International Journal of Power Electronics and Drive System (IJPEDS)

Vol. 16, No. 2, June 2025, pp. 1083~1093

ISSN: 2088-8694, DOI: 10.11591/ijpeds.v16.i2.pp1083-1093  1083

A novel adaptive control scheme for dynamic voltage restorer

Tummala Kranti Kiran1, Balakrishnan Rajagopal1, Yerramilli Butchi Raju2

1
Department of Electrical Engineering, Annamalai University, Chidambaram, India
2
Department of Electrical and Electronic Engineering, Sir C. R. Reddy College of Engineering, Eluru, India

Article Info ABSTRACT

Article history: This study introduces a novel approach to improve the performance of the
dynamic voltage restorer (DVR) using a control structure based on the
Received Apr 29, 2024 inverse hyperbolic sine function-based least mean square (IHSF-LMS). The
Revised Jan 18, 2025 proposed control structure efficiently extracts the fundamental source
Accepted Mar 1, 2025 voltage component with rapid convergence. The primary objective of this
control scheme is to enhance compensation capability and maintain load
voltage stability. Key features include superior steady-state performance and
Keywords: heightened robustness with reduced computational complexity. This
approach achieves a reduction in peak overshoot and settling time compared
Adaptive algorithm to the traditional LMS method. The validation of system performance is
Dynamic voltage restorer conducted through MATLAB/Simulink simulations, demonstrating
Inverse hyperbolic sine compliance with the IEEE-519-2014 standards for harmonic spectra.
function-based LMS
Power quality This is an open access article under the CC BY-SA license.
Voltage distortion

Corresponding Author:
Tummala Kranti Kiran
Department of Electrical Engineering, Annamalai University
Annamalainagar 608002, Chidambaram, Tamil Nadu, India
Email: [email protected]

1. INTRODUCTION
Power electronic devices have gained popularity in both consumer and industrial sectors due to
advantages such as energy conservation, cost-effectiveness, smaller size, and reduced maintenance
requirements [1], [2]. However, these devices introduce power quality concerns, including voltage sag, swell,
unbalancing, and distortions at the distribution end. These issues lead to undesirable consequences such as
increased losses and malfunctioning protective equipment. Power quality (PQ) problems are prevalent in
distribution networks and pose significant challenges [3]. Addressing these concerns, [4] provides a list of
custom power devices (CPDs) designed to enhance PQ in distribution networks. In a study by Das et al. [5], a
brief overview of deteriorating power quality is presented. Within the realm of power quality, a series-
connected device referred to as the dynamic voltage restorer (DVR) has been introduced to address voltage-
related issues [6].
The control strategy implemented in the DVR significantly influences operational performance
during various disturbances in grid voltages. Preview study [7], [8] proposed software PLL (SPLL) for DVR
control, focusing on estimating phase jumps during system faults to protect loads. The basic control scheme
is a synchronous reference frame (SRF), which is implemented for DVR for mitigation of voltage sag and
swell [9], [10]. However, its complexity arises from inherent abc/dq transformations and the presence of a
phase lock loop [11], potentially introducing errors if the SRF is not identified accurately. The dual-pq
theory-based DVR with embedded EZ source inverter topology is articulated in [12]. Various filtering
techniques, such as modified dual-second-order generalized integrator (MD-SOGI) [13], least mean square
(LMS) [14], and least mean fourth (LMF), among others, are available in the literature. The distinct
comparison between LMS, LMF, and combined LMS/F is made by the author in [15], indicating the
capability of the LMS/F-based control algorithm. However, it is noted that the LMS algorithm is susceptible

Journal homepage: http://ijpeds.iaescore.com

1084  ISSN: 2088-8694

to optimization errors and increased computational burden, while the LMF algorithm exhibits a poor steady-
state response, focusing primarily on dynamic response. Furthermore, the combined LMS LMF-based control
scheme overcomes the shortfalls of the LMS control technique, such as a lower convergence rate [16]. To
handle voltage distortions and unbalances, the recursive least squares (RLS) technique is implemented for
DVR operation [17]. The second-order generalized integrator (SOGI) algorithm has limitations in DC offset
rejection and performance degradation during a distorted grid state. Preview study [18], a frequency adaptive
enhanced reduced-order generalized integrator (FAEROGI)-based control algorithm is implemented for DVR
operation. The FAEROGI-based control scheme offers notable benefits, including reduced settling time,
lower peak overshoot, and faster-unbalanced compensation.
The effectiveness of adaptive filters heavily relies on the selection of an appropriate cost function, a
critical factor in diverse noise environments [19]. To achieve superior steady-state alignment and enhanced
robustness in impulsive noise conditions, a generalized hyperbolic secant cost function has been
employed [20]. Similarly, the arctangent cost function has been integrated into the normalized least mean
square (NLMS) framework, demonstrating heightened resilience against impulsive interference [21].
A multi-convex adaptive least mean square (LMS) filter-based control strategy is proposed in [22] for DVR
applications, where each LMS sub-filter excels in harmonic suppression and dynamic tracking by utilizing
step-size adaptation for superior voltage distortion management. Further advancements are seen in [23],
where a modified fractional least mean square (M-FLMS) control scheme optimizes DVR performance. By
incorporating an adjustable gain parameter into the original FLMS structure, the M-FLMS method achieves
faster convergence, reduced steady-state errors, and higher computational efficiency during voltage
distortions. Alternatively, transformer-based impedance source inverters (Trans-Z source inverters) have
been explored for DVR operation in [24]. Notably, while methods in [21]–[23] showcase improved
robustness and performance metrics, they are accompanied by higher computational demands. Therefore,
there is a need for an algorithm with a novel cost function that delivers both robust steady-state and dynamic
responses, even under unbalanced grid conditions, ensuring appropriate DVR operation. In [25], a study
employed the inverse hyperbolic sine function as a cost function, showcasing improved steady-state
accomplishment and increased resilience against impulsive interference when compared to several existing
algorithms. Hence, this work refers [25] to the effective operation of DVR.
The prime contributions of the technical manuscript are: i) The IHSF-LMS-based control scheme is
utilized for the efficient extraction of the fundamental component (FC) from a distorted source voltage. It
offers improved steady-state performance and demonstrates strong resilience to impulsive interference.
Additionally, its implementation is straightforward, eliminating the need for complex mathematical
computations; ii) A comparative analysis between the proposed IHSF-LMS control scheme and the
traditional LMS technique highlights its superior performance. Notably, the proposed algorithm's
computational burden is significantly lower than the methods discussed in [21]–[23]; iii) The proposed
control algorithm effectively ensures that the total harmonic distortion (THD) of the load voltage remains
well within acceptable limits, complying with the 5% threshold specified in the IEEE-519-2014 standard.

