ramasamy,
“Artificial neural network approach-an
application to radial loadflow algorithm”, IEICE

-6-
 Yanbu Journal of Engineering and Science Vol. 1 (2010)

TABLE 1: SYSTEM UNDER STUDY

Fundamental

From Bus

To Bus
S. No.

R (Ω)

Q (kvar)

V (p.u.)
P (kW)
X (Ω)
1 1 2 0.0922 0.047 100 60 1
2 2 3 0.493 0.2511 90 40 0.997
3 3 4 0.366 0.1864 120 80 0.9829
4 4 5 0.3811 0.1941 60 30 0.9754
5 5 6 0.819 0.707 60 20 0.9679
6 6 7 0.1872 0.6188 200 100 0.9495
7 7 8 1.7114 1.2351 200 100 0.946
8 8 9 1.03 0.74 60 20 0.9323
9 9 10 1.044 0.74 60 20 0.926
10 10 11 0.1966 0.065 45 30 0.9201
11 11 12 0.3744 0.1238 60 35 0.9192
12 12 13 1.468 1.155 60 35 0.9177
13 13 14 0.5416 0.7129 120 80 0.9115
14 14 15 0.591 0.526 60 10 0.9092
15 15 16 0.7463 0.545 60 20 0.9078
16 16 17 1.289 1.721 60 20 0.9064
17 17 18 0.732 0.574 90 40 0.9043
18 2 19 0.164 0.1565 90 40 0.9037
19 19 20 1.5042 1.3554 90 40 0.9965
20 20 21 0.4095 0.4784 90 40 0.9929
21 21 22 0.7089 0.9373 90 40 0.9922
22 3 23 0.4512 0.3083 90 50 0.9916
23 23 24 0.898 0.7091 420 200 0.9793
24 24 25 0.896 0.7011 420 200 0.9726
25 6 26 0.203 0.1034 60 25 0.9693
26 26 27 0.2842 0.1447 60 25 0.9475
27 27 28 1.059 0.9337 60 20 0.945
28 28 29 0.8042 0.7006 120 70 0.9335
29 29 30 0.5075 0.2585 200 600 0.9253
30 30 31 0.9744 0.963 150 70 0.9217
31 31 32 0.3105 0.3619 210 100 0.9176
32 32 33 0.341 0.5302 60 40 0.9167
Losses: 210.9983 kW

-7-
 Yanbu Journal of Engineering and Science Vol. 1 (2010)

TABLE 2A
3rd Harmonic, load = 1% of total load, voltage magnitude: 0.1pu
System Parameters Load
Real Reactive Voltage Voltage
Line Sending Receiving Resistance Reactance Percentage
Power Power Solution in pu Solution from
number Bus Bus Ω Ω Accuracy
KW KVAr (conventional) BPN in pu
1 1 2 0.0922 0.141 1 0.6 100
0.1000 0.1000
2 2 3 0.493 0.7533 0.9 0.4 0.0995 0.0997 99.98

