Professional Documents
Culture Documents
Textbook Neural Information Processing 25Th International Conference Iconip 2018 Siem Reap Cambodia December 13 16 2018 Proceedings Part Vii Long Cheng Ebook All Chapter PDF
Textbook Neural Information Processing 25Th International Conference Iconip 2018 Siem Reap Cambodia December 13 16 2018 Proceedings Part Vii Long Cheng Ebook All Chapter PDF
https://textbookfull.com/product/neural-information-
processing-25th-international-conference-iconip-2018-siem-reap-
cambodia-december-13-16-2018-proceedings-part-vii-long-cheng/
https://textbookfull.com/product/neural-information-
processing-25th-international-conference-iconip-2018-siem-reap-
cambodia-december-13-16-2018-proceedings-part-ii-long-cheng/
https://textbookfull.com/product/neural-information-
processing-25th-international-conference-iconip-2018-siem-reap-
cambodia-december-13-16-2018-proceedings-part-iii-long-cheng/
https://textbookfull.com/product/neural-information-
processing-25th-international-conference-iconip-2018-siem-reap-
cambodia-december-13-16-2018-proceedings-part-v-long-cheng/
Neural Information Processing: 25th International
Conference, ICONIP 2018, Siem Reap, Cambodia, December
13–16, 2018, Proceedings, Part VI Long Cheng
https://textbookfull.com/product/neural-information-
processing-25th-international-conference-iconip-2018-siem-reap-
cambodia-december-13-16-2018-proceedings-part-vi-long-cheng/
https://textbookfull.com/product/neural-information-
processing-27th-international-conference-iconip-2020-bangkok-
thailand-november-18-22-2020-proceedings-part-iv-haiqin-yang/
https://textbookfull.com/product/neural-information-
processing-27th-international-conference-iconip-2020-bangkok-
thailand-november-18-22-2020-proceedings-part-v-haiqin-yang/
https://textbookfull.com/product/neural-information-
processing-27th-international-conference-iconip-2020-bangkok-
thailand-november-23-27-2020-proceedings-part-i-haiqin-yang/
https://textbookfull.com/product/neural-information-
processing-27th-international-conference-iconip-2020-bangkok-
thailand-november-23-27-2020-proceedings-part-ii-haiqin-yang/
Long Cheng
Andrew Chi Sing Leung
Seiichi Ozawa (Eds.)
LNCS 11307
Neural
Information Processing
25th International Conference, ICONIP 2018
Siem Reap, Cambodia, December 13–16, 2018
Proceedings, Part VII
123
Lecture Notes in Computer Science 11307
Commenced Publication in 1973
Founding and Former Series Editors:
Gerhard Goos, Juris Hartmanis, and Jan van Leeuwen
Editorial Board
David Hutchison
Lancaster University, Lancaster, UK
Takeo Kanade
Carnegie Mellon University, Pittsburgh, PA, USA
Josef Kittler
University of Surrey, Guildford, UK
Jon M. Kleinberg
Cornell University, Ithaca, NY, USA
Friedemann Mattern
ETH Zurich, Zurich, Switzerland
John C. Mitchell
Stanford University, Stanford, CA, USA
Moni Naor
Weizmann Institute of Science, Rehovot, Israel
C. Pandu Rangan
Indian Institute of Technology Madras, Chennai, India
Bernhard Steffen
TU Dortmund University, Dortmund, Germany
Demetri Terzopoulos
University of California, Los Angeles, CA, USA
Doug Tygar
University of California, Berkeley, CA, USA
Gerhard Weikum
Max Planck Institute for Informatics, Saarbrücken, Germany
More information about this series at http://www.springer.com/series/7407
Long Cheng Andrew Chi Sing Leung
•
Neural
Information Processing
25th International Conference, ICONIP 2018
Siem Reap, Cambodia, December 13–16, 2018
Proceedings, Part VII
123
Editors
Long Cheng Seiichi Ozawa
The Chinese Academy of Sciences Kobe University
Beijing, China Kobe, Japan
Andrew Chi Sing Leung
City University of Hong Kong
Kowloon, Hong Kong SAR, China
This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Preface
speakers, authors, reviewers, volunteers, and participants for their contribution and
support in making ICONIP 2018 a successful event.
