(EBook PDF) Recent Developments in Electrical and Electronics Engineering 1st edition by Poonam Singhal, Sakshi Kalra, Bhim Singh, Bansal 303108280X 9783031082801 full chapters - The full ebook with all chapters is available for download
(EBook PDF) Recent Developments in Electrical and Electronics Engineering 1st edition by Poonam Singhal, Sakshi Kalra, Bhim Singh, Bansal 303108280X 9783031082801 full chapters - The full ebook with all chapters is available for download
https://ebookball.com/product/ebook-pdf-electronics-and-electrical-
engineering-1st-editon-by-alan-zhao-1317411811-9781317411819-full-
chapters-22862/
Poonam Singhal
Sakshi Kalra
Bhim Singh
R. C. Bansal Editors
Recent
Developments
in Electrical
and Electronics
Engineering
Select Proceedings of ICRDEEE 2022
Lecture Notes in Electrical Engineering
Volume 979
Series Editors
Leopoldo Angrisani, Department of Electrical and Information Technologies Engineering, University of Napoli
Federico II, Naples, Italy
Marco Arteaga, Departament de Control y Robótica, Universidad Nacional Autónoma de México, Coyoacán,
Mexico
Bijaya Ketan Panigrahi, Electrical Engineering, Indian Institute of Technology Delhi, New Delhi, Delhi, India
Samarjit Chakraborty, Fakultät für Elektrotechnik und Informationstechnik, TU München, Munich, Germany
Jiming Chen, Zhejiang University, Hangzhou, Zhejiang, China
Shanben Chen, Materials Science and Engineering, Shanghai Jiao Tong University, Shanghai, China
Tan Kay Chen, Department of Electrical and Computer Engineering, National University of Singapore,
Singapore, Singapore
Rüdiger Dillmann, Humanoids and Intelligent Systems Laboratory, Karlsruhe Institute for Technology,
Karlsruhe, Germany
Haibin Duan, Beijing University of Aeronautics and Astronautics, Beijing, China
Gianluigi Ferrari, Università di Parma, Parma, Italy
Manuel Ferre, Centre for Automation and Robotics CAR (UPM-CSIC), Universidad Politécnica de Madrid,
Madrid, Spain
Sandra Hirche, Department of Electrical Engineering and Information Science, Technische Universität
München, Munich, Germany
Faryar Jabbari, Department of Mechanical and Aerospace Engineering, University of California, Irvine, CA,
USA
Limin Jia, State Key Laboratory of Rail Traffic Control and Safety, Beijing Jiaotong University, Beijing, China
Janusz Kacprzyk, Systems Research Institute, Polish Academy of Sciences, Warsaw, Poland
Alaa Khamis, German University in Egypt El Tagamoa El Khames, New Cairo City, Egypt
Torsten Kroeger, Stanford University, Stanford, CA, USA
Yong Li, Hunan University, Changsha, Hunan, China
Qilian Liang, Department of Electrical Engineering, University of Texas at Arlington, Arlington, TX, USA
Ferran Martín, Departament d’Enginyeria Electrònica, Universitat Autònoma de Barcelona, Bellaterra,
Barcelona, Spain
Tan Cher Ming, College of Engineering, Nanyang Technological University, Singapore, Singapore
Wolfgang Minker, Institute of Information Technology, University of Ulm, Ulm, Germany
Pradeep Misra, Department of Electrical Engineering, Wright State University, Dayton, OH, USA
Sebastian Möller, Quality and Usability Laboratory, TU Berlin, Berlin, Germany
Subhas Mukhopadhyay, School of Engineering and Advanced Technology, Massey University,
Palmerston North, Manawatu-Wanganui, New Zealand
Cun-Zheng Ning, Electrical Engineering, Arizona State University, Tempe, AZ, USA
Toyoaki Nishida, Graduate School of Informatics, Kyoto University, Kyoto, Japan
Luca Oneto, Department of Informatics, BioEngineering, Robotics and Systems Engineering, University of
Genova, Genova, Genova, Italy
Federica Pascucci, Dipartimento di Ingegneria, Università degli Studi “Roma Tre”, Rome, Italy
Yong Qin, State Key Laboratory of Rail Traffic Control and Safety, Beijing Jiaotong University, Beijing, China
Gan Woon Seng, School of Electrical and Electronic Engineering, Nanyang Technological University,
Singapore, Singapore
Joachim Speidel, Institute of Telecommunications, Universität Stuttgart, Stuttgart, Germany
Germano Veiga, Campus da FEUP, INESC Porto, Porto, Portugal
Haitao Wu, Academy of Opto-electronics, Chinese Academy of Sciences, Beijing, China
Walter Zamboni, DIEM—Università degli studi di Salerno, Fisciano, Salerno, Italy
Junjie James Zhang, Charlotte, NC, USA
The book series Lecture Notes in Electrical Engineering (LNEE) publishes the
latest developments in Electrical Engineering—quickly, informally and in high
quality. While original research reported in proceedings and monographs has
traditionally formed the core of LNEE, we also encourage authors to submit books
devoted to supporting student education and professional training in the various
fields and applications areas of electrical engineering. The series cover classical and
emerging topics concerning:
• Communication Engineering, Information Theory and Networks
• Electronics Engineering and Microelectronics
• Signal, Image and Speech Processing
• Wireless and Mobile Communication
• Circuits and Systems
• Energy Systems, Power Electronics and Electrical Machines
• Electro-optical Engineering
• Instrumentation Engineering
• Avionics Engineering
• Control Systems
• Internet-of-Things and Cybersecurity
• Biomedical Devices, MEMS and NEMS
For general information about this book series, comments or suggestions, please
contact [email protected].
To submit a proposal or request further information, please contact the Publishing
Editor in your country:
China
Jasmine Dou, Editor ([email protected])
India, Japan, Rest of Asia
Swati Meherishi, Editorial Director ([email protected])
Southeast Asia, Australia, New Zealand
Ramesh Nath Premnath, Editor ([email protected])
USA, Canada
Michael Luby, Senior Editor ([email protected])
All other Countries
Leontina Di Cecco, Senior Editor ([email protected])
** This series is indexed by EI Compendex and Scopus databases. **
Poonam Singhal · Sakshi Kalra · Bhim Singh ·
R. C. Bansal
Editors
Recent Developments
in Electrical and Electronics
Engineering
Select Proceedings of ICRDEEE 2022
Editors
Poonam Singhal Sakshi Kalra
J.C. Bose University of Science J.C. Bose University of Science
and Technology, YMCA and Technology, YMCA
Faridabad, India Faridabad, India
© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature
Singapore Pte Ltd. 2023
This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether
the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse
of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and
transmission or information storage and retrieval, electronic adaptation, computer software, or by similar
or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors, and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or
the editors give a warranty, expressed or implied, with respect to the material contained herein or for any
errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.
This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd.
The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721,
Singapore
Contents
v
vi Contents
Bhim Singh received his B.E. from the University of Roorkee, India, in 1977 and
M.Tech. (Power Apparatus & Systems) and Ph.D. from the Indian Institute of Tech-
nology (IIT) Delhi, India, in 1979 and 1983, respectively. Prof. Singh is also a
member of Governing Council of Central Power Research Institute. Prof. Singh has
guided 101 Ph.D. dissertations and 171 M.E./M.Tech./M.S.(R) theses. He has filed 94
patents. He has executed more than eighty sponsored and consultancy projects. He
has co-authored a textbook on power quality: Power Quality Problems and Miti-
gation Techniques published by John Wiley & Sons Ltd. in 2015. His areas of
interest include solar PV grid interface systems, microgrids, power quality moni-
toring, and mitigation, solar PV water pumping systems, improved power quality
ACDC converters, power electronics, electrical machines, drives, flexible alternating
transmission systems, and high-voltage direct current systems.
ix
x About the Editors
renewable energy and conventional power systems that further include wind, photo-
voltaics, hybrid power systems, distributed generation, grid integration of renewable
energy, power systems analysis, smart grid, and power quality. He has 25+ years
of experience in academia, research, and industry. He has over 350 publications in
reputed journals, conferences, and edited volumes.