2. DVR CONFIGURATION
Figure 1 illustrates the MATLAB/Simulink model of the 3P3W-VSC-based DVR configuration
with a capacitor support. The major portion includes a three-phase supply with distortions created, such as
voltage sag, swell, distortions, and imbalance, and the three-phase-sensitive load, which needs to be protected
from these undesirable changes in supply voltage. The DC-link capacitor provides voltage as input during
different disturbances. The filter has been used at the output of the voltage source converter (VSC) to remove
switching harmonics generated by the VSC. The voltage is to be inserted into the system through a single-
phase series-connected injection transformer. The supply voltage is represented by 𝑣𝑠𝑎𝑏𝑐 and the load voltage
as well as the currents are depicted as 𝑣𝐿𝑎𝑏𝑐 and 𝑖𝐿𝑎𝑏𝑐 , respectively. 𝑉𝐷𝐶 is the DC-link voltage across the
capacitor and 𝑣𝑐𝑎𝑏𝑐 is the DVR-injected voltage during the disturbances.

3. IHSF-LMS BASED-CONTROL ALGORITHM

The main function of a DVR is to inject a compensating voltage in series with the power supply,
maintaining stable voltage regulation at the load terminal. The effectiveness of the control algorithm hinges
on isolating the fundamental source voltage component that aligns with the unit load current template. For
precise evaluation, the unit load current template must maintain phase synchronization with the load current
while exhibiting a unit magnitude.

Int J Pow Elec & Dri Syst, Vol. 16, No. 2, June 2025: 1083-1093
Int J Pow Elec & Dri Syst ISSN: 2088-8694  1085

Figure 1. MATLAB/Simulink model of the 3P3W-VSC-based DVR configuration with a capacitor support

3.1. Inverse hyperbolic sine function-based (IHSF) LMS adaptive filtering algorithm
Figure 2(a) illustrates a complete circuit representation of the inverse hyperbolic sine function-based
(IHSF) LMS adaptive filtering algorithm for DVR. The IHSF-LMS control-based extractor for phase ‘a’ is
depicted in Figure 2(b). The line voltages are represented as 𝑣𝑎𝑏 , and 𝑣𝑏𝑐 . After measuring it at the point of
common coupling (PCC), the conversion into phase voltages 𝑣𝑎 , 𝑣𝑏 and 𝑣𝑐 is as (1).
𝑣𝑎 2 1 𝑣
1 𝑎𝑏
𝑣
( 𝑏 ) = [−1 1 ] [𝑣 ] (1)
3 𝑏𝑐
𝑣𝑐 −1 −2
From the power circuit, the load currents will be sensed. Three-phase load current (𝑖𝐿 )is used to determine in-
phase and quadrature unit templates. Let 𝐼𝐿𝑎 is the magnitude of the three-phase load current which is given by (2).

2
𝐼𝐿𝑎 = √ (𝑖 2 𝐿𝑎 + 𝑖 2 𝐿𝑏 + 𝑖 2 𝐿𝑐 ) (2)
3

The in-phase as well as quadrature unit templates are given by (3) and (4).
𝑖𝐿𝑎
𝜎𝑝𝑎 (𝑛) =
𝐼𝐿𝑎
𝑖𝐿𝑏
𝜎𝑝𝑏 (𝑛) = (3)
𝐼𝐿𝑎
𝑖𝐿𝑐
𝜎𝑝𝑐 (𝑛) =
𝐼𝐿𝑎 }

𝜎𝑞𝑎 (𝑛) = 0.577 ∗ [𝜎𝑝𝑐 (𝑛) − 𝜎𝑝𝑏 (𝑛)]

𝜎𝑞𝑏 (𝑛) = 0.289 ∗ [3𝜎𝑝𝑎 (𝑛) + 𝜎𝑝𝑏 (𝑛) − 𝜎𝑝𝑐 (𝑛)] } (4)
𝜎𝑞𝑐 (𝑛) = 0.289 ∗ [−3𝜎𝑝𝑎 (𝑛) + 𝜎𝑝𝑏 (𝑛) − 𝜎𝑝𝑐 (𝑛)]

The utilized cost function in this context is the inverse hyperbolic sine function (IHSF) with the squared error
as its argument, represented by (5).

