Volume 9, Issue 11, November – 2024 International Journal of Innovative Science and Research Technology

ISSN No:-2456-2165

Calculation of Stable Controller Values for

Single Area Isolated Power System using
Boundary Locus Method
1
L. Advila,
1
Assistant Professor, Dept. of EEE, NSRIT, Visakhapatnam, AP
NSRIT-Nadimpalli Satyanarayana Raju Institute of Technology.

Abstract:- The power system's operating point fluctuates is provided. Plotting the stability boundary locus in the (kp;
constantly due to its extremely nonlinear nature. As a ki) plane and then determining the stabilising values of a PI
result, both actual and reactive power are impacted by the controller's parameters form the basis of the suggested
extremely low system performance. Real power shifts approach. By solving a series of inequalities without the use
mostly impact the Changes in voltage magnitude are the of linear programming, the method does not need sweeping
primary determinant of changes in reactive power and over the parameters or the Pade approximation. As a result, it
system frequency. Reactive and real capabilities can provides a number of significant benefits above the current
therefore be managed independently. The Automatic findings in this area. In addition to stabilisation, stabilising PI
Voltage Regulator (AVR) regulates the voltage magnitude controllers that meet predetermined gain and phase margins
and, therefore, the reactive power, whereas the Load are calculated using this technique. PID controllers for
Frequency Control (LFC) controls the actual power. The control systems with and without time delay are also designed
regulating of generator power output is known as load using the suggested methodology. The (kp; ki), (kp; kd), and
frequency control, or LFC, in an interconnected system. (ki; kd) planes yield the limiting values of a PID controller
Generally speaking, fixed gain controllers are made for that stabilise a particular system with time delay.
nominal operating settings and don't offer the optimal
control performance under a variety of operating II. BOUNDARY LOCUS METHOD
circumstances. Therefore, it is preferable to monitor
operating circumstances and compute the control using A. Design of Pi Controller
updated parameters in order to maintain system Examine Figure 1's single-input single-output (SISO)
performance close to optimal. Using the "Boundary Locus control system.
Method," a novel approach to identifying stabilising PID
controllers for the LFC control system loop has been put
forth in this study.

I. INTRODUCTION

Unusual circumstances like generating failures and load

variations lead to the system frequency degrading from the
intended level. With an appropriate LFC design, the generator
loads must be controlled based on the ideal frequency value Fig 1: A SISO Control System
in order to guarantee the power supply's quality. Several
𝑁(𝑆)
strategies have been used in the past to account for parametric Where 𝐺(𝑆) 𝐷(𝑆) (1)
uncertainty and nonlinear limitations while maintaining
system performance close to its ideal operating state [1].
Is the plant that has to be managed, and C(s) is the form's PI
controller.
A controller is a device that has the ability to rectify
deviations by comparing regulated values with intended 𝐾𝑖 (𝐾𝑝 𝑆+𝐾𝑖)
values. Achieving system stability is the primary goal of 𝐺(𝑆) = 𝐾𝑝 + 𝑆
= 𝑆
(2)
every controller. By reducing steady state errors, controllers
increase steady state accuracy. Stability increases together The challenge is to calculate the PI controller's settings
with the precision of the steady state. Additionally, they aid using Eq. (2), which stabilises the system in Fig. 1.
in lowering the offsets generated inside the system. With Substituting s = jω after breaking down the numerator and
these controllers, the system's maximum overrun may be denominator polynomials of Eq. (3) into their even and odd
managed. Additionally, they aid in lowering the system's components yields
noise signals. With these controls, the over-damped system's
slow reaction may be accelerated. This study uses a novel 𝑁𝑒 (−𝜔 2 )+𝑗𝜔𝑁𝑜(−𝜔 2 )
𝐺(𝑗𝜔) = 𝐷𝑒 (−𝜔 2 )+𝑗𝜔𝐷𝑜(−𝜔 2 )
(3)
approach to calculate all stabilising PI controllers [2]. and [3]

IJISRT24NOV841 www.ijisrt.com 1741

Volume 9, Issue 11, November – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165

The system's characteristic polynomial ∆(s) is shown in Hurwitz stability is 1. The characteristic equation from
Fig. Fig. 1 may be expressed as ∆(s)=1+G(s).C(s)
The system's closed loop characteristic polynomial may be
expressed as