Design and Implementation
of Fractional-Order PID Controller
for Magnetic Levitation System Using
Genetic Algorithm-Based Optimization
1 Introduction
Magnetic levitation is a most effective and popular contactless and frictionless tech-
nique in which a ferromagnetic object is draped in the air with the support of
magnetic fields. Use of magnetic fields is to counteract the gravitational force and
any other counteracceleration. Magnetic levitation has given frictionless, efficient,
and outstanding technologies [1].
In a magnetic levitation system, a steel ball made of ferromagnetic material is
levitated against gravity and its levitation position is controlled by electromagnetic
force of attraction generated by coil and force is controlled by controlling voltage
which is applied to the coil [1]. Due to the absence of mechanical contact the fric-
tional force in magnetic levitation system gets eliminated. This results in improved
system efficiency. There are various engineering applications of magnetic levitation
technology such as high-speed trains, wind turbine, and personal rapid transit. Posi-
tion of the suspended object is being measured by the use of optical sensor. Based on
the error generated, i.e., difference of desired and actual position of the suspended
object, a controller takes action and provides suitable current to electromagnetic coil.
It generates a required force which is applied on the ferromagnetic object to control
its position and make it stable and able to track a reference signal [2]. Without an
efficient controller, Maglev system is inherently unstable.
© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 1
P. Singhal et al. (eds.), Recent Developments in Electrical and Electronics Engineering,
Lecture Notes in Electrical Engineering 979,
https://doi.org/10.1007/978-981-19-7993-4_1
2 N. K. Soni et al.
The nonlinearity and instability of Maglev makes very challenging system. Most
of time traditional controllers are not perfectly able to handle the nonlinearity and
instability of system. Generally for controlling of such systems some traditional
controllers like PID, PI, fuzzy controller, and other controllers are used [3]. If we
use FOPID controller, it is advantageous in that it has two extra tuning parameters
(λ and μ) which provides additional degree of freedom for better control, speed in
response, flexibility, stability, and better handling of a nonlinear system [4]. Settling
time must be decreased to increase the stability, speed of response, and efficiency of
the system. Instability and nonlinearity of the open-loop magnetic levitation system
are two aspects that are represented by an extremely nonlinear differential equation
[5].
In this paper, magnetic levitation system is controlled by FOPID controller which
is a nonconventional control technique. Organization of this paper is as follows.
Section 2 explains about the mathematical model of the magnetic levitation system.
The description of GA algorithm is given in Section 3. FOPID controllers have been
introduced and explained in Section 4. Implementation of FOPID controller is shown
in Section 5. The results and comparative analysis are shown in Section 6. Finally,
Section 7 shows the conclusions and future scope.
The block diagram of magnetic levitation system is shown in Fig. 1. Magnetic levi-
tation system consists of an electromagnetic coil along with a pair of infrared trans-
mitter and sensor which can accurately sense the position of the ball [6]. The elec-
tromagnetic force exerted by the coil is controlled by controlling the current in the
electromagnetic coil. The current is controlled by controller in such a way so that the
metal object can float in the air space [7–10]. The electromagnetic force F which is
generated by electromagnetic coil and the gravitational force g are opposite which
manages to sustain the metal object in the air space. The generated electromagnetic
force F by the coil depends on the current I and the air gap X between the metal ball
and the electromagnet coil, the generated force F is given by [11–13]
( )2
im
F = Mg − K m (1)
xb
where
i m = current in electromagnetic coil (Ampere);
xb = ball distance from the( coil
/ (m)
)
g = gravitational constant m s 2 ;
km = magnetic force constant
M = mass of the metal ball (K g )
Design and Implementation of Fractional-Order PID Controller … 3
The metal ball position and value of electromagnetic coil current at the operating
2
point can be derived by putting ddtx2b = 0 in Eq. (2) which gives
/
Km
xbss = im (3)
Mg ss
where
xbss = metal ball position; i m ss = electromagnetic coil current.
In theoretical sense, this coil current is sufficient to generate force F to levitate the
metal ball to the desired location. But practically there is fair chance to fail due to any
variation at operating point which may be caused by external disturbances, uncer-
tainties in parameter, and others. That’s why we require an efficient controller which
is capable of handling such irregularities of the system. The system linearization can
be done by taking the approximates of xb and i m as
i m (t) ≜ î m + i m ss (5)
4 N. K. Soni et al.
where
xbss = ball position variations; i m ss = coil current variation.
Above variations of ball position and coil current are around the operating point.
The dynamic equation of the magnetic levitation system is given by Eq. (6).
( )2
d 2 x̂b îm + i m ss
M 2 = Mg − K m (6)
dt x̂b + xbss
Therefore
( )
d 2 x̂b 1 2K m i m2 ss 2K m i m ss
= x̂b − î m (8)
dt 2 M xb3ss xb2ss
X̂ b (s) K2
G(s) = =− 2 (9)
ˆIm (s) s − K1
2K m i m2 ss 2K m i m ss
where K 1 = M xb3ss
and K 2 = M xb2ss
−24.5250
G(s) = (10)
s 2 + 13.08
Design and Implementation of Fractional-Order PID Controller … 5
The step response and frequency response of the open-loop system are shown
in Figs. 2 and 4, respectively. In Fig. 3, we can see that the poles of the open-loop
system, which are located on the imaginary axis, i.e., at s = ± j3.6166 which means
that the system goes toward instability or having sustained oscillations which can be
seen in Fig. 2. When we make this system closed loop the system becomes unstable
6 N. K. Soni et al.
as shown in Fig. 5. One of the pole is lying on the right-hand side of the s-plane and
the closed-loop poles are located at s = ±3.3823. Therefore, a controller is required
for making closed-loop system stable which can control the position of the metal
ball so that it can levitate in the air space.
3 Description of GA Algorithm
The step response of closed-loop magnetic levitation system with PID controller
is shown in Fig. 8. We found that system is stable for the tuned parameter using
GA of PID controller. The time response characteristics for step input of considered
magnetic levitation system with PID controller tuned by the genetic algorithm is
shown in Table 3. It can be noted that PID controller has stabilized and improved
performance of system. Now we will implement the FOPID controller for comparison
of the performance with PID.
Design and Implementation of Fractional-Order PID Controller … 9
where n − 1 ≺ α ≺ n
n and α are an integer number and initial condition, respectively.
PID controller [18, 19] has been very popular and widely used controllers, but in the
last couple of decades the improvement in fractional calculus in the field of control
engineering has introduced new controller which is known as the fractional-order
PID controller and it becomes very popular in control applications. Fractional-order
P I λ D μ controller has two extra parameters λ and μ which makes it complex and
also provides extra degree of freedom as compared to integer-order PID controller.
Fractional-order P I λ D μ controller is described as
The generalization of PID controller can be done for different values of λ and μ
in (14) through fractional-order PID controller as shown in Fig. 9. PID (λ = 1, μ =
1), PI (λ = 1, μ = 0), PD (λ = 0, μ = 1), and P (λ = 0, μ = 0) are the special cases
of P I λ D μ controller.
After considering the Laplace transform of Eq. 14, the expression of FOPID
controller in s-domain is obtained as
U (s) KI
C(s) = = K p + λ + K D Sμ (15)
E(s) S
Fig. 9 Generalization of
PID
Design and Implementation of Fractional-Order PID Controller … 11
U (s) KI
C(s) = = Kp + + KDS (16)
E(s) S
PID controller can be implemented by FOPID so that in the present work PID
controller is also studied along with the study of P I λ D μ controller because PID
controller is easily implemented by fractional-order controller. The performance
comparison of PID with P I λ D μ is carried out on the basis of same design specifi-
cations to show that the fractional-order P I λ D μ controller has better performance
in terms of performance index considered.
The hardware and software realization of FOPID controller is difficult. Therefore,
an integer-order approximation method is used for the fractional-order elements [20].