𝐽𝐻𝑆𝐹 (𝑒(𝑛)) = 𝐸[𝑠𝑖𝑛ℎ−1 ( 𝑒 2 (𝑛))] (5)

A novel adaptive control scheme for dynamic voltage restorer (Tummala Kranti Kiran)
1086  ISSN: 2088-8694

The term 𝑠𝑖𝑛ℎ−1 [. ] is inverse hyperbolic sine function. As far as the definition of the IHSF is
concerned, it is vigorously convex at 𝑒(𝑛) ∈ (−∞, ∞). The cost function employed to derive the proposed
IHCF adaptive filtering algorithm is expressed as (6).

𝛼(𝑛) = 𝑎𝑟𝑔𝑚𝑖𝑛 𝐽 [𝛼(𝑛)] = 𝑎𝑟𝑔𝑚𝑖𝑛 𝐽 [𝑠𝑖𝑛ℎ−1 ( 𝑒 2 (𝑛))] (6)

Its gradient with respect to the filter coefficient is as (7).

𝜕𝐽[𝛼(𝑛)] 𝜕[𝑠𝑖𝑛ℎ−1 (𝑒 2 (𝑛))]
= = −2𝑒(𝑛)𝑥(𝑛) (7)
𝜕[𝛼(𝑛)] 𝜕[𝛼(𝑛)]

Hence, the updated weight equation for IHSF-LMS is as (8).

1
𝛼(𝑛) = 𝛼(𝑛 − 1) + 2𝛾 × 𝑒(𝑛)𝑥(𝑛) (8)
√1+𝑒 2 (𝑛)

The term 𝛾 is adaptation constant and 𝑥(𝑛) is input vector. The term 𝑒𝑝𝑎(𝑛) in (9) is the error
between substantial and fundamental estimated signals. Furthermore, the weight transfer rule for active
components of phase ‘b’ and ‘c’ is given by (9).
1
𝛼𝑝 𝑏 (𝑛) = 𝛼𝑝 𝑏 (𝑛 − 1) + 2𝛾 𝑒𝑝𝑏 (𝑛)𝜎𝑝𝑏 (𝑛)
√1+𝑒𝑝𝑏 2 (𝑛)

𝑒𝑝𝑏 (𝑛) = 𝑣𝑠𝑏 (𝑛) − 𝜎𝑝𝑏 (𝑛) ∗ 𝛼𝑝𝑏 (𝑛 − 1)

1 (9)
𝛼𝑝 𝑐 (𝑛) = 𝛼𝑝 𝑐 (𝑛 − 1) + 2𝛾 𝑒𝑝𝑐 (𝑛)𝜎𝑝𝑐 (𝑛)
√1+𝑒𝑝𝑐 2 (𝑛)

𝑒𝑝𝑐 (𝑛) = 𝑣𝑠𝑐 (𝑛) − 𝜎𝑝𝑐 (𝑛) ∗ 𝛼𝑝𝑐 (𝑛 − 1) }

The term  sp is the mean of 𝛼𝑝𝑎 (𝑛), 𝛼𝑝𝑏 (𝑛), 𝛼𝑝𝑐 (𝑛) which active component of all three phases is given by (10).

𝛼𝑠𝑝 = 𝑎𝑣𝑒𝑟𝑎𝑔𝑒(𝛼𝑝𝑎 (𝑛), 𝛼𝑝𝑏 (𝑛), 𝛼𝑝𝑐 (𝑛)) (10)

It represents the mean weight of the fundamental d-axis component in the reference supply voltage.
Similarly, the weight transfer rule for a reactive component of phase ‘a,’ ‘b’, and ‘c’ is given by (11).
1
𝛼𝑞 𝑎 (𝑛) = 𝛼𝑞 𝑎 (𝑛 − 1) + 2𝛾 𝑒𝑞𝑎 (𝑛)𝜎𝑞𝑎 (𝑛)
√1+𝑒𝑞𝑎 2 (𝑛)

𝑒𝑞𝑎 (𝑛) = 𝑣𝑠𝑎 (𝑛) − 𝜎𝑞𝑎 (𝑛) ∗ 𝛼𝑞𝑎 (𝑛 − 1)

1
𝛼𝑞 𝑏 (𝑛) = 𝛼𝑞 𝑏 (𝑛 − 1) + 2𝛾 𝑒𝑞𝑏 (𝑛)𝜎𝑞𝑏 (𝑛)
√1+𝑒𝑞𝑏 2 (𝑛) (11)
𝑒𝑞𝑏 (𝑛) = 𝑣𝑠𝑏 (𝑛) − 𝜎𝑞𝑏 (𝑛) ∗ 𝛼𝑞𝑏 (𝑛 − 1)
1
𝛼𝑞 𝑐 (𝑛) = 𝛼𝑞 𝑐 (𝑛 − 1) + 2𝛾 𝑒𝑞𝑐 (𝑛)𝜎𝑞𝑐 (𝑛)
√1+𝑒𝑞𝑐 2 (𝑛)

𝑒𝑞𝑐 (𝑛) = 𝑣𝑠𝑐 (𝑛) − 𝜎𝑞𝑐 (𝑛) ∗ 𝛼𝑞𝑐 (𝑛 − 1) }

In (11), 𝛼𝑠𝑞 is the mean of 𝛼𝑞𝑎 (𝑛), 𝛼𝑞𝑏 (𝑛), 𝛼𝑞𝑐 (𝑛) which reactive component of all the three phases and it is
given by (12).