3 3 4 0.366 0.5592 1.2 0.8 0.0972 0.0975 99.97

4 4 5 0.3811 0.5823 0.6 0.3 0.0960 0.0958 99.98

5 5 6 0.819 2.121 0.6 0.2 0.0947 0.0950 99.97

6 6 7 0.1872 1.8564 2 1 0.0911 0.0914 99.97

7 7 8 1.7114 3.7053 2 1 0.0901 0.0901 100

8 8 9 1.03 2.22 0.6 0.2 0.0877 0.0874 99.97

9 9 10 1.044 2.22 0.6 0.2 0.0866 0.0870 99.96

10 10 11 0.1966 0.195 0.45 0.3 0.0856 0.0850 99.94

11 11 12 0.3744 0.3714 0.6 0.35 0.0855 0.0849 99.94

12 12 13 1.468 3.465 0.6 0.35 0.0852 0.0852 100

13 13 14 0.5416 2.1387 1.2 0.8 0.0841 0.0837 99.96

14 14 15 0.591 1.578 0.6 0.1 0.0836 0.0832 99.96

15 15 16 0.7463 1.635 0.6 0.2 0.0834 0.0830 99.96

16 16 17 1.289 5.163 0.6 0.2 0.0831 0.0835 99.96

17 17 18 0.732 1.722 0.9 0.4 0.0827 0.0837 99.9

18 2 19 0.164 0.4695 0.9 0.4 0.0826 0.0822 99.96

19 19 20 1.5042 4.0662 0.9 0.4 0.0994 0.0994 100

20 20 21 0.4095 1.4352 0.9 0.4 0.0989 0.0994 99.95

21 21 22 0.7089 2.8119 0.9 0.4 0.0988 0.0972 99.84

22 3 23 0.4512 0.9249 0.9 0.5 0.0986 0.0976 99.9

23 23 24 0.898 2.1273 4.2 2 0.0967 0.0980 99.87

24 24 25 0.896 2.1033 4.2 2 0.0956 0.0956 100

25 6 26 0.203 0.3102 0.6 0.25 0.0951 0.0954 99.97

26 26 27 0.2842 0.4341 0.6 0.25 0.0907 0.0927 99.8

27 27 28 1.059 2.8011 0.6 0.2 0.0902 0.0912 99.9

28 28 29 0.8042 2.1018 1.2 0.7 0.0877 0.0707 98.3

29 29 30 0.5075 0.7755 2 6 0.0859 0.0879 99.8

30 30 31 0.9744 2.889 1.5 0.7 0.0852 0.0832 99.8

31 31 32 0.3105 1.0857 2.1 1 0.0844 0.0834 99.9

32 32 33 0.341 1.5906 0.6 0.4 0.0842 0.0812 99.7

Loss
Total Loss 2.3947KW from 2.52KW
88.47%
BPN

-8-
 Yanbu Journal of Engineering and Science Vol. 1 (2010)

TABLE 2B
5th Harmonic, load = 0.5% of total load, voltage magnitude: 0.075pu
System Parameters Load
Voltage Voltage
Real Reactive
Line Sending Receiving Resistance Reactance Solution in pu Solution Percentage
Power Power
number Bus Bus Ω Ω (conventional) from BPN in Accuracy
KW KVAr
pu
1 1 2 0.0922 0.235 0.5 0.3 0.0750 0.0750 100

2 2 3 0.493 1.2555 0.45 0.2 0.0745 0.0723 99.78

3 3 4 0.366 0.932 0.6 0.4 0.0723 0.0743 99.8

4 4 5 0.3811 0.9705 0.3 0.15 0.0711 0.0722 99.89

5 5 6 0.819 3.535 0.3 0.1 0.0698 0.0596 98.98

6 6 7 0.1872 3.094 1 0.5 0.0658 0.0648 99.9

7 7 8 1.7114 6.1755 1 0.5 0.0646 0.0743 99.03

8 8 9 1.03 3.7 0.3 0.1 0.0621 0.0635 99.86

9 9 10 1.044 3.7 0.3 0.1 0.0609 0.0629 99.8

10 10 11 0.1966 0.325 0.225 0.15 0.0598 0.0548 99.5

11 11 12 0.3744 0.619 0.3 0.175 0.0597 0.0596 99.99

12 12 13 1.468 5.775 0.3 0.175 0.0595 0.0595 100

13 13 14 0.5416 3.5645 0.6 0.4 0.0583 0.0575 99.92

14 14 15 0.591 2.63 0.3 0.05 0.0577 0.0583 99.94

15 15 16 0.7463 2.725 0.3 0.1 0.0574 0.0563 99.89

16 16 17 1.289 8.605 0.3 0.1 0.0572 0.0532 99.6

17 17 18 0.732 2.87 0.45 0.2 0.0567 0.0565 99.98

18 2 19 0.164 0.7825 0.45 0.2 0.0565 0.0546 99.81

19 19 20 1.5042 6.777 0.45 0.2 0.0745 0.0775 99.7

20 20 21 0.4095 2.392 0.45 0.2 0.0739 0.0735 99.96

21 21 22 0.7089 4.6865 0.45 0.2 0.0738 0.0733 99.95

22 3 23 0.4512 1.5415 0.45 0.25 0.0737 0.0735 99.98

23 23 24 0.898 3.5455 2.1 1 0.0718 0.0735 99.83

24 24 25 0.896 3.5055 2.1 1 0.0708 0.0608 99

25 6 26 0.203 0.517 0.3 0.125 0.0703 0.0623 99.2

26 26 27 0.2842 0.7235 0.3 0.125 0.0654 0.0654 100

27 27 28 1.059 4.6685 0.3 0.1 0.0649 0.0669 99.8

28 28 29 0.8042 3.503 0.6 0.35 0.0621 0.0631 99.9

29 29 30 0.5075 1.2925 1 3 0.0601 0.0631 99.7

30 30 31 0.9744 4.815 0.75 0.35 0.0593 0.0583 99.9

31 31 32 0.3105 1.8095 1.05 0.5 0.0585 0.0485 99

32 32 33 0.341 2.651 0.3 0.2 0.0582 0.0602 99.8

Loss from
Total Loss 1.189KW 1.03KW
BPN 84.1

-9-
 Yanbu Journal of Engineering and Science Vol. 1 (2010)