In MATLAB, FOPID controller can be implemented using FOMCOM toolbox [4]
where Oustaloup’s approximation is realized.
5 Simulation Results
The different parameters of the FOPID controller for magnetic levitation system
are tuned and the optimized values of different controller parameters [K p , K i ,K d ,λ,
μ] are obtained. The considered performance index is integral time absolute error
(ITAE) and it is minimized by use of GA algorithm optimizer toolbox [14].
For fractional-order system, different performance index criterions like ITAE,
ISCO, ITSE, etc. can be applied. Different integral performance index criterion has
different advantages for every design of PID and FOPID controller. Among them,
the ITAE performance index criterion optimizes the absolute error and settling time,
which is not possible to achieve by other performance index. From the different liter-
ature study, it has been noted that the most popular and commonly used performance
index is ITAE for tuning of PID and FOPID controller. In this paper, we considered
the same objective function (ITAE) which is minimized using GA and it is given as
{∞
J= t|e(t)|dt (17)
0
Table 5 Parameter of GA
Parameter Value
used for FOPID tuning
Lower bound [1, 1, 1, 0, 0]
Upper bound [2, 8]
Size of population 60
Crossover type Scattered
Crossover fraction 0.8
Mutation fraction 0.01
Iterations 400
Selection Stochastic uniform
Table 6 Comparison of
Response PID controller FOPID controller
response characteristics of
FOPID and PID Rise time 0.0860 s 0.0644 s
Peak overshoot 32.0189% 17.5006%
Settling time 1.1182 s 0.9776 s
Steady-state error 0.00 0.00
input in which both FOPID and IOPID controllers are tuned by genetic algorithm
and the response graph clearly shows that there is a better improvement in the closed
step response with FOPID controller compared to the integer-order PID controller.
The various characteristics of time response of PID and FOPID are shown in
Table 6.
Design and Implementation of Fractional-Order PID Controller … 13
6 Conclusions
References
1. Ahmad I, Javaid MA (2010) Nonlinear model & controller design for magnetic levitation
system. In: 9th WSEAS international conference on signal processing, robotics and automation.
Cambridge, UK
2. Yaghoubi H, Barazi N, Aoliaei MR (2012) “Maglev, ch 6,” in Infrastructure design, signalling
and security in railway, pp 123–176, InTech, Rijeka, Croatia
3. Podlubny I, Dorcak L, Kostial I (1997) On fractional derivatives, fractional-order dynamic
systems and PIλDμ controllers. In: 36th IEEE conference on decision and control. San Diego,
US
4. Tepljakov A, Petlenkov E, Belikov J (2012) A flexible MATLAB tool for optimal fractional-
order PID controller design subject to specifications. In: Proceedings of the 31st Chinese control
conference, Hefei, Anhui, China, pp 4698–4703
5. Kumar T, Shimi S, Karanjkar D, Rana S (2014) Modeling, simulation and control of single
actuator magnetic levitation system. In: 2014 recent advances in engineering and computational
sciences (RAECS), pp 1–6. IEEE
6. Knospe CR, Collins EG (1996) Special issue on magnetic bearing control. IEEE Trans Cont
Syst Technol 4(5)
7. Rule RG (1980) Combined magnetic levitation and propulsion: the mag-transit concept. IEEE
Trans Veh Technol VT 29:41–49
8. Ion B, Trica A, Papusoiu G, Nasar AS (1988) Field tests on a maglev with passive guideway
linear inductor motor transportation system. IEEE Trans. Veh Technol VT 38:230–236
9. Yadav S, Tiwari JP, Nagar SK (2012) Digital control of magnetic levitation system using fuzzy
logic controller. Int J Comput Appl (0975–8887) 41(21)
14 N. K. Soni et al.
10. Yadav S, Verma SK, Nagar SK (2016) Optimized PID controller for magnetic levitation system.
IFAC-PapersOnLine 49(1):778–782
11. Kumar EV, Jerome J (2013) LQR based optimal tuning of PID controller for trajectory tracking
of magnetic levitation system. Proc Eng 64:254–264
12. Verma SK, Yadav S, Nagar SK (2015) Optimal fractional order PID controller for magnetic
levitation, Presented at 39th NSC-2015, 14th–16th December, Shiv Nadar University, Dadri,
India
13. Golob M, Tovornik B (2003) Modeling and control of the magnetic suspension system. ISA
Trans 42:89–100
14. Yadav S et al (2017) Performance enhancement of magnetic levitation system using teaching
learning bsased optimization. Alexandria Eng J
15. Ortigueira MD (2011) Fractional calculus for scientists and engineers. Springer, Berlin,
Germany
16. Caponetto R, Dongola G, Fortuna L, Petras I (2010) Fractional order systems: modeling and
control applications, vol. 72 of World Scientific Series on Nonlinear Science Series A, world
Scientific Publishing, Singapore
17. Baleanu D, Diethelm K, Scalas E, Trujillo JJ (2012) Fractional calculus models and numer-
ical methods. Nonlinearity and Chaos, World Scientific Publishing, Singapore, Series on
Complexity
18. Ogata K (2002) Modern control engineering. Prentice-Hall, Upper Saddle River, NJ, USA
19. Astrom KA, Hagglund T (1995) PID controller: theory, design and tuning, Instrument Society
of America, Research Triangle Park, NC, USA
20. Outstaloup A, Levron F, Mathieu B, Nanot FM (2000) Frequency-based complex noninteger
differentiator: Characterization and synthesis. IEEE Trans Circuits Syst I: Fundamental Theory
Appl 47(1):25–39
Spin-Transfer Torque MRAM
with Emerging Sensing Techniques
1 Introduction
Spin-transfer torque MRAM shows fast access speed, excellent endurance, retention,
simple cell structure high array density, low-power switching characteristics, and
excellent CMOS compatibility and scalability [1–3]. It has been used in embedded
systems, last-level cache, a microcontroller unit (MCU), and so on. Data is saved
as two or more distinct resistance conditions of an MTJ device in an MRAM bit.
Making working memory non-volatile and turning off the power of the memory
region in stand-by are effective ways to lower the operational power of computer
systems. STT- MRAM is currently the most probable choice for non-volatile working
memory in terms of operating parameters such as speed and low-voltage operability.
The rise in switching current as the technology shrinks down is one of MRAM’s
significant drawbacks. The spin-polarized current is applied vertically through the
MTJ component in STT-RAM, which overcomes the switching current problem.
Unlike CMOS dynamic RAM and static RAM, STT-MRAM is a technology in which
there is no leakage current in the data array. A decrease in the access transistor’s
threshold voltage exhibits two useful effects. First, a small transistor could be used
to provide the device with the necessary write current. Another one is the access
transistor’s resistance is decreased, further improving both the current and voltage
sense margins.
© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 15
P. Singhal et al. (eds.), Recent Developments in Electrical and Electronics Engineering,
Lecture Notes in Electrical Engineering 979,
https://doi.org/10.1007/978-981-19-7993-4_2
16 S. Kumari and R. Yadav
MTJ is a suitable device for implementing non-volatile memories [1]. These memo-
ries can solve the power issues in a nanoscale integrated circuit due to data holding
after power shutdown. The MTJ structure is the basic component of the MRAM.
MTJs are popular in the market due to their easy integration with the CMOS process
and low fabrication cost. The perpendicular spin-transfer torque (p-STT) is preferred
for low-power operation and improving scalability. In p-STT, the two FMs’ polar-
ization is perpendicular to the MTJ plane, indicating decreased switching current for
the same retention time. The main magnetism of MTJs for logic is their non-volatile
aspect, which permits in principle instant-on operation and zero leakage (Fig. 1).
An oxide barrier layer separates two ferromagnetic layers in a magnetic tunnel
junction (MTJ). The resistance is high or low dependent on whether the ferromagnetic
layers are magnetically anti-parallel or parallel as shown in Fig. 2. When both layers
have the same directions, it is called the RAP resistance of MTJ. Whereas when the
magnetic directions of layers are different it is called anti-parallel resistance of MTJ.