Δ(S) = [KiNe(−ω 2 )– Kpω 2N0(−ω 2 )– ω 2D0(−ω 2 )] + j[KpωNe(−ω 2 ) + KiωN0(−ω 2 ) + ωD0(−ω 2 )] = 0 (4)

The real and imaginary components of ∆(s) are thus equal to zero, yielding

[Kpω Ne(−ω 2 ) + KiωN0(−ω 2 ) + ωD0(−ω 2 )] = 0 (5)

And

KpωNe(−ω 2 ) + KiωN0(−ω 2 ) = − ωD0(−ω 2 ) (6)

Let
Q(ω) = -ω 2N0(-ω 2 ), R(ω) = Ne(-ω 2 ) B. Design of Pid Controllers:
S(ω) = ωNe(-ω 2 ), U(ω) = ωN0(-ω 2 ) Assume that Fig. 1's C(s) is a PID controller of type
X(ω) = ωD0(-ω 2 ), Y(ω) = -ωDe(-ω 2 ) (7)
𝐾𝑖
𝐶(𝑆) = 𝐾𝑝 𝐾𝑑 𝑆 (11)
𝑆
Equations (5) and (6) may therefore be expressed as
The stability boundary locus in the (Kp, Ki) plane can
𝐾𝑝 𝑄(𝜔) + 𝐾𝑖 𝑅(𝜔) = 𝑋(𝜔)
be found by following the steps in Section 1. be readily
produced in the (kp, kd)-plane for a fixed value of ki, or it can
𝐾𝑝 𝑆(𝜔) + 𝐾𝑖 𝑈(𝜔) = 𝑌(𝜔) (8) be acquired for a fixed value of Kd. For a given value of kp,
however, the stability boundary locus in the (ki, kd)-plane
From these equations cannot be obtained because will, in this instance, equal zero.
While the stability area in the (ki,kd) plane for a fixed value
𝑋(𝜔)𝑈(𝜔)−𝑌(𝜔)𝑅(𝜔)
𝐾𝑝 = 𝑄(𝜔)𝑈(𝜔)−𝑅(𝜔)𝑆(𝜔) (9) of Kp is a convex polygon, it is not a convex polygon and
might not even be a convex set in the (kp, li)-plane for a fixed
value of kd or in the (kp,kd)-plane for a fixed value of ki.
And
However, by employing the stability region found in the (kp,
𝑌(𝜔)𝑄(𝜔)−𝐾(𝜔)𝑆(𝜔) ki) plane and (kp, kd)-plane as follows, the stability region in
𝐾𝑖 = 𝑄(𝜔)𝑈(𝜔)−𝑅(𝜔)𝑆(𝜔) (10) the (ki, kd)-plane may be determined for a fixed value of kp.

The stability boundary locus, l(kp,ki,w), in the (k,.ki)- 𝜔3 𝑁0𝐷0 + 𝜔𝑁𝑒 𝐷0

𝐾𝑝 =
plane may be found by concurrently solving these two −𝜔 2 𝑁02 − 𝑁𝑒2
equations. The parameter plane ((kp,ki)-plane) is separated
into stable and unstable areas by the stability boundary locus. 𝜔2 𝑁0 𝐷𝑒 − 𝜔2 𝑁𝑒 𝐷0
𝐾𝑖 =
−𝜔 2 𝑁02 − 𝑁𝑒2
The stable zone containing the stabilising kp and ki
parameter values may be identified by selecting a test point
within each region.

Substituting Ki in terms of Kd

𝜔2 𝑁0 𝐷𝑒 − 𝑘𝑑 𝜔4 𝑁02 − 𝜔2 𝑁𝑒 𝐷0 − 𝐾𝑑 𝜔2 𝑁𝑒2
𝐾𝑖 =
−𝜔 2 𝑁02 − 𝑁𝑒2

(𝜔2 𝑁0 𝐷0 − 𝜔2 𝑁𝑒 𝐷0 ) − 𝐾𝑑 𝜔2(−𝜔2 𝑁02 − 𝑁𝑒2 )

𝐾𝑖 =
−𝜔 2 𝑁02 − 𝑁𝑒2

−(𝜔2 𝑁0 𝐷𝑒 − 𝜔2 𝑁𝑒 𝐷0) − 𝐾𝑖 (−𝜔2 𝑁02 − 𝑁𝑒2 )

𝐾𝑑 =
𝜔 2 (−𝜔 2 𝑁02 − 𝑁𝑒2 )

Thus, the stability areas in the (kp, ki) and (kp, kd) planes may be found using these equations.