General Chair
Jun Wang City University of Hong Kong,
Hong Kong SAR, China
Advisory Chairs
Akira Hirose University of Tokyo, Tokyo, Japan
Soo-Young Lee Korea Advanced Institute of Science and Technology,
South Korea
Derong Liu Institute of Automation, Chinese Academy of Sciences,
China
Nikhil R. Pal Indian Statistics Institute, India
Program Chairs
Long Cheng Institute of Automation, Chinese Academy of Sciences,
China
Andrew C. S. Leung City University of Hong Kong, Hong Kong SAR,
China
Seiichi Ozawa Kobe University, Japan
Tutorial Chairs
Hiroaki Gomi NTT Communication Science Laboratories, Japan
Takashi Morie Kyushu Institute of Technology, Japan
Kay Chen Tan City University of Hong Kong, Hong Kong SAR,
China
Dongbin Zhao Institute of Automation, Chinese Academy of Sciences,
China
VIII ICONIP 2018 Organization
Publicity Chairs
Zeng-Guang Hou Institute of Automation, Chinese Academy of Sciences,
China
Tingwen Huang Texas A&M University at Qatar, Qatar
Chia-Feng Juang National Chung-Hsing University, Taiwan
Tomohiro Shibata Kyushu Institute of Technology, Japan
Publication Chairs
Xinyi Le Shanghai Jiao Tong University, China
Sitian Qin Harbin Institute of Technology Weihai, China
Zheng Yan University Technology Sydney, Australia
Shaofu Yang Southeast University, China
Registration Chairs
Shenshen Gu Shanghai University, China
Qingshan Liu Southeast University, China
Ka Chun Wong City University of Hong Kong,
Hong Kong SAR, China
Conference Secretariat
Ying Qu Dalian University of Technology, China
Program Committee
Hussein Abbass University of New South Wales at Canberra, Australia
Choon Ki Ahn Korea University, South Korea
Igor Aizenberg Texas A&M University at Texarkana, USA
Shotaro Akaho National Institute of Advanced Industrial Science
and Technology, Japan
Abdulrazak Alhababi UNIMAS, Malaysia
Cecilio Angulo Universitat Politècnica de Catalunya, Spain
Sabri Arik Istanbul University, Turkey
Mubasher Baig National University of Computer and Emerging
Sciences Lahore, India
Sang-Woo Ban Dongguk University, South Korea
Tao Ban National Institute of Information and Communications
Technology, Japan
Boris Bačić Auckland University of Technology, New Zealand
Xu Bin Northwestern Polytechnical University, China
David Bong Universiti Malaysia Sarawak, Malaysia
Salim Bouzerdoum University of Wollongong, Australia
Ivo Bukovsky Czech Technical University, Czech Republic
ICONIP 2018 Organization IX
Potential and Sampling Based RRT Star for Real-Time Dynamic Motion
Planning Accounting for Momentum in Cost Function. . . . . . . . . . . . . . . . . 209
Saurabh Agarwal, Ashish Kumar Gaurav, Mehul Kumar Nirala,
and Sayan Sinha
Biomedical Applications
Hardware
NVM Weight Variation Impact on Analog Spiking Neural Network Chip . . . 676
Akiyo Nomura, Megumi Ito, Atsuya Okazaki, Masatoshi Ishii,
Sangbum Kim, Junka Okazawa, Kohji Hosokawa, and Wilfried Haensch
Yuqi Liu, Haojie Li, Wenjie Wu, Dan Wang, and Zhouhua Peng(B)
1 Introduction
In recent years, double-fed induction generator has been widely used as the core
component of variable speed constant frequency wind power generation system
[1]. Compared with synchronous generator and fixed speed asynchronous genera-
tor, double-fed induction generator has many advantages, such as variable speed
operation, four quadrant active power, reactive power regulation, low converter
cost and power loss [2].
D. Wang—This work was supported in part by the National Natural Science Founda-
tion of China under Grants 61673081, 51579023, and in part by the Innovative Talents
in Universities of Liaoning Province under Grant LR2017014, and in part by High Level
Talent Innovation and Entrepreneurship Program of Dalian under Grant 2016RQ036,
and in part by the National Key Research and Development Program of China under
Grant 2016YFC0301500, and in part by the Fundamental Research Funds for the Cen-
tral Universities under Grants 3132016313, 3132018306.