To design the STT-RAM, one reference layer’s magnetic orientation is fixed whereas
the free layer’s magnetic orientation can be modified by delivering a switching current
polarized by the reference layer magnetization.
The tunneling magnetoresistance ratio (TMR) determined the difference between
anti-parallel resistance and parallel resistance in an MTJ.
T1
Reference Layer
Insulator
Jsst
Free Layer
T2 T3
Tunneling Tunneling
oxide oxide
There are various ways to design MTJs-based logic circuits, but mostly two cate-
gories are preferred: in first one MTJs are arranged and used as a 2D memory array,
and in another category MTJs are used as a particular switch which is either in
parallel/series configuration to keep logic functionality or having the status of register
replacement devices.
The voltage sense amplifier approach improves the bit-error-rate performance signif-
icantly. The VSenseAmp is less affected by data and reference cell resistance fluc-
tuations. The VSenseAmp, which uses a single pMTJ cell as a reference, is a viable
choice for high-speed, low-power read operations [3].
Convolutional neural networks are popular for deep learning structures. These
models are made up of several layers, each with a separate function, that work
together to do tasks that are generally difficult for traditional algorithms to complete.
To overcome the complexity of metadata management, they proposed a grouping
mechanism in which some blocks are combined to further decrease the storage over-
head. To solve issues with MLC STT-RAM, including reliability and excessive power
consumption, a simple yet effective technique is proposed to concurrently enhance
their reliability and degrade power consumption [4]. Toggle Spin Torques (TST)-
MRAM is an excellent choice for upper level caches when rapid operation and long
write durability are required. The TST mechanism significantly improves the write
endurance of the toggle spin-torque MRAM as compared to STT-MRAM. As a result,
the TST-MRAM is particularly well suited to ultrafast and long-duration applications
like upper level caches. As one of the most important issues in magnetic RAMs, the
Spin Hall-assisted STT approach reduces writing energy [5].
To improve the bit error rate (BER) and sensing margin dynamic dual-reference
sensing technique (DDRS) is used. If the reference cell maintains the same structure
as those of the data cell, then no regularity difficulties exist in the suggested DDRS
scheme. In deeply scaled STT-MRAM, read reliability is the main challenge that can
be dealt with by using the DDRS technique. The main feature of the suggested DDRS
scheme is to enhance the sensing margin without increasing the read disturbance.
So that for STT MRAM, the DDRS scheme offers a powerful solution in deep
submicrometric skill nodes [6].
Two magneto-tunnel junctions (MTJs) and two transistors make up a non-volatile,
dependable static gain cell for cache memory applications. The variation in the gain
cell threshold voltage and the wide distribution in MTJs resistance in anti-parallel
and parallel states can be overcome by self-reference sensing circuits and connecting
two complementary MTJs, respectively. The high voltage is applied across the stack
complementary MTJ which enhances the sensing margin. The self-reference sensing
structure has a greater ON/OFF ratio as compared to 1T1MTJ STT-MRAM and a
smaller cell area as compared to SRAM [7].
The DMTJ consists of two MgO oxide barriers which act as the interefence layer. The
free layer in DMTJ is sandwiched between two MgO oxide barriers. The reference
lines have anti-parallel magnetizations, which increase the total torque acting on the
Spin-Transfer Torque MRAM with Emerging Sensing Techniques 19
Fig. 3 DMTJ
NSRS scheme does not require write operation so that it maintains the STT-RAM
short read latency and the non-volatility design. This scheme is used to overcome
MTJ’s resistance variation from bit to bit. The robustness of NSRS can be improved
by using three techniques: yield-driven cell current choosing, R-I curve skewing,
and ratio matching technique. Theoretical, “R-I curve skewing” and techniques
significantly increased the STT-RAM cell current with the least yield-driven cell
current choosing impact on memory reliability. By modifying the cell current at the
post-manufacture stage, the "ratio matching" technology can balance the process
variance in the voltage divider. NSRS can provide superior energy, performance,
and yield trade-off. These strategies significantly enhance chip yield by eliminating
sensing failures with low-power consumption and a high sense margin, according to
measurements on a 16Kb test chip [1] (Fig. 4).
20 S. Kumari and R. Yadav
The VDRS scheme increases sense margin in STT-RAM chips, resulting in higher
yield. The VDRS technique improves device tolerance while maintaining a low read
latency, high sense margin, and power consumption. In comparison to an NSRS
technique, the VDRS is more resilient (Fig. 5).
The main point of DMSS is to identify the exception state without any operational
or energy consumption in the normal mode. The data-cell-variation-tolerant DMSS
mostly operates in normal mode when appropriate sensing is assured. Non-DCVT-SS
can operate in normal mode to achieve high performance and low energy consump-
tion. When correct sensing is uncertain these schemes operate in exception mode.
The DMSS can attain the target read yield by using the dual-mode strategy. The
target read yield was calculated using the D-SRSS as DCVT-SS in exception mode
[10].
The electrical current measurement of the device is the main aspect when deter-
mining the energy converter’s regulation and efficiency. The magnetoresistive-based
electrical current sensor is an optimal measurement approach due to the absence
of thermal losses, low-power consumption, and isolation. Several microfabricated
or commercial prototypes MR current sensors used in instrumentation or industrial
application offer different electrical properties.
The sensor based on MTJ consists of an MTJ, a gold covering layer right beneath
a bio-coating layer, and an isolation layer. The free ferromagnet of the top of the
MTJ stack is heated by a current pulse above its blocking temperature. During the
cool-down time, the top free magnet takes up its magnetization direction dependent
on the externally applied magnetic field.
During the sensing phase, to excite the superparamagnetic bio-labels a vertical
excitation is applied. The biosensor design can attain high sensitivity in three distinct
ways:
(1) In the time domain, the electronic readout operation and the magnetic sensing
technique are decoupled.
(2) The binary readout enhances its signal-to-noise ratio substantially.
(3) The sensor sensitivity can be decoupled by thermally assisted magnetic sensing
from the magnetic rigidity of the sensing layer.
To prevent damage to the bio-coating layer, the temperature should be below 50C
gold layer’s top surface (Table 1).
3 Conclusion
References
1 Introduction
© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 27
P. Singhal et al. (eds.), Recent Developments in Electrical and Electronics Engineering,
Lecture Notes in Electrical Engineering 979,
https://doi.org/10.1007/978-981-19-7993-4_3
28 C. J. Khare et al.
et al. [5] illustrated that modern controller is much faster than conventional integral
control and gives better results.
The proposed work introduced a new variation in classical PSO in terms of statis-
tical parameters called hybrid statistically tracked particle swarm optimization
(hybrid STPSO) to search global best value with fast convergence speed. In multi-
dimensional shifted, rotated problem, basic PSO may enmesh on local best value
and then no further changes are observed in velocity of particles and at this point
statistical parameter gives a new path to get global best value. This paper illustrates a
statistical-based comparison of six evolutionary algorithms: ant colony optimization
[6], chaotic PSO (CPSO) [7, 8], cuckoo search [9, 10], bacteria foraging optimization
(BFO) [11, 12], teaching learning-based optimization (TLBO) [13, 14], and hybrid
STPSO to find optimal PID controller gains of automatic generation control.
Modeling is the framework, which provide optimum design of any system, which
consists of all the essential parameters, which is used for performance enhancement
of that system and if a system is modeled out with mathematical equation then it
is called the problem formulation of a particular system. Here we are discussing
modeling and problem formulation of automatic and generation control system.
The main objectives of automatic generation control (AGC) are
i. To maintain the frequency within specified limits.
ii. To control tie-line power flows.
iii. Distribute the loads among the generating units of multi-area system.
In the present, workload-frequency dynamics are illustrated by developing a math-
ematical model of the interconnected reheat thermal system in the state variable form
and Fig. 1 shows the transfer function model for jth area representation of the reheat
thermal system.