𝛼𝑠𝑞 = 𝑎𝑣𝑒𝑟𝑎𝑔𝑒(𝛼𝑞𝑎 (𝑛), 𝛼𝑞𝑏 (𝑛), 𝛼𝑞𝑐 (𝑛)) (12)

It represents the mean weight of the FC of the q-axis in the reference supply voltage. The signal
𝛼𝑠𝑝 is passed through a LPF that separates the non-oscillating portion. It extracts DC apparatus and
harmonics. The error (𝑣𝐷𝐶𝑒 ) is obtained by subtracting the actual DC-bus voltage 𝑉𝐷𝐶 from reference DC bus
voltage (𝑣𝑑𝑐 ∗ ). This is provided to the DC side PI controller. The output of the DC-PI controller is 𝑉𝑑𝑃𝐼
regulates 𝑉𝐷𝐶 . The term 𝑉𝑝𝑑 will be (13).

𝑉𝑝𝑑 = 𝛼𝑠𝑝 − 𝑉𝑑𝑝𝐼 (13)

Let the load terminal voltage (𝑣𝐿𝑡 ) is obtained through (14).

Int J Pow Elec & Dri Syst, Vol. 16, No. 2, June 2025: 1083-1093
Int J Pow Elec & Dri Syst ISSN: 2088-8694  1087

2
𝑣𝐿𝑡 = √ (𝑣 2 𝐿𝑎 + 𝑣 2 𝐿𝑏 + 𝑣 2 𝐿𝑐 ) (14)
3

VDC
V * DC +
VDCe
vsa ( n ) IHSF based LMS for Extraction of  pa (n) PI
fundamental active power
 pa (n) component of phase 'a' + VdPI
 pb (n)  sp V pd
v (n) IHSF based LMS for Extraction of
sb
fundamental active power
+ 1/3 LPF + V * pabc
X
 pb (n) component of phase 'b'  pc (n)  pa (n)
+
vsc ( n ) IHSF based LMS for Extraction of

 pc (n)
fundamental active power  pb (n) X
component of phase 'c' V * Labc

 pc (n) X

(a) V *
t
vLabc  qa (n)
Amplitude
+ X
Calculation
Vte  qb (n)

IHSF based LMS for Extraction of  qa (n) PI X

vsa ( n ) V *qabc
fundamental reactive power
 qa (n) +  sq VtPI  qc (n)
component of phase 'a'  qb (n) X
+
IHSF based LMS for Extraction of +
v (n) + 1/3 LPF + VLabc
sb Vql
fundamental reactive power
 qb (n) component of phase 'b'  qc (n)
+
IHSF based LMS for Extraction of PWM Gate
v sc ( n ) Pulse
fundamental reactive power
 qc (n) component of phase 'c' iLabc
Amplitude Calculation Generator

 pa (n)  pb (n)  pc (n)  qa (n)  qb (n)  qc (n)

−1
Z

 pa (n − 1)  p a ( n ) =  p a ( n − 1) + 2
1
e pa ( n ) pa ( n )
vsa ( n ) 1 + e pa 2 ( n )
e pa (n) = vsa ( n ) −  pa ( n ) *  pa ( n−1)  pa (n)
 pa (n)
(b)
−1
Z

 qa (n − 1) 1
 q a ( n ) =  q a ( n − 1) + 2  e q a ( n ) qa (n)
1 + eqa 2 ( n )
vsa ( n )
eqa (n) = vsa ( n ) −  qa ( n ) *  qa ( n−1)  qa (n)
 qa (n)

Figure 2. Control diagram of DVR: (a) IHSF-LMS-based control diagram for DVR and (b) IHSF-LMS
control algorithm-based extractor

The reference load terminal voltage is (𝑣 ∗ 𝐿𝑡 ). The 𝑣𝐿𝑡 is compared with reference load terminal
voltage that generates an error (𝑣𝑡𝑒𝑟 ). The error signal is given to the AC side PI controller which gives 𝑣𝐿𝑃𝐼
as output. This output is essential to regulate the terminal load voltage of the DVR. 𝛼𝑠𝑞 is passed through LPF
to extract load-reactive components. The output of AC-PI controller 𝑣𝐿𝑃𝐼 is added to the average three-phase
reactive fundamental component 𝛼𝑠𝑞 that gives signal 𝑉𝑞𝑙 .

𝑉𝑞𝑙 = 𝛼𝑠𝑞 + 𝑉𝐿𝑃𝐼 (15)

∗ ∗
Three-phase active and reactive components of reference voltage 𝑣𝑝𝑎𝑏𝑐 and 𝑣𝑞𝑎𝑏𝑐 are given by (16) and (17).

𝑣 ∗ 𝑝𝑎𝑏𝑐 = 𝑉𝑝𝑑 ∗ 𝜎𝑝𝑎𝑏𝑐 (𝑛) (16)

𝑣 ∗ 𝑞𝑎𝑏𝑐 = 𝑉𝑞𝑙 ∗ 𝜎𝑞𝑎𝑏𝑐 (𝑛) (17)

The signal (𝑣 ∗ 𝐿𝑎𝑏𝑐 ) will be determined with the help of (16) and (17) as (18).