IJISRT24NOV841 www.ijisrt.com 1742

Volume 9, Issue 11, November – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165

III. BASIC GENERATION CONTROL LOOPS moment of inertia-time constant, the power frequency control
is slow acting and the excitation control is rapid acting. As a
Each generator in an integrated power system has its result, the load frequency and excitation voltage control are
own LFC and AVR control loop. Fig. 2 shows the schematic examined separately, and there is very little cross coupling
representation of the voltage and frequency control loop. The between the LFC loop and the AVR. Active and reactive
frequency and voltage magnitude are kept within the power demands in a power system are never constant; they
designated bounds by the controllers, which are configured fluctuate constantly in response to growing or dropping
for a certain operating state and handle slight variations in trends.
load demand. Variations in the rotor angle δ and,
consequently, the frequency f, are the primary determinants While reactive power is primarily dependent on changes
of slight variations in actual power. The magnitude of the in voltage magnitude and less sensitive to frequency
voltage, or the generator excitation, is the primary variations, changes in actual power have an impact on the
determinant of reactive power [4]. A brief variation in system frequency. System generation control's primary goal
generator speed results in a change in angle ⁤. For minor is to maintain the intended frequency and power exchanges
variations, load frequency and excitation voltage controls are across adjacent systems by balancing system generation
therefore non-interactive and amenable to separate modelling against load and losses. The Automatic Voltage Regulator
and analysis. Additionally, because the generating field's time (AVR) and the Load Frequency Controller (LFC) are a
constant is significantly less than the turbine's and generator's generation's two primary control loops.

Fig 2: Schematic Diagram of LFC and AVR of a Synchronous Generator

IV. LOAD FREQUENCY CONTROL It could not be cost-effective to send all of the data over
great distances. The control, which depends on load demand,
In the design and operation of electrical power systems, is a function of the states. Realising the ideal controller may
load frequency management is crucial. depend on an accurate load demand prediction.

Furthermore, an LFC system that controls generator By controlling the system frequency, LFC seeks to
loading based on frequency must be designed in order to preserve the system's actual power balance. There is a
guarantee the quality of the power supply. For the following frequency variation whenever the actual power demand
reasons, conventional controllers are frequently impractical varies. The turbine governor receives this frequency error
for implementation. All of the system's states influence the after it has been amplified, combined, and converted into a
ideal control. In reality, not every state could be accessible. It command signal. By altering the turbine's output, the
is necessary to assess the states that are unavailable or absent. governor works to bring the input and output back into
balance.

Megawatt frequency or power-frequency (P-f) control

are other names for this technique. Take into account the
following LFC Loop Parameters for the analysis:

IJISRT24NOV841 www.ijisrt.com 1743

Volume 9, Issue 11, November – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165

Table 1: LFC
S.No Block Gain Time Constant
1 Governor Kg=1 Tg=0.2 Sec
2 Turbine Kt=1 Tt=0.5 Sec
3 Generator Inertia Constant H=5 Sec
4 Governor Speed Regulation R=0.05 Per Unit
5 The load varies by 0.8 percent for a 1 percent change in frequency, i.e D=0.8

A. Design of Pi Controller
Single Area Load Frequency Control's analogous transfer function is provided by

△ 𝜴(𝑆) (1 + 𝑇𝑔 𝑠)(1 + 𝑇𝑡 𝑠)
𝐺(𝑆) = =
△ 𝑷𝑳 (𝑺) (2𝐻𝑠 + 𝐷)(1 + 𝑇 𝑆)(1 + 𝑇 𝑆) + 1
𝑔 𝑡 𝑅