• Effect of governor dead band
Fig. 1 Block diagram of jth area AGC implies dead band in the speed governing loop
The state space equation of jth area can be written in the following form:
[ ]
Ẋ j = [A] j [X ] j + [U ] j , (1a)
where the value of a21 , a41 , U2 , and U4 depends whether the effect of dead band is
considered or not. For the linear system (dead band not considered) this values is
expressed as [2]
30 C. J. Khare et al.
1
a21 = − (2a)
R j Tg, j
K r, j
a41 = − . (2b)
R j Tg, j Tt, j
1
a21 = − (3a)
R j Tg, j
K r, j
a41 = −
R j Tg, j Tt, j
DB ( )
U2 = sign Δ f j (3b)
R j Tg, j
K r, j ( )
U4 = D Bsign Δ f j
R j Tg, j Tt, j
The proposed research work analyzed four-area systems with and without effect
of speed governing dead band of 0.06%.
Many academics have improved PSO algorithms in recent years, and while all of
these changes have helped to speed up convergence, a multidimensional problem
still requires a large number of iterations to get an optimal value. The current study
uses various statistical parameters as a tracker to get the best global value with
less iteration. To generate optimal AGC controller gains, a newly designed hybrid
statistically tracked particle swarm optimization (STPSO) method is applied in this
study. Basic PSO is a well-known evolutionary method with a random initialization.
This algorithm updates particle positions to find the best solution in a given search
space. Using the equations below, each particle sets its position and velocity [6]:
and
xmin and xmax : minimum and maximum values of particles in search space, and
vmin and vmax : minimum and maximum value velocity vectors.
The velocity and position of each (ith) particle are updated using following
equations:
( ) ( )
vik+1 = wvki + c1 r1 pbest ki − xik + c2 r2 gbest k − xik (5)
x k+1
i = xik + vik+1 , (6)
where r1 and r2 are two separate random numbers ranging from 0 to 1. The accel-
eration constants c1 and c2 help particles advance toward the best possible value
(gbestt ), which is set to 1.5 and “w” is the inertia weight used to balance between
best and best value. The weight of inertia changes with each iteration:
(wmax − wmin)
w = wmax − ∗ iter, (7)
iter max
where itermax is the maximum number of iterations; wmax and wmin, the upper and
lower limits of inertia weights. In the proposed paper, these values are 0.9 and 0.4
for each benchmark function. In complex multimodal issues, such an inertia weight
method may not provide satisfactory results. To update particle velocity, the proposed
hybrid STPSO employs the following modified relations:
( ) ( )
k+1
vnew,i = wvki + c1r1 pbest ki − xik + c2 r2 gbest k − xik + β k , (8)
∑
Xi
X sk =X= . (10)
N
The current study investigates hybrid STPSO by employing two statistical param-
eters to get an optimum value with increased variety. The mean value (X) of the
particle’s position is the first parameter, and the standard deviation () is the second.
When there is no improvement in the global best value of the basic PSO, statis-
tical tracking based on mean value is used, and when mean tracking becomes slow,
standard deviation tracking is used to update particle position.
k+1
Now, calculate fitness function value corresponding for xnew,i . Now particle is
selected for next iteration as
{ k+1 ( k+1 ) ( )}
xnew,i if f xnew,i ≤ f xik
xi =
k+1 ( ) ( k ) (13)
xik if f xik < f xnew,i
Step 7 If |f (gbest k )—f (gbest k+1 )|≤ ε then go to step-9. Here ε is a switch criteria
(set at 0.0001) to change tracks from basic PSO to mean.
Step 8 Increase the generation count k = k + 1, and go to step-4.
Step 9 Calculate mean of particle position
X and determine the acceleration factor using relations 10 and 9.
Step 10 Update velocity and position of each particle using relations 8, 12, and 13,
respectively.
Step 11 Calculate fitness function further for updated particles and obtain pbest,i k
and gbest k .
Step 12 If |f (gbest k )—f (gbest k+1 )|≤ ε then go to step 14.
Step 13 Increment the generation count k = k + 1 and go to step 9.
Step 14 Calculate standard deviation vector
Step 15 σ using relation 11.
Step 16 Update each particles using relations 12 and 13.
Step 17 Set k = k + 1. If k ≥ itermax then stop otherwise repeat from step 11.
In the proposed research, hybrid STPSO is implemented for four-area linear and
nonlinear AGC systems as shown in Fig. 2.
The goal of the AGC problem is to find a gain parameter that allows for desired
system performance such as lowest overshoot, undershoot, and settling time. To
attain the aforesaid objectives, the current study applied the following optimization
function [2]:
t ∑n ∑n−1 ( )2
fitness function = ∫ (Δ f j )2 + ΔPtie, j dt (14)
0 j=1 j=1
∑n−1
ΔPtie,n = − ΔPtie,i . (16)
i=1
Substituting this condition and modifying state matrix of Eq. (1a) as follows:
⎡ · ⎤T
Δ f , ΔPE,1 , ΔPT ,1 , ΔPT ,1 , ΔPc,1 , Δ f 2 , ΔPE,2 ,
⎢ 1 ⎥
⎢ · ⎥
⎢ ⎥
⎢ ΔPT,2 , ΔPT,2 , ΔPc,2 , Δ f 3 , PE,3 , ΔPT ,3 , ⎥
⎢
x =⎢ ⎥ (17)
⎥
⎢ · · ⎥
⎢ ΔP , ΔP , Δ f , ΔP , ΔP , ΔP , ⎥
⎣ T,3 c,3 4 E,4 T,4 T ,4 ⎦
ΔPc,4 , ΔPtie,1 , ΔPtie,2 , ΔPtie,3 , ΔPtie,4
In which ΔPtie,1 , ΔPtie,2 , ΔPtie,3 , and Δ Ptie,4 for four-area system can be
written as
Similarly equation for change in tie-line power for other areas can be written.
The performance of hybrid STPSO was compared to that of chaotic PSO (CPSO),
ant colony, bacterium foraging optimization, cuckoo search, and teaching learning-
based optimization (TLBO) algorithms in this section. For the sake of comparison,
100 runs were done for each technique. The proposed hybrid STPSO provides the
following benefits.
Convergence profile: Fig. 3 compares the convergence of the best fitness value
with the number of iterations for the four-area test system. The hybrid STPSO with
mean and standard deviation discovers the best value faster than the other approaches,
as seen in this graph.
0.014
1.00E-02 mean(f)_CPSO mean(f)_CPSO
0.012
8.00E-03 0.01 mean(f)_BFO
mean(f)_BFO
6.00E-03 0.008 mean(f)_Ant
mean(f)_Ant 0.006 colony
4.00E-03 colony
0.004 mean(f)_TLBO
2.00E-03 mean(f)_TLBO
0.002 mean(f)_cuckoo
0.00E+00 mean(f)_Cuckoo 0 search
search mean(f)_STPSO
14
27
40
53
66
79
92
14
27
40
53
66
79
92
1
(a) (b)
Fig. 3 Convergence profile for four-area (a linear, b nonlinear) AGC system of best mean fitness
function value with respect to iteration count
Optimum Design of Controller Parameters for Automatic Generation … 35
Table 1 a Parameter values for test system without considering dead band. b: Parameter values
for test system with dead band (nonlinear system)
Test system Algorithms Parameters Fitness function
(a) Parameter values for test system without considering dead band
Four-area AGC KI B
CPSO 0.32203068 0.9660046 0.0054784203
BFO 0.3515 0.8800 0.0054350498
Ant Colony 0.3451 0.9000 0.0054300418
TLBO 0.4239 0.7342 0.0054867217
Cuckoo Search 0.3580 0.8977 0.0054330442
Hybrid STPSO 0.3467 0.8993 0.0053300418
(b) Parameter values for test system with dead band (nonlinear system)
Four-area AGC KI B
CPSO 0.2746904 0.87208495 0.0088929749
BFO 0.2656 0.8086 0.0087297675
Ant Colony 0.2386 0.8949 0.008708383
TLBO 0.29568052 0.88709475 0.0088690639
Cuckoo Search 0.23619314 0.89047862 0.0087363357
Hybrid STPSO 0.23480 0.8106 0.0086047532
Dynamic performance evaluation of AGC test system: The current study eval-
uates the best controller for AGC of four-area systems, both with and without
nonlinearity, and implements various heuristics algorithms for the test system. When
compared to alternative evolutionary algorithms for similar systems, such as CPSO,
BFO, ant colony, TLBO, and cuckoo search optimization, the effectiveness of the
proposed hybrid STPSO is demonstrated. The integral controller is discussed in this
section of the text. The optimum values for test system corresponding to performance
indices considered in Eq. 14 are summarized in Table 1a, b.