𝑣 ∗ 𝐿𝑎𝑏𝑐 = 𝑣 ∗ 𝑝𝑎𝑏𝑐 + 𝑣 ∗ 𝑞𝑎𝑏𝑐 (18)

The error generated by comparing (𝑣 ∗ 𝐿𝑎𝑏𝑐 ) and (𝑣𝐿𝑎𝑏𝑐 ) is utilized for PWM generation.
A novel adaptive control scheme for dynamic voltage restorer (Tummala Kranti Kiran)
1088  ISSN: 2088-8694

4. SIMULATION RESULT OF DVR USING IHSF-LMS-BASED ADAPTIVE FILTERING

ALGORITHM
The 3P3W DVR utilizing the inverse hyperbolic sine function-based LMS (IHSF-LMS) control
algorithm has been modeled and simulated in MATLAB Simulink. A sample time of 20 microseconds was
utilized for the simulations. Voltage swell and sag events were applied for durations of 0.4 ≤ 𝑡 ≤ 0.48 s. and
0.54 ≤ 𝑡 ≤ 0.62 s., respectively, while source voltage distortion was introduced over a range of 0.7 ≤ 𝑡 ≤
0.78 s. Additionally, voltage unbalance was observed at 0.84 ≤ 𝑡 ≤ 0.9 s. The subsequent sections provide a
detailed analysis of DVR operation. Intermediate signals are analyzed to assess the DVR's performance under
the specified control scheme. Simulation results are discussed in the following sub-sections.

4.1. Intermediate signals of the IHSF-LMS-based control scheme

Figures 3(a) and 3(b) represent the internal signals of IHSF-LMS from the supply voltage of phase
‘a’ (vsa). The first subplot in Figures 3(a) and 3(b) indicates source side voltage (vsabc) with aforementioned
PQ issues. The active FC of the vsabc of phase ‘a’ is represented by 𝛼𝑝𝑎 in subplot two. The signal 𝛼𝑠𝑝
∗
obtained from (10) is the third subplot of Figures 3(a) and 3(b). The term 𝑣𝑝𝑎 calculated from (18) which is
the reference active component of load voltage for phase ‘a’ can be depicted in subplot four. Moreover, the
reactive FC of the vsabc with the phase ‘a’ i.e., 𝛼𝑞𝑎 is represented by the subplot fifth of Figures 3(a) and 3(b).
∗
A signal 𝛼𝑠𝑞 obtained from (12) is illustrated in the sixth subplot of Figures 3(a) and 3(b). The term 𝑣𝑞𝑎
calculated from (17) which is the reference reactive component of load voltage for phase ‘a’ is illustrated in
subplot seventh. Finally, the reference load voltage obtained from (18) is articulated in subplot eight. From
Figures 3(a) and 3(b), it is observed that the offered IHSF-LMS-based control scheme produces a balanced
three-phase reference load voltage even though the source voltage is distorted.

4.2. DVR overall performance with IHSF-LMS-based control scheme

Figure 4 demonstrates the comprehensive performance of the DVR using the IHSF-based LMS control
technique under various dynamic conditions. Figure 4(a) includes voltage swell and sag conditions, whereas
Figure 4(b) includes voltage distortion and unbalanced conditions. For both Figures 4(a) and 4(b), the subplots
are the same viz. supply voltages (vsabc), load voltages (vLabc), compensated voltages (vca, vcb, and vcc), load
currents (iLabc), and DC-bus voltage (Vdc) and terminal voltage (Vt). From both Figures 4(a) and 4(b), it is
noteworthy that the load voltage is balanced and of constant magnitude. This indicates the effectiveness of the
proposed control scheme. Harmonic analysis vsabc proposed IHSF-LMS and conventional LMS are shown in
Figure 5(a). Significantly improved using the proposed control scheme, achieving 1.61%, compared to 3.03%
with the conventional LMS, as illustrated in Figure 5(b). Similarly, the load current THD is reduced to 1.48% for
the proposed controller, whereas the conventional LMS achieves 2.05%, as seen in Figure 5(c). These results
highlight the superior harmonic mitigation capabilities of the IHSF-LMS controller over the conventional LMS
algorithm, particularly in improving the quality of load voltage and current. Additionally, Table 1 provides the
simulation performance of the IHSF-based LMS algorithm.

(a) (b)

Figure 3. Intermediate signals for proposed IHSF-LMS-based control scheme during: (a) voltage swell and sag
and (b) voltage distortion and unbalance

Int J Pow Elec & Dri Syst, Vol. 16, No. 2, June 2025: 1083-1093
Int J Pow Elec & Dri Syst ISSN: 2088-8694  1089

(a) (b)

Figure 4. Comprehensive effectiveness of the DVR with the IHSF-LMS control scheme:
(a) voltage swell and sag (b) voltage distortion and voltage unbalance

Table 1. Simulation performance of the IHSF-based LMS algorithm

Sr. No. Disturbance nature Parameter in RMS values
1. DVR behavior 0.82 p.u. sag 𝑣𝑠𝑎𝑏𝑐 = 339.2 V, 𝑣𝑐𝑎𝑏𝑐 = 75.83 V, 𝑣𝐿𝑎𝑏𝑐 = 415 V
2. (dynamic condition) 1.2 p.u. swell 𝑣𝑠𝑎𝑏𝑐 = 443 V, 𝑣𝑐𝑎𝑏𝑐 = 31.54 V, 𝑣𝐿𝑎𝑏𝑐 = 415 V
3. Distortion 𝑣𝑠𝑎𝑏𝑐 = 316.4 V, 𝑣𝑐𝑎𝑏𝑐 = 30 V, 𝑣𝐿𝑎𝑏𝑐 = 415 V
4. Unbalance (resistive load is 𝑣𝑠𝑎𝑏𝑐 = 415 V, 𝑣𝑐𝑎𝑏𝑐 = 30.68 V, 𝑣𝐿𝑎𝑏𝑐 = 415 V
connected between phases ‘a’)
5. DVR behavior (steady-state condition) Voltage distortion 𝑣𝑠𝑎𝑏𝑐 = 9.79%, 𝑣𝐿𝑎𝑏𝑐 = 9.79 %