TABLE 2C
7th Harmonic, load = 0.25% of total load, voltage magnitude: 0.0.05pu

System Parameters Load

Voltage Percentage
Real Reactive Voltage Solution from Accuracy
Line Sending Receiving Resistance Reactance
Power Power Solution in pu BPN in pu
number Bus Bus Ω Ω
KW KVAr (conventional)

1 1 2 0.0922 0.329 0.25 0.15 0.0500 0.0490 99.9

2 2 3 0.493 1.7577 0.225 0.1 0.0494 0.0483 99.89
3 3 4 0.366 1.3048 0.3 0.2 0.0467 0.0456 99.89
4 4 5 0.3811 1.3587 0.15 0.075 0.0450 0.0450 100
5 5 6 0.819 4.949 0.15 0.05 0.0433 0.0422 99.89
6 6 7 0.1872 4.3316 0.5 0.25 0.0376 0.0386 99.9
7 7 8 1.7114 8.6457 0.5 0.25 0.0356 0.0346 99.9
8 8 9 1.03 5.18 0.15 0.05 0.0320 0.0323 99.97
9 9 10 1.044 5.18 0.15 0.05 0.0302 0.0313 99.89
10 10 11 0.1966 0.455 0.1125 0.075 0.0286 0.0276 99.9
11 11 12 0.3744 0.8666 0.15 0.0875 0.0285 0.0265 99.8
12 12 13 1.468 8.085 0.15 0.0875 0.0282 0.0262 99.8
13 13 14 0.5416 4.9903 0.3 0.2 0.0263 0.0250 99.87
14 14 15 0.591 3.682 0.15 0.025 0.0254 0.0244 99.9
15 15 16 0.7463 3.815 0.15 0.05 0.0249 0.0238 99.89
16 16 17 1.289 12.047 0.15 0.05 0.0245 0.0255 99.9
17 17 18 0.732 4.018 0.225 0.1 0.0236 0.0236 100
18 2 19 0.164 1.0955 0.225 0.1 0.0234 0.0238 99.96
19 19 20 1.5042 9.4878 0.225 0.1 0.0494 0.0484 99.9
20 20 21 0.4095 3.3488 0.225 0.1 0.0488 0.0468 99.8
21 21 22 0.7089 6.5611 0.225 0.1 0.0487 0.0477 99.9
22 3 23 0.4512 2.1581 0.225 0.125 0.0486 0.0486 100
23 23 24 0.898 4.9637 1.05 0.5 0.0461 0.0451 99.9
24 24 25 0.896 4.9077 1.05 0.5 0.0451 0.0442 99.91
25 6 26 0.203 0.7238 0.15 0.0625 0.0445 0.0345 99
26 26 27 0.2842 1.0129 0.15 0.0625 0.0371 0.0377 99.94
27 27 28 1.059 6.5359 0.15 0.05 0.0365 0.0368 99.97
28 28 29 0.8042 4.9042 0.3 0.175 0.0325 0.0321 99.96
29 29 30 0.5075 1.8095 0.5 1.5 0.0297 0.0295 99.98
30 30 31 0.9744 6.741 0.375 0.175 0.0287 0.0289 99.98
31 31 32 0.3105 2.5333 0.525 0.25 0.0273 0.0270 99.97
32 32 33 0.341 3.7114 0.15 0.1 0.0270 0.0258 99.88
Loss
Total Loss 0.9358KW from 0.84KW
90.42
BPN

- 10 -
 ‫)‪Yanbu Journal of Engineering and Science Vol. 1 (2010‬‬

‫ﺗﻄﺒﻴﻖ أﺳﻠﻮب ﺷﺒﻜﺎت اﻟﻌﺼﺐ اﻟﺼﻨﺎﻋﻲ ﻋﻠﻰ ﺳﺮﻳﺎن اﻻﺣﻤﺎل ﻓﻲ اﻻﻧﻈﻤﺔ اﻟﻤﺤﻮرﻳﺔ ﻓﻲ ﻇﻞ ﺗﻮاﻓﻘﻴﺎت ﻏﻴﺮ ﻣﺮﻏﻮب ﻓﻴﻬﺎ‬