Dynamic performance
Fig. 5 Dynamic performance evaluation of area control error of area 1 (e), area 4 (f) with respect
to time
search, change in tie-line power flow among various areas also gives better perfor-
mance as shown in Figs. 6 and 7, which depict that area control error of all areas
reach desired value faster with hybrid STPSO.
• Comparison based on statistical evaluation: Table 2 shows the results of
a four-area system’s autonomous generation control. In terms of minimum
fitness function, the statistical performance of hybrid STPSO is better than
that of other heuristic algorithms, as shown in the table (four-area linear:
CPSO-0.00547, BFO-0.00545, Ant Colony-0.00545, TLBO-0.00546, Cuckoo
search: 0.00543, STPSO-0.00534, four-area nonlinear: CPSO-0.00871, BFO-
0.00871, Ant Colony-0.00871, TLBO-0.00881, Cuckoo Search-0.00870, STPSO-
0.00861), mean of fitness function, maximum of fitness function, etc.
Optimum Design of Controller Parameters for Automatic Generation … 37
Fig. 6 Comparison of hybrid STPSO with other evolutionary algorithms for change in frequency
of area 1 (k), area 4 (l)
Fig. 7 Comparison of different heuristic algorithms for area control error of area 1(m), area
4(p) w.r.t time
5 Conclusion
A modified PSO algorithm was created in this study by introducing statistical factors
to get the global optimal value. Table 2 compares the proposed hybrid STPSO method
to existing optimization methodologies such as basic PSO and CPSO in terms of best
value, maximum value, standard deviation mean, and other statistical metrics. For
linear and nonlinear systems, the performance of hybrid STPSO is compared to BFO,
CPSO, ant colony, cuckoo search, and TLBO of autonomous generation control. In
terms of convergence profile and statistical parameters, experimental results show
that hybrid STPSO performs better for four-area test systems with and without dead
band. The proposed technique can successfully prevent the trapping on local best
value, according to numerical data.
38
Table 2 Statistical comparison of four-area (linear and nonlinear) AGC with confidence level 0.95
System Algorithm Mean (fbest) Median Std (fbest) (σ) Best value Max. value Freq. of Standard error Length of
(μ) (fbest) (fbest) (fbest) convergence of the mean confidence
fitness function
√ 2*C*ε
(ε) = C*σ/ n;
n = pop size; C
= 97.5th
percentile of
distribution
Four area CPSO 0.00548 0.00542 0.00014 0.00547 0.00585 78 0.0000247 0.0000481
(linear) BFO 0.00544 0.00544 0.00002 0.00545 0.00548 50 0.0000032 0.0000063
Ant Colony 0.00543 0.00543 0.00000 0.00545 0.00543 77 0.0000000 0.0000000
TLBO 0.00549 0.00549 0.00001 0.00546 0.00551 65 0.0000018 0.0000036
Cuckoo 0.00543 0.00543 0.00001 0.00543 0.00543 86 0.0000001 0.0000001
Search
PSO mean 0.00533 0.00535 0.00007 0.00534 0.00537 88 0.0000131 0.0000255
and std
(STPSO)
Four area CPSO 0.00889 0.00893 0.00014 0.00871 0.00913 48 0.00003 0.00005
(nonlinear) BFO 0.00873 0.00872 0.00003 0.00871 0.00886 62 0.00001 0.00001
Ant colony 0.00871 0.00871 0.00013 0.00871 0.00871 81 0.00002 0.00005
TLBO 0.00887 0.00885 0.00005 0.00881 0.00894 42 0.00001 0.00002
Cuckoo 0.00874 0.00874 0.00000 0.00870 0.00874 64 0.00000 0.00000
search
PSO mean 0.00860 0.00878 0.00010 0.00861 0.00868 83 0.00002 0.00003
and std
(STPSO)
C. J. Khare et al.
Optimum Design of Controller Parameters for Automatic Generation … 39
Appendix
System parameter for four area AGC system [2, 4̄]
k g,1 = k g,2 = k g,3 = k g,4 = 1,
Tg,1 = Tg,2 = Tg,3 = Tg,4 = 0.08,
kt,1 = kt,2 = kt,3 = kt,4 = 1,
Tt,1 = Tt,2 = Tt,3 = Tt,4 = 0.3,
K r,1 = K r,2 = K r,3 = K r,4 = 0.5,
Tr,1 = Tr,2 = Tr,3 = Tr,4 = 10,
K p,1 = K p,2 = K p,3 = K p,4 = 120,
T p,1 = T p,2 = T p,3 = T p,4 = 20,
β1 = β2 = β3 = β4 = 0.872,
R1 = R2 = R3 = R4 = 2.2568,
2π T1 = 2π T2 = 2π T3 = 2π T4 = 0.05.
References
1. Elgerd OI, Fosha C (1970) Optimum megawatt-frequency control of multiarea electric energy
systems. IEEE Trans Power Apparatus Syst PAS-89(4):556–563
2. Sinha N, Lai LL, Rao VG (2008) GA optimized PID controllers for automatic generation
control of two area reheat thermal system under deregulated environment. In: Proceedings of
the IEEE international conference on electric utilizes deregulation and restructuring and power
technologies, 6–9 April, pp 1186–1191
3. Raj H (2006) A distributed algorithm for an AGC in hydrothermal power plant. IFAC Proc Vol
39(7):1–6
4. Sahu R, Panda S, Raut UK (2013) DE optimized parallel 2-DOF PID controller for load
frequency control of power system with governor dead-band nonlinearity. Int J Electr Power
Energy Syst 49:19–33
40 C. J. Khare et al.
5. Jain D, Bhasker MK, Kumar M (2014) Comparative analysis of different methods of tuning the
PID controller parameters for load frequency control problem. Int J Adv Res Electr, Electron
Instrument Eng 3(11):13144–13154
6. Dorigo M, Blum C (2005) Ant colony optimization theory: a survey. Theoret Comput Sci
344:243–278
7. Nikabadi A, Ebadzadeh MM, Safabaksh R (2011) A novel particle swarm optimization with
adaptive inertia weight. Elsevier, Appl Soft Comput 11:8658–3670
8. Gao WF, Liu S-Y, Hauang L-L (2012) Particle swarm optimization with chaotic apposition
based population initialization and stochastic search technique. Elsevier 17:4316–4327
9. Payne B (2005) Robert. Oxford University Press, The Cuckoos
10. Rajabioun R (2011) Cuckoo optimization algorithm. Appl Soft Comput 11:5508–5518
11. Agrawal V, Sharma H, Bansal JC (2011) Bacterial foraging optimization: a survey. In: The
international conference on soft computing for problem solving (SOCPROS), pp 227–242
12. Ali ES, Abd-Elazim SM (2011) Bacteria foraging optimization algorithm based load frequency
controller for interconnected power system. Electr Power Energy Syst 33:633–638
13. Rao RV, Savsaniand VJ, Vakharia DP (2012) Teaching–learning-based optimization: an
optimization method for continuous non-linear large scale problems. Elsevier, Inform Sci
183:1–15
14. VenkataRao R, Patel V (2012) An elitist teaching-learning-based optimization algorithm for
solving complex constrained optimization problems. Int J Industr Eng Comput 3:535–560
Other documents randomly have
different content
Cela était insuffisant pour ses besoins d'âme. De pratique, elle aurait
eu assez, mais c'est la vie intérieure qui lui faisait défaut. L'action
d'un milieu indifférent est plus dissolvante que celle d'un milieu
hostile. Heurtées, les idées religieuses d'Élisabeth se fussent exaltées
par réaction naturelle. Ce qui même contribuait le plus efficacement
à maintenir sa foi, c'étaient les paroles de sarcasme et de
dénigrement qui sortaient de la bouche de certains amis de la
maison. Mais ces propos ne lui étaient point personnellement
adressés et nul jamais n'avait fait de tentative sur sa conscience.