c Springer Nature Switzerland AG 2018
L. Cheng et al. (Eds.): ICONIP 2018, LNCS 11307, pp. 3–14, 2018.
https://doi.org/10.1007/978-3-030-04239-4_1
4 Y. Liu et al.
As the prime mover of wind power generation system, wind turbine is the pri-
mary component to convert the wind energy to mechanical energy in the entire
system. The performance of wind turbine determines the effective output power
of the wind power system and the safe, stable and reliable operation of wind tur-
bine. From the aerodynamic analysis, the mechanical power Pa and aerodynamic
torque Ta captured by the wind turbine from the wind are:
1 Pa 1 Cp (λ, β)
Pa = ρπR2 Cp (λ, β)v 3 , Ta = = ρπR2 v 3 (1)
2 ωr 2 ωr
where ρ is the air density, R is the radius of the rotor, Cp (λ, β) is the wind
energy utilization factor, v is the wind speed, and ωr is the rotation speed of the
wind wheel. The value of Cp (λ, β) depends on the pitch angle β and tip speed
ratio λ. According to Betz theory, the theoretical maximum value of Cp (λ, β) is
0.593, which is also known as the Betz limit. For horizontal axis wind turbines,
the value of Cp is generally less than 0.5. The tip speed ratio is expressed as:
Prescribed Performance Control 5
where urd , urq , usd and usq are the rotor and the stator d-axis and q-axis voltage
components, respectively. ird , irq , isd and isq are the rotor and the stator d-axis
and q-axis current components, respectively. φrd , φrq , φsd and φsq are the rotor
and the stator d-axis and q-axis flux components, respectively. Rr and Rs are
the rotor and the stator winding resistances. Lr and Ls are the self-inductances
of the rotor winding and the stator winding in a synchronous rotating coordinate
system. np is the number of pole pairs. Lm is the equivalent mutual inductance
in the stator windings in a synchronous rotating coordinate system.
According to the stator voltage orientation rule, the d-axis of the synchronous
rotational coordinate system is aligned with the stator voltage vector Us . Neglect-
ing the influence of stator resistance, the voltage equation is given by
urd = Rr ird + σLr i̇rd − ωsl (− ωL1m
Ls Us + σLr irq ) (4)
urq = Rr irq + σLr i̇rd + σLr ωsl ird
3 Problem Formulation
In different operating areas, the control objectives of wind power generators are
different. The wind turbines running at low wind speed are focused in this paper.
The control objective is to make the wind power generator run in the vicinity of
the maximum power point. When the wind power generator runs at the maximum
value of Cp , the reference value of the electromagnetic torque is [2]:
πρ5 Cpmax
Tref = kopt ωg2 , kopt = . (6)
2n3g λ3opt
The stator reactive power needs to be controlled to satisfy the grid require-
ments. When wind power plants do not require wind turbines to generate reactive
power, the DFIG can run under the unit power factor, i.e., Q = 0. From (5), the
reference value of the q-axis component of the rotor current can be calculated
Us
irq ref =− . (7)
ω1 Lm
In actual wind power system, electrical subsystems respond more quickly
than mechanical subsystems. Therefore, the controllers of generator and wind
power generator are designed separately. A double closed loop control structure
is used, in which the inner loop is concerned with the electrical subsystem and the
outer loop is concerned with the mechanical subsystem. The outer loop provides
reference torque input Tref for the inner loop. When designing the outer loop
controller, it is usually assumed that the inner loop of the electrical subsystem
can track the reference torque Tref in real time [7–12].
For the convenience of analysis, the electromagnetic torque tracking error
eTem and the reactive current tracking error eirq are defined as:
In this paper, the transient and stability performance of the controller are
considered. Therefore, the control target is described in detail as follows. By
controlling the dq-axis components of the rotor voltage urd , urq , we can achieve:
(1) Electromagnetic torque Tem tracks its reference signal Tref to achieve max-
imum power tracking;
(2) Reactive current irq tracks its reference signal irq ref to achieve unit power
factor operation;
(3) The system tracking error satisfies the prescribed transient and steady state
control performance.
eTem < eTem < ēTem , eirq < eirq < ēirq (9)
Prescribed Performance Control 7
where ēTem and eTem are the upper and lower bounds of torque tracking error
eTem respectively, and satisfy eTem (t) < 0, ēTem (t) > 0. ēirq and eirq are the upper
and lower bounds of current tracking error eirq respectively, and satisfy eirq (t) <
0, ēirq (t) > 0. By preselecting the upper and lower bound functions of the
appropriate tracking error, the pre-defined transient and stability performance
can be achieved. Taking the torque tracking error eTem as an example, select the
initial value eTem (0) and ēTem (0). The overshoot must be between eTem (0) and
ēTem (0). The shapes of eTem (t) and ēTem (t) determine the convergence rate of the
tracking error in the transient process. The steady-state error signal eTem (∞) is
maintained in a prescribed error domain, i.e., eTem (∞) < eTem (∞) < ēTem (∞).
For the design of the controller, the following assumption is needed.
Assumption 1. The upper and lower bounds of tracking error ēTem , eTem , ēirq ,
eirq and their derivatives are known and can be used directly in the controller
design.
o quente da cama
Com Gonçalo, e com sua ama,
Dizendo estava comei-me, &.
Por vida do meu Gonçalo,
Custodia formosa e linda, &.
Madrasta do Gonçalinho,
Que é lindo enteado a fé, &.
Sim, por vida de Gonçalo, &.
Mas por vida de Gonçalo, &.
GLOSA.
AOS VICIOS
TERCETOS