5. COMPARISION OF IHSF-LMS-BASED CONTROL WITH CONVENTIONAL LMS CONTROL

TECHNIQUE
The effectiveness of IHSF-LMS is compared with conventional LMS. The fundamental active
weight component of phase 'a' is compared to assess its effectiveness with the LMS and IHSF-LMS-based
control schemes. It is depicted in Figure 6. The wave shapes show that there are persistent oscillations with
considerable magnitude in the case of the LMS controller. Oscillations with reduced amplitude are observed
with the IHSF-LMS controller. It shows that the weight converges faster, and a smaller static error is
achieved in steady-state conditions with (t = 0.1 to 0.5 s.) for the proposed IHSF-LMS control scheme. Now,
the unbalanced load condition with a time frame of t = 0.8 to t = 0.85 s is considered. It is noteworthy that
during dynamic situations, the convergence of the proposed IHSF-LMS-based control technique is better
with lesser steady-state error. It shows its effectiveness during voltage anomalies.
Figures 5(a)-5(c) illustrate the harmonic analysis of load voltage using LMS, IHSF-LMS, and source
voltage, respectively. Meanwhile, Figure 7 highlights the DC-link voltage response for both the proposed
IHSF-LMS control scheme and the conventional LMS algorithm, with an acceptable range of 2% applied for
DC-bus voltage settling. Analyzing the settling time and overall response in Figure 7, it becomes evident that
the IHSF-LMS control scheme outperforms the conventional LMS in dynamic performance. The proposed
design excels in key areas, including weight convergence, error minimization, reduced peak overshoot, faster
rise time, improved settling time, and effective harmonic mitigation. The comparative results summarized in
Table 2 further emphasize the superior performance of the IHSF-LMS controller over the conventional LMS,
validating its effectiveness as a robust and efficient control strategy.

Table 2. A comparative study of the conventional algorithm with the proposed IHSF-LMS-based control scheme
Operations LMS IHSF-LMS
Type Adaptive Adaptive
Convergence Slower Faster than LMS
Complexity Less Less
Convergence of weight Steady-state condition Higher amplitude having sustained oscillations Lower amplitude having sustained oscillations
Unbalanced condition Sustained oscillations Faster convergent
THD analysis Source voltage 25.53% 25.53%
Load voltage 3.03% 1.61%
Load current 2.05% 1.48%
Steady state error More than IHSF-LMS Significantly less than LMS
Max. overshoot (%Mp) 4.45% 2.92%
Rise time (tr s) 0.084 s 0.082 s
Settling time (ts s) 0.135 s 0.118 s

A novel adaptive control scheme for dynamic voltage restorer (Tummala Kranti Kiran)
1090  ISSN: 2088-8694

(a)

(b)

(c)

Figure 5. Harmonic analysis for (a) source voltage, (b) load voltage, and (c) load current,
with proposed IHSF-LMS and conventional LMS control scheme

Figure 6. Comparison between LMS and IHSF-LMS control scheme based on the fundamental active
component of phase ‘a’

Int J Pow Elec & Dri Syst, Vol. 16, No. 2, June 2025: 1083-1093
Int J Pow Elec & Dri Syst ISSN: 2088-8694  1091

Figure 7. Response of DVR DC-link voltage with tuned PI controller gains:

(a) IHSF-LMS and (b) conventional LMS

6. CONCLUSION
A 3P3W VSC-based DVR, integrated with the IHSF-LMS control scheme, underwent a
comprehensive evaluation to assess its performance across diverse voltage conditions, including sag, swell,
unbalance, and distortion in the supply voltage. The IHSF-LMS framework showcased superior adaptability,
effectively tracking supply voltage variations while maintaining an optimal balance between convergence
speed and misalignment. The results demonstrate a remarkable improvement in DVR functionality,
delivering rapid convergence, enhanced robustness, and reliable compensation capabilities. Compared to the
conventional LMS algorithm, the proposed IHSF-LMS control strategy achieved notable advantages,
including reduced peak overshoot, quicker settling time, and shorter rise time. Notably, the load voltage's
THD consistently adhered to IEEE-519-2014 standards, underscoring the algorithm's efficiency in harmonic
suppression. This control scheme excelled in both dynamic and steady-state scenarios, reinforcing its
reliability and robustness. System validation through MATLAB/Simulink further confirmed the significant
enhancements in the DVR's overall performance, solidifying the IHSF-LMS control algorithm as a superior
solution for power quality improvement. Parameters: system parameters for simulation: grid voltage with
voltage anomalies- 3 phase, 415 V, 50 Hz; Load of 10 kVA with 0.82 p.f. (Lagg.); rating of injection-
transformer = 11 kVA, 200/200 V; DC-link voltage (Vdc) = 300 V; DC bus capacitor (Cdc) = 3300 µF;
Interfacing inductor (Li) = 3 mH; passive ripple filter: Rr = 6 Ω and Cr =10 µF; switching frequency of VSC:
(fs) = 10 kHz., adaptation constant  = 0.1.

FUNDING INFORMATION
The authors received no specific funding for this work.