Aussi, dans la liberté dédaigneuse qui lui était laissée, ne trouvait-
elle pas à puiser cette généreuse ferveur qu'inspire la persécution.
Et, au point de vue même le plus largement ésotérique, tous autour
d'elle étaient tellement étrangers à ces choses, que le sentiment
qu'elle en avait s'étiolait faute d'aliments. Nature un peu molle,
l'énergie lui manquait pour lutter contre son isolement moral. Puis
cela était bien abstrait, bien complexe pour sa jeunesse. En sorte
que, dans l'ambiance de matérialisme où elle vivait, les rites
finissaient par prendre pour elle ce caractère machinal qui en
amoindrit le sens divin. Fidèle à la lettre, Élisabeth sentait en elle se
refroidir l'esprit.
De ce desséchement de son être spirituel, elle souffrait. Sujette à
des réchauffements subits, comme celui qui s'était manifesté pour la
cérémonie nuptiale de sa cousine Jeanne, ce lui était l'occasion de
ces petites crises intimes auxquelles on attribuait faussement une
cause si concrète. Car elle ne s'en ouvrait à personne, sachant que
personne ne la comprendrait.
Sa tante toutefois ne se méprenait point en voyant dans ses élans de
ferveur l'ascendant de son ancienne compagne de couvent. Jamais
elles ne s'étaient perdues de vue. Plusieurs étés de suite, Élisabeth
avait passé un mois ou deux chez une sœur de sa mère qui habitait
un petit manoir sur la rivière de Morlaix. Monique Le Huédé était la
fille d'un officier supérieur de la marine retiré dans le voisinage très
proche. Déjà, lorsque les deux jeunes filles avaient été séparées par
les circonstances, chez l'aînée s'était déclaré ce penchant pour la vie
religieuse assez commun aux alentours de la seizième année. Le
temps semblait l'avoir confirmé, quoique, soumise à ses directeurs
spirituels, Monique n'en parlât guère, la prudence ecclésiastique lui
ayant imposé l'essai de la vie du siècle avant qu'elle envisageât
sérieusement l'éventualité d'une prise de voile. En attendant, elle
s'adonnait aux pratiques de la piété la plus exaltée. Moralement
dépaysée comme l'était, avenue de Messine, la nièce du grand
chirurgien athée, sur cette terre de Bretagne si intensément
catholique, où Élisabeth se retrempait dans le sang de sa race
maternelle, il lui semblait retrouver son équilibre rompu. Ce n'était
pas encore tout à fait cela néanmoins. La dévotion de Monique était
trop rigide, trop exaltée aussi pour satisfaire complètement à ses
propres tendances la portant vers une foi simple, calme, douce.
Aussi ne suivait-elle pas son amie jusqu'au bout du chemin où celle-
ci, dévorée d'un zèle d'apôtre qu'exagérait l'absolutisme de l'extrême
jeunesse, s'employait ardemment à l'entraîner sur ses pas.
N'empêche que les impressions subies laissaient leur empreinte et,
rentrée à Paris, jusqu'à ce que l'eussent effacée des contacts si
différents, Élisabeth ressentait le malaise créé par cette dualité
morale.
Sans qu'aucune confidence épistolaire eût fait prévoir l'événement,
comme Monique venait d'atteindre sa vingtième année elle se maria.
Ayant eu dans leur famille la douloureuse aventure d'une fausse
vocation au lendemain amer, ses parents l'avaient voulu détourner
du cloître. Soit qu'à ce moment une détente se fût produite en elle,
soit que la personne du prétendant qu'on lui fit connaître eût
triomphé de son éloignement réel ou factice pour les fins normales
des filles, elle consentit à épouser l'aimable et galant homme qu'était
Alain Guivarch. Lui, las de la vie de garçon assez joyeusement
menée, s'était volontiers épris de cette jolie personne un peu
austère et frigide, mais qu'un mari aurait d'autant plus de mérite et
d'agrément sans doute à conquérir à l'amour. Quittant le
commissariat de marine, il venait d'entrer dans les bureaux de la
Compagnie Transatlantique à Saint-Nazaire. Puis un emploi supérieur
l'appela à l'administration centrale. Ainsi furent plus étroitement
rapprochées les amies d'enfance. Leur intimité s'était encore accrue
du fait qu'Élisabeth ayant perdu sa tante de Bretagne, ce deuil
l'avait, six mois durant, tenue éloignée du train mondain, et la
famille Bertereau ne le portant pas, un peu isolée dans la maison. A
ce moment, Monique se trouvait séparée de son mari, en voyage de
service aux ports des Antilles. Elle vivait très retirée. Il allait de soi
que la jeune fille fréquentât beaucoup chez elle, et d'autant plus que
celle-ci n'allait qu'à contre-cœur avenue de Messine, où tout lui était
sujet de scandale. Car le mariage n'avait guère adouci cette âpre
dévotion qui faisait d'elle—comme le remarquait le docteur
Bertereau—une religieuse égarée dans le monde, avec plus
d'intransigeance que n'en ont d'ordinaire les saintes filles nourries
d'esprit de charité. Ses occupations conjugales et maternelles—un
fils lui était né—l'empêchaient de donner autant de soi
qu'auparavant aux pratiques pieuses, mais sans qu'eût fléchi la
rigidité de sa religion. M. Guivarch aurait souhaité chez sa femme
une foi plus amène. De trouver son foyer si morose, il commençait à
le délaisser. Monique s'en affligeait. En humanisant son austérité, il
n'eût tenu qu'à elle de retenir ce mari un peu léger, encore épris
pourtant. Mais, scrupuleuse observatrice de tous les devoirs de
l'épouse chrétienne, il en est un qu'elle ne savait pas remplir: celui
de se faire aimer. Son humeur allait s'assombrissant de ce chagrin
dont elle était l'artisan, et cela rejaillissait en sévérités envers le
prochain. Élisabeth était si tendrement attachée aux parents très
bons à qui elle devait d'oublier la tristesse d'être orpheline, qu'il lui
eût déplu d'entendre sur eux des paroles de blâme. Aussi se taisait-
elle avec son amie du malaise spirituel qui parfois l'oppressait. Et
puis n'était-elle point assez heureuse par ailleurs pour éviter de s'y
attarder, troublant ainsi la joie de vivre la vie douce et tiède de ses
vingt ans en fleur?
III
C'était maison ouverte chez les Bertereau. Pour paysan du Danube
que fût le grand chirurgien, dans ses façons rudes comme dans sa
massive personne, il n'ignorait pas l'art de mettre en valeur le
mérite. Il savait notamment que les échos sous la rubrique
«Mondanités» servent mieux que les comptes rendus de l'Académie
de médecine à entretenir la réputation d'un praticien. Tout le monde
lit ceux-là, et qui donc jette un œil sur ceux-ci? Puis l'eau va à la
rivière; aussi, dans toute profession, montrer qu'on gagne beaucoup
d'argent est le moyen d'en gagner davantage. Et c'était bals,
matinées de musique, soirées de comédie et de tableaux vivants,
très somptueuses fêtes où se pressait «l'aristocratie républicaine»,
de compagnie avec le monde médical et scientifique, ainsi que le
personnel bigarré et cosmopolite de la clientèle, sans omettre les
sommités de la presse, cette puissance que se doit concilier
quiconque vit du public et a besoin de réclame.