‫‪١‬‬
‫أروﻧﺎﺟﻴﺮي‪ ،١‬ﺳﻮرﻳﺶ‪ ،٢‬ﻓﻨﻜﺎﺗﻴﺶ‪ ،٣‬راﻣﻴﺶ آﻮﻣﺎر‪ ١‬و ﻣﺴﺘﺠﺎب أﺣﻤﺪ ﺧﺎن‬
‫‪ ١‬آﻠﻴﺔ ﻳﻨﺒﻊ اﻟﺼﻨﺎﻋﻴﺔ‪ ،‬اﻟﻤﻤﻠﻜﺔ اﻟﻌﺮﺑﻴﺔ اﻟﺴﻌﻮدﻳﺔ‬
‫‪ ٢‬ﺟﺎﻣﻌﺔ ﺑﺎراث‪ ،‬اﻟﻬﻨﺪ‬
‫‪ ٣‬ﺟﺎﻣﻌﺔ رﻳﺴﻮن‪ ،‬آﻨﺪا‬

‫اﻟﻤﻠﺨﺺ‪:‬‬

‫ﺗﺘﻄﻠﺐ اﻻﻧﻈﻤﺔ اﻟﻤﺤﻮرﻳﺔ ﻟﺘﻮزﻳﻊ اﻟﻄﺎﻗﺔ ﻃﺮق ﺧﺎﺻﺔ ﻟﺴﺮﻳﺎن اﻻﺣﻤﺎل ﻟﻜﻲ ﻳﺤﻞ ﻣﻌﺎدﻻت ﺳﺮﻳﺎن اﻟﻘﻮي‪ .‬ﻧﺒﺰﻳﺎدة‬

‫اﺳﺘﻌﻤﺎل اﻻﺟﻬﺰة اﻻﻟﻜﺘﺮوﻧﻴﺔ وﺗﺄﺛﻴﺮ اﻟﺘﺸﺒﻊ اﻟﻤﻐﻨﺎﻃﻴﺴﻲ أدي اﻟﻰ ﻇﻬﻮر ﺗﻮاﻓﻘﻴﺎت ﻏﻴﺮ ﻣﺮﻏﻮب ﻓﻴﻬﺎ ﻓﻲ اﻧﻈﻤﺔ‬

‫اﻟﺘﻮزﻳﻊ اﻟﻤﺤﻮرﻳﺔ‪.‬‬

‫ﺗﻘﺪم هﺬﻩ اﻟﻤﻘﺎﻟﺔ اﻟﺒﺤﺜﻴﺔ ﻧﻈﺎم ﻣﺘﻌﺪد اﻟﻄﺒﻘﺎت ﻟﻠﺘﻐﺬﻳﺔ اﻻﻣﺎﻣﻴﺔ ﻣﻊ وﺟﻮد ﺑﺮﻧﺎﻣﺞ ﻟﺴﺮﻳﺎن اﻻﺧﻄﺎء ﻓﻲ اﻻﺗﺠﺎﻩ‬

‫اﻟﻤﻌﺎآﺲ وذﻟﻚ ﻟﺤﺴﺎب اﻟﺠﻬﺪ وﻓﻘﺪ اﻟﻘﻮة ﻻآﺜﺮ ﻣﻦ ﻣﺮآﺒﺔ ﺗﻮاﻓﻘﻴﺔ ﻏﻴﺮ ﻣﺮﻏﻮب ﻓﻴﻬﺎ‪.‬‬

‫ﺗﻢ اﺧﺘﺒﺎر اﻟﻄﺮﻗﺔ اﻟﻤﻘﺘﺮﺣﺔ ﻋﻠﻰ ‪ ٣٣‬ﻣﺴﺎر ﻟﻠﺘﻮزﻳﻊ اﻟﻤﺤﻮري وﺗﻢ ﺗﺒﻮﻳﺐ اﻟﻨﺘﺎﺋﺞ ﻃﺒﻘﺎ ﻟﻜﻞ ﺗﺮدد ﻣﻦ اﻟﺘﻮاﻓﻘﻴﺎت‬

‫ﻏﻴﺮ اﻟﻤﺮﻏﻮب ﻓﻴﻬﺎ‪ .‬وﻗﺪ اﺛﺒﺘﺖ هﺬﻩ اﻻﺧﺘﺒﺎرات ﻟﻠﻄﺮﻳﻘﺔ اﻟﻤﻘﺘﺮﺣﺔ ﺟﺪوي هﺬﻩ اﻟﻄﺮﻳﻘﺔ ﻟﺘﻘﻴﻴﻢ ﺳﺮﻳﺎن اﻻﺣﻤﺎل ﻓﻲ‬

‫اﻻﻧﻈﻤﺔ اﻟﺤﻮرﻳﺔ ﻓﻲ وﺟﻮد ﺗﻮاﻓﻘﻴﺎت ﻏﻴﺮ ﻣﺮﻏﻮب ﻓﻴﻬﺎ‪.‬‬

‫‪- 11 -‬‬