AUTHOR CONTRIBUTIONS STATEMENT

This journal uses the Contributor Roles Taxonomy (CRediT) to recognize individual author
contributions, reduce authorship disputes, and facilitate collaboration.

Name of Author C M So Va Fo I R D O E Vi Su P Fu
Tummala Kranti Kiran ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓
Balakrishnan Rajagopal ✓ ✓ ✓ ✓ ✓ ✓ ✓
Yerramilli Butchi Raju ✓ ✓ ✓ ✓ ✓ ✓ ✓

C : Conceptualization I : Investigation Vi : Visualization

M : Methodology R : Resources Su : Supervision
So : Software D : Data Curation P : Project administration
Va : Validation O : Writing - Original Draft Fu : Funding acquisition
Fo : Formal analysis E : Writing - Review & Editing

A novel adaptive control scheme for dynamic voltage restorer (Tummala Kranti Kiran)
1092  ISSN: 2088-8694

CONFLICT OF INTEREST STATEMENT

Authors state no conflict of interest.

DATA AVAILABILITY
Data availability is not applicable to this paper as no new data were created or analyzed in this study.

REFERENCES
[1] B. Singh, A. Chandra, and K. Al-Haddad, Power Quality Problems and Mitigation Techniques, vol. 9781118922. 2015. doi:
10.1002/9781118922064.
[2] M. H. J. Bollen, “What is power quality?,” Electric Power Systems Research, vol. 66, no. 1, pp. 5–14, 2003, doi: 10.1016/S0378-
7796(03)00067-1.
[3] A. Javadi, A. Hamadi, L. Woodward, and K. Al-Haddad, “Experimental Investigation on a Hybrid Series Active Power
Compensator to Improve Power Quality of Typical Households,” IEEE Transactions on Industrial Electronics, pp. 1–1, 2016,
doi: 10.1109/TIE.2016.2546848.
[4] A. Domijan, A. Montenegro, A. Keri, and K. Mattern, “Custom Power Devices: An Interaction Study,” IEEE Transactions on
Power Systems, vol. 20, no. 2, pp. 1111–1118, May 2005, doi: 10.1109/TPWRS.2005.846101.
[5] S. Das, A. K. Pradhan, A. Kedia, S. Dalai, B. Chatterjee, and S. Chakravorti, “Diagnosis of Power Quality Events Based on
Detrended Fluctuation Analysis,” IEEE Transactions on Industrial Electronics, vol. 65, no. 9, pp. 7322–7331, Sep. 2018, doi:
10.1109/TIE.2018.2795559.
[6] D. Rajasekara, S. Sekhar Das, and P. Vignesh, “Mitigation of Voltage Sags and Voltage Swells by Dynamic Voltage Restorer,” in
International Conference on Advances in Recent Technologies in Communication and Computing, 2011, vol. 5, no. 3, pp. 36–40.
doi: 10.3923/ijepe.2011.139.143.
[7] Changjiang Zhan et al., “Dynamic voltage restorer based on voltage-space-vector PWM control,” IEEE Transactions on Industry
Applications, vol. 37, no. 6, pp. 1855–1863, 2001, doi: 10.1109/28.968201.
[8] Changjiang Zhan, C. Fitzer, V. K. Ramachandaramurthy, A. Arulampalam, M. Barnes, and N. Jenkins, “Software phase-locked
loop applied to dynamic voltage restorer (DVR),” in 2001 IEEE Power Engineering Society Winter Meeting. Conference
Proceedings (Cat. No.01CH37194), vol. 3, pp. 1033–1038. doi: 10.1109/PESW.2001.917210.
[9] K. Chandrasekaran and V. K. Ramachandaramurthy, “An improved Dynamic Voltage Restorer for power quality improvement,”
International Journal of Electrical Power and Energy Systems, vol. 82, pp. 354–362, 2016, doi: 10.1016/j.ijepes.2016.02.036.
[10] P. Kanjiya, B. Singh, A. Chandra, and Kamal-Al-Haddad, “‘SRF theory revisited’ to control self-supported dynamic voltage
restorer (DVR) for unbalanced and nonlinear loads,” in 2011 IEEE Industry Applications Society Annual Meeting, Oct. 2011, pp.
1–8. doi: 10.1109/IAS.2011.6074356.
[11] P. Shukla and B. Singh, “Synchronization of Solar PV System to Grid with Enhanced Power Quality,” in 2021 4th Biennial
International Conference on Nascent Technologies in Engineering (ICNTE), Jan. 2021, pp. 1–6. doi:
10.1109/ICNTE51185.2021.9487702.
[12] V. Deshpande Chinmay and V. Deshpande Chaitanya, “Optimum design of dynamic voltage restorer for voltage sag mitigation in
distribution network,” International Journal of Power Electronics and Drive Systems, vol. 10, no. 3, pp. 1364–1372, 2019, doi:
10.11591/ijpeds.v10.i3.pp1364-1372.
[13] K. R. Patil and H. H. Patel, “Modified dual second-order generalised integrator FLL for synchronization of a distributed generator
to a weak grid,” in 2016 IEEE 16th International Conference on Environment and Electrical Engineering (EEEIC), Jun. 2016, pp.
1–5. doi: 10.1109/EEEIC.2016.7555824.
[14] B. Singh, P. Jayaprakash, and D. P. Kothari, “Adaline-based control of capacitor supported DVR for distribution System,”
Journal of Power Electronics, vol. 9, no. 3, pp. 386–395, 2009.
[15] H. M. M. Alhaj, N. M. Nor, V. S. Asirvadam, and M. F. Abdullah, “Power system harmonics estimation using LMS, LMF and
LMS/LMF,” in 2014 5th International Conference on Intelligent and Advanced Systems (ICIAS), Jun. 2014, pp. 1–5. doi:
10.1109/ICIAS.2014.6869521.
[16] M. Srinivas, I. Hussain, and B. Singh, “Combined LMS–LMF-Based Control Algorithm of DSTATCOM for Power Quality
Enhancement in Distribution System,” IEEE Transactions on Industrial Electronics, vol. 63, no. 7, pp. 4160–4168, Jul. 2016, doi:
10.1109/TIE.2016.2532278.
[17] J. R. S. Martins, D. A. Fernandes, F. F. Costa, M. B. R. Corrêa, A. J. Sguarezi Filho, and E. R. C. da Silva, “Optimized voltage
injection techniques for protection of sensitive loads,” International Journal of Electrical Power and Energy Systems, vol. 116,
2020, doi: 10.1016/j.ijepes.2019.105569.
[18] C. V. Deshpande, R. Chilipi, and S. R. Arya, “Frequency Adaptive Enhanced Reduced-Order Generalized Integrator Control for
Operation of DVR,” in 2023 IEEE International Conference on Energy Technologies for Future Grids (ETFG), Dec. 2023, pp. 1–
6. doi: 10.1109/ETFG55873.2023.10408206.
[19] A. Khalifi, Q. Mayyala, N. Iqbal, A. Zerguine, and K. Abed-Meraim, “Adaptive algorithm based on a new hyperbolic sine cost
function,” in 2017 51st Asilomar Conference on Signals, Systems, and Computers, Oct. 2017, pp. 812–815. doi:
10.1109/ACSSC.2017.8335459.
[20] L. Lu, L. Chen, Z. Zheng, Y. Yu, and X. Yang, “Behavior of the LMS algorithm with hyperbolic secant cost,” Journal of the
Franklin Institute, vol. 357, no. 3, pp. 1943–1960, Feb. 2020, doi: 10.1016/j.jfranklin.2019.12.040.
[21] J. Zeng, Y. Lin, and L. Shi, “A Normalized Least Mean Square Algorithm Based on the Arctangent Cost Function Robust Against
Impulsive Interference,” Circuits, Systems, and Signal Processing, vol. 35, no. 8, pp. 3040–3047, Aug. 2016, doi:
10.1007/s00034-015-0175-5.
[22] C. Deshpande, R. Chilipi, and S. R. Arya, “A multi-convex composition of adaptive LMS filter-based control for dynamic voltage
restorer with SSO-optimized PI gains,” Electrical Engineering, Feb. 2024, doi: 10.1007/s00202-023-02119-4.
[23] C. V. Deshpande, R. Chilipi, and S. R. Arya, “Modified fractional least mean square-based control scheme for dynamic voltage
restorer to improve power quality,” Electrical Engineering, vol. 106, no. 4, pp. 5069–5087, Aug. 2024, doi: 10.1007/s00202-024-
02270-6.