On dînait aussi beaucoup, avenue de Messine, forme de réception
préférée du docteur, lequel, matineux par goût comme par obligation
de métier, ne se retirait jamais passé minuit, même lorsque ses
lustres demeuraient allumés bien plus tard. Le dîner dominical était
réservé à la famille, augmentée de quelques intimes, parfois d'un ou
deux de ses élèves favoris. Le mariage de sa fille cadette ne l'avait
pas éloignée de cette réunion hebdomadaire, la proximité de
Beauvais permettant au jeune ménage de venir souvent à Paris. Non
que Jeanne se déplût dans sa résidence provinciale. Naïvement
pénétrée de l'importance de son nouveau personnage, par cette
royauté sur le personnel médiocre qui hante les salons préfectoraux
elle se sentait relevée de l'état insignifiant dont le sentiment l'avait
faite si timide. Mais c'était pour Gaston. Tellement Parisien, ce
Bordelais débarqué au «Quartier» pour y faire son droit, quelque
quinze ans plus tôt, s'étant poussé par le bagou et l'intrigue dans
des cabinets de ministre, d'où le testament de son dernier patron lui
avait ouvert la bonne porte de la carrière administrative. La paisible
cité bellovaque lui semblait un fâcheux exil, et l'air du boulevard lui
était indispensable, assurait-il, pour se retremper périodiquement
dans le mouvement d'esprit. Il s'y retrempait si bien que d'ordinaire,
un instant avant qu'on se mît à table chez son beau-père, un petit
bleu arrivait, l'excusant sur quelque rencontre faite au Cercle
National, un homme politique avec qui il avait intérêt à causer... Il
viendrait dans la soirée, le plus tôt possible... Jeanne en soupirait un
peu. Elle aimait son mari. Elle tirait quelque vanité aussi, elle, le
laideron de la famille, de ce joli garçon bien tourné, aux yeux
luisants, aux lèvres vermeilles dans la fine barbe noire, dont elle
prenait pour de l'esprit la faconde gasconne mâtinée de blague
boulevardière, comme sa suffisance lui semblait du mérite, et ses
allures faraudes habillées par un bon tailleur lui donnaient l'illusion
de la distinction et de l'élégance. Elle était moins riche qu'Hélène;
mais ainsi avait-elle l'avantage sur le gros homme lourd et vulgaire,
rougeaud et rageur, quoique bon diable au demeurant, qu'était son
beau-frère Gustave Percheron.
Ce dimanche-là, sur le coup de huit heures, fut apporté le
pneumatique. On ne se trouvait pas encore au complet. Le retard
était usuel dans cette famille, où les principes d'indépendance sur
lesquels se fondait l'éducation prenaient volontiers la forme du
désordre. Jeanne seulement et Élisabeth, qui étaient allées à un
concert classique, avaient achevé l'après-midi auprès de Mme
Bertereau, l'aidant à écrire des invitations pour un raout prochain.
Les premiers arrivèrent M. et Mme Biscaras, habituels commensaux
du dimanche. Puis le maître du logis, revenant de Saint-Germain, où
il suivait la lente et atroce agonie d'un premier ministre déchu, rongé
d'un cancer au foie, et chez qui la décomposition du sang s'aggravait
des rancœurs que donne le sentiment d'une vie mal vécue, sans
avoir recueilli le fruit de ses compromissions, de ses vilenies, de ses
lâchetés, triste exemple de la faillite d'une intelligence supérieure
s'accordant avec une conscience amorphe et un caractère dénué de
noblesse.
«Il est perdu, dit le docteur. Je me suis nettement prononcé contre
l'opération. Il mourrait sous le bistouri. On appellera, si on veut, de
Berlin ou de Londres, Vogel ou Mackay. Moi, je ne commettrai pas
un assassinat.
—Dix mille francs de manque à gagner, gouailla le normalien qui
entrait. Qu'est-ce que cela peut te faire, papa? Ce serait même une
bonne action, puisqu'il cesserait de souffrir.»
Le chirurgien fronça ses gros sourcils.
«Et la probité professionnelle, qu'en fais-tu?»
Marcel eut ce geste léger et ironique qu'il opposait aux mots à ses
yeux démodés.
«Ce sera un grand deuil pour la démocratie, soupira Alcide Biscaras,
dont la maigre figure aux lignes aiguës, qu'allongeait encore la
pointe d'une barbiche poivre et sel et qu'éclairaient des yeux de jais,
vifs et perçants comme des vrilles, présentait bien les traits
physionomiques du jacobin buveur d'eau.
«La démocratie? riposta Marcel... Ce qu'elle s'en fiche de votre grand
homme! Qu'a-t-il fait pour elle, je vous le demande?
—Il a sauvé la république, tout simplement. Nieras-tu que la
coalition des forces cléricales et réactionnaires ait les reins cassés?
—Diable! il fallait qu'elle fût bien branlante, votre république,
remarqua gaiement un nouveau venu. Car, en vérité, la fameuse
alliance du sabre et du goupillon... pistolet de paille et sabre de bois,
monsieur Biscaras... Croyez-moi, puisque j'en suis.»
Cette remarque du capitaine Maurice Briffault ne sembla point au
vieux sectaire digne d'être relevée. Qu'attendre du raisonnement
d'un soldat? Le normalien, d'ailleurs, avec son ton mordant et
dédaigneux de raffiné, appuyait aussitôt l'offensive blagueuse de son
cousin.
«Aussi faudrait-il démontrer que république est synonyme de
démocratie. Vous y auriez de la peine. Son analogie avec
ploutocratie s'impose davantage. Votre grand manitou a défendu le
râtelier unguibus et rostro... Geste rempli d'intérêt pour ceux qui y
mangent. Mais a-t-il avancé d'une ligne l'évolution vers l'humanité
intégrale? Y a-t-il dans la masse populaire une once de moins
d'asservissement moral ou de misère matérielle?»
IV
Cinq ans plus tard, sur l'impassible registre de l'état civil de la mairie
du huitième arrondissement, en marge de l'acte de mariage de M. et
Mme Edmond Lambertier était inscrit leur divorce.
Ce naufrage du bonheur qu'il s'était réjoui de donner à sa nièce fut
pour l'optimisme du docteur Bertereau une cruelle déconvenue. C'est
qu'il avait ignoré—ne l'ayant pas assez cherché peut-être—le
véritable mobile de la recherche dont il s'émerveillait. Loin que le
jeune viveur eût cédé à un entraînement auquel n'était pas
accessible sa nature positive et passablement grossière, c'est d'un
esprit très réfléchi au contraire que, sans mettre de visage sur cette
détermination, il avait résolu de prendre femme. Menacé par sa
mère d'un conseil judiciaire que justifiaient amplement ses
extravagances, souhaitant d'autre part complaire à un oncle de qui il
attendait une grosse succession—le tiers des parts du charbonnage
—et qui voulait des petits-neveux, en enrayant temporairement sa
vie de désordres il pacifiait les siens, et à chaque jour sa peine.
N'ayant point à se préoccuper de la dot, il tenait seulement à la
beauté. Plus encore que pour son agrément, c'était par vanité, par
habitude de se payer ce qui se fait de mieux en tous genres. Donc, il
s'était mis à regarder les jeunes filles, variété féminine jusqu'alors
totalement dédaignée par lui. Le destin aussitôt avait mis celle-ci sur
É
sa route. Élisabeth était assez jolie pour lui plaire. Mais davantage
est-ce le charme de la pureté, en elle si profond et si vif, qui avait
fixé son choix, non moins rapide que ses fantaisies libertines.
Impulsif, en effet, volontaire extrêmement, chez lui il n'y avait guère
de place entre un désir et sa satisfaction.
Our website is not just a platform for buying books, but a bridge
connecting readers to the timeless values of culture and wisdom. With
an elegant, user-friendly interface and an intelligent search system,
we are committed to providing a quick and convenient shopping
experience. Additionally, our special promotions and home delivery
services ensure that you save time and fully enjoy the joy of reading.
ebookball.com