Changing the parameter values shown in Table 1 now

(1 + 0.2𝑠)(1 + 0.5𝑠)
𝐺(𝑆) =
1
(2 ∗ 5𝑠 + 0.8)(1 + 0.2𝑆)(1 + 0.5𝑆) +
0.05

0.1𝑆 2 + 0.7𝑆 + 1
𝐺(𝑆) =
𝑆3 + 7.08𝑆 2 + 10.56𝑆 + 20.8

Substituting S = j ω;

(−0.1𝜔2 + 1) + 𝑗𝜔(0.7)
𝐺(𝑆) =
(−7.08𝜔 3 + 20.8) + 𝑗𝜔(−𝜔 2 + 10.56)

From above equation, we have

R(ω) = −0.1𝜔2 + 1, 𝑈(𝜔) = 0.7𝜔

Q(ω) = −0.7𝜔2 , 𝑆(𝜔) = −0.1𝜔3 + 𝜔

Y(ω) = 7.08𝜔3 − 20.8𝜔, 𝑋(𝜔) = −𝜔4 + 10.56𝜔2

0.008𝜔 5 −1.768𝜔3 +20.8𝜔 Next, determine the Kp and Ki planes' relative

𝐾𝑝 = (12)
−0.01𝜔 5 −0.29𝜔 3 − 𝜔 stabilisation and ρ =0.5. By repeating the previous step, we
determine Kp and Ki.
−0.1𝜔 7 −2.9𝜔 5 +4.396𝜔3
𝐾𝑖 = (13)
−0.01𝜔 5 −0.29𝜔3 −𝜔3
0.058𝜔 5 −0.117𝜔3 +38.465𝜔
𝐾𝑝 = −0.01𝜔 5−0.365𝜔 3−1.8906𝜔 (14)
Shifting all of the poles of a control system's
characteristic equation to a desired area in the complex −0.1𝜔 7 −3.65𝜔 5+2.9𝜔 3
plane—for instance, to a shifted half plane that ensures a 𝐾𝑖 = (15)
−0.01𝜔 5 −0.365𝜔3 −1.8906𝜔
given response settling time—is crucial for control system
analysis and design. From the equations 12, 13, 14 and 15, the stability
region is shown in Fig 3.
This section's goal is to identify all values of S=ρ, where
(ρ= constant).

Using s+ρ instead of s in quation 1 & 2.

IJISRT24NOV841 www.ijisrt.com 1744

Volume 9, Issue 11, November – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165

Fig 3: Stability Regions for ρ=0 and ρ=0.5

B. Design of Pid Controller For Lfc

Now Consider Kd. Step 1: Let us consider Ki,

𝑎1 𝜔7 + 𝑎2 𝜔5 + 𝑎3 𝜔3
𝐾𝑖 =
−0.01𝜔 5 −0.29𝜔 3 − 𝜔

Where 𝑎1 = −0.1 − 0.01𝐾𝑑 , 𝑎2 = −2.9 − 0.29𝐾𝑑 , 𝑎3 = 4 − 𝐾𝑑 ,

Stability region for 𝐾𝑑 = 0

−0.1𝜔7 −2.9𝜔5 + 4𝜔3

𝐾𝑖 =
−0.01𝜔 5 −0.29𝜔 3 − 𝜔

Stability region for 𝐾𝑑 = 1

−0.11𝜔7 −3.19𝜔5 + 4𝜔3

𝐾𝑖 =
−0.01𝜔 5 −0.29𝜔 3 − 𝜔

Now the plots are obtained for 𝐾𝑝 𝑎𝑛𝑑 𝐾𝑖

Fig 4: Stability Region in the (Kp, Ki) plane for Kd=0 and Kd=1

IJISRT24NOV841 www.ijisrt.com 1745

Volume 9, Issue 11, November – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165

 Step 2: 0.1𝜔7 + 2.9𝜔5 − 4𝜔3

𝐾𝑑 =
−0.1𝜔 7 − 2.9𝜔 5 − 𝜔 3
Obtain equation 𝐾𝑑 om terms of 𝐾𝑖
Stability region for 𝐾𝑖 = 1
0.1𝜔 7 +𝑏1 𝜔 5+𝑏2 𝜔 3+ 𝑏3 𝜔
𝐾𝑖 = −0.01𝜔 7 −0.29𝜔5 −𝜔 3
0.1𝜔7 + 2.89𝜔5 − 4.29𝜔3
𝐾𝑑 =
Where 𝑏1=2.9-0.01 𝐾𝑖 ,𝑏2= -4-0.29𝐾𝑖 ,𝑏3 = 𝐾𝑖 −0.1𝜔 7 − 2.9𝜔 5 − 𝜔 3