Int J Pow Elec & Dri Syst, Vol. 16, No. 2, June 2025: 1083-1093
Int J Pow Elec & Dri Syst ISSN: 2088-8694  1093

[24] V. Deshpande Chinmay, V. Deshpande Chaitanya, and A. Deokar Sanjay, “Performance evaluation of dynamic voltage restorer
based on transformer-based Z source inverter,” International Journal of Power Electronics and Drive Systems, vol. 8, no. 3, pp.
1101–1108, 2017, doi: 10.11591/ijpeds.v8i3.pp1101-1108.
[25] S. Guan, Q. Cheng, Y. Zhao, and B. Biswal, “Robust adaptive filtering algorithms based on (inverse)hyperbolic sine function,”
PLoS ONE, vol. 16, no. 10 October, 2021, doi: 10.1371/journal.pone.0258155.

BIOGRAPHIES OF AUTHORS

Tummala Kranti Kiran is an Assistant Professor in the Department of Electrical

Engineering at Sir C. R. Reddy College of Engineering, Eluru, Andhra Pradesh, India. He
received his B.Tech. and M.Tech. degrees in Electrical Engineering in 2003 and 2006
respectively. He has been working as an assistant professor since 2007. He is, at present, the
external research scholar of Annamalai University, Annamalainagar, Tamil Nadu, India. His
research interests include power systems, power electronics, motor drives, and renewable
energy. He can be contacted at email: [email protected].

Balakrishnan Rajagopal is an Associate Professor in the Department of

Electrical Engineering at Annamalai University, Annamalainagar, Tamil Nadu, India. He
received his B.E. M.E., and Ph.D. Degrees in Electrical Engineering. He has been working as
an associate professor since 2001. He published 10 papers in International Journals. His
research interests include electrical machines, power electronics, IoT-based systems,
renewable energy, high voltage engineering, and power systems. He can be contacted at email:
[email protected].

Yerramilli Butchi Raju is a Professor in the Department of Electrical

Engineering at Sir C. R. Reddy College of Engineering, Eluru, Andhra Pradesh, India. He
received his B.Tech, M.Tech., and Ph.D. degrees in Electrical Engineering in 1999, 2004 and
2012 respectively. He has been working as a professor since 2016. He published 15 papers in
International Journals. His research interests include power system dynamics, power
electronics, IoT-based systems, and renewable energy. He can be contacted at email:
[email protected].

A novel adaptive control scheme for dynamic voltage restorer (Tummala Kranti Kiran)