Now the plots are obtained for 𝐾𝑝 𝑎𝑛𝑑 𝐾𝑑

Stability regions for 𝐾𝑖 = 0

Fig 5: Stability Region in the (Kp, Ki) Plane for Ki=0 and Ki=1

From the above two graphs we obtain 8 points for Kp= 𝐾𝑑 = 0.015𝐾𝑖 c -5.71, 𝐾𝑑 =0
0 are (36.57, 0), (42.97, 1), (0, 0), (0, 1), (0, - 5.71), (1, -5.55),
(0, 0), (1, 0) 𝐾𝑑 = 0.016𝐾𝑖 c -5.71, 𝐾𝑑 =0

From these we obtain 4 straight lines Now, sketch these lines' patch shape. Plots for Kp, Ki,
and Kd are now available.

Fig 6(a): Stabilizing Kp, Ki, and Kd Region

IJISRT24NOV841 www.ijisrt.com 1746

Volume 9, Issue 11, November – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165

Fig 6(b): Stabilizing Kp, Ki, and Kd Region

Fig 7: Simulink LFC Block

Fig 8: Plot for Change in Frequency for a Step Load Change of 0.1 p.u

IJISRT24NOV841 www.ijisrt.com 1747

Volume 9, Issue 11, November – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165

Fig 9: Plot for Change in Frequency for a Step Load Change of 0.1 p.u

V. CONCLUSION [7]. A.R. Hasan, T.S. Martis , Design and Implementation

of a Fuzzy Based Automatic Voltage Regulator for a
The quality of the power supply is defined by the Synchronous Generator, IEEE Transactions on energy
stability of its frequency and voltage. In this context, the conversion, 1994.
stabilizing PID control parameters for the Load Frequency [8]. D.M. Vinod Kumar, Intelligent Controllers for
Control (LFC) system, which manages real power and Automatic Generation Control, IEEE Transactions on
frequency, are determined using the "Stability Boundary global connectivity in energy, computer,
Locus Method." Simulation results indicate that the proposed communication and control, 1988, pp557-574.
PID controller, based on this method, efficiently and rapidly [9]. Ho MT, Datta A, Bhattacharyya SP.A new approach
achieves optimal LFC parameters. The results further show to feedback stabi-lization. In: Proc of the 35th CDC,
that the PID controllers offer satisfactory stability by 1996. p. 46438.
balancing frequency overshoot and transient oscillations, [10]. Power System Analysis by Haadi Sadat
while maintaining zero steady-state error. Overall, the
simulations demonstrate the effectiveness of the control
response.

REFERENCES

[1]. J. Talaq, F. Al-Basri, Adaptive Fuzzy Gain Scheduling

for Load Frequency Control, IEEE Transactions in
power system,pp145- 150, 1999.
[2]. Tan, N., Computation of Stabilizing PI and PID
controllers for processes with time delay, ISA
Transactions, Vol. 44, No. 2, pp. 213- 223, 2005.
[3]. Tan, N., I. Kaya, C. Yeroglu and D. P. Atherton,
Computation of stabilizing PI and PID controllers
using the stability boundary locus, Energy Conversion
and Management, Vol. 47, No. 18-19, pp. 3045-3058,
2006.
[4]. A.Soundarrajan, Member ,IAENG, Dr.S.Sumathi,
C.Sundar, Particle Swarm Optimization Based LFC
and AVR of Autonomous Power Generat-ing System,
IAENG International journal of computer science ,
2010.
[5]. H.D. Mathur and S. Ghosh, A Comprehensive
Analysis of Intelligent Control for Load Frequency
Control, IEEE Power india conference, 2006.
[6]. Elyas Rakhshani, Kumars Rouzbehi and Sedigheh
Sadeh , A New Com-bined Model for Simulation of
Mutual Effects Between LFC and AVR Loops, IEEE
Transactions on power system, 2009.

IJISRT24NOV841 www.ijisrt.com 1748