Methods and Applications for Modeling

and Simulation of Complex Systems

18th Asia Simulation Conference
AsiaSim 2018 Kyoto Japan October 27
29 2018 Proceedings Liang Li
Visit to download the full and correct content document:
https://textbookfull.com/product/methods-and-applications-for-modeling-and-simulatio
n-of-complex-systems-18th-asia-simulation-conference-asiasim-2018-kyoto-japan-oct
ober-27-29-2018-proceedings-liang-li/
More products digital (pdf, epub, mobi) instant
download maybe you interests ...

Methods and Applications for Modeling and Simulation of

Complex Systems 19th Asia Simulation Conference AsiaSim
2019 Singapore October 30 November 1 2019 Proceedings
Gary Tan
https://textbookfull.com/product/methods-and-applications-for-
modeling-and-simulation-of-complex-systems-19th-asia-simulation-
conference-asiasim-2019-singapore-
october-30-november-1-2019-proceedings-gary-tan/

PRIMA 2018 Principles and Practice of Multi Agent

Systems 21st International Conference Tokyo Japan
October 29 November 2 2018 Proceedings Tim Miller

https://textbookfull.com/product/prima-2018-principles-and-
practice-of-multi-agent-systems-21st-international-conference-
tokyo-japan-october-29-november-2-2018-proceedings-tim-miller/

Simulation and Modeling Methodologies Technologies and

Applications 8th International Conference SIMULTECH
2018 Porto Portugal July 29 31 2018 Revised Selected
Papers Mohammad S. Obaidat
https://textbookfull.com/product/simulation-and-modeling-
methodologies-technologies-and-applications-8th-international-
conference-simultech-2018-porto-portugal-july-29-31-2018-revised-
selected-papers-mohammad-s-obaidat/

Internet of Things Smart Spaces and Next Generation

Networks and Systems 18th International Conference
NEW2AN 2018 and 11th Conference ruSMART 2018 St
Petersburg Russia August 27 29 2018 Proceedings Olga
Galinina
https://textbookfull.com/product/internet-of-things-smart-spaces-
and-next-generation-networks-and-systems-18th-international-
conference-new2an-2018-and-11th-conference-rusmart-2018-st-
System Analysis and Modeling Languages Methods and
Tools for Systems Engineering 10th International
Conference SAM 2018 Copenhagen Denmark October 15 16
2018 Proceedings Ferhat Khendek
https://textbookfull.com/product/system-analysis-and-modeling-
languages-methods-and-tools-for-systems-engineering-10th-
international-conference-sam-2018-copenhagen-denmark-
october-15-16-2018-proceedings-ferhat-khendek/

Computer Information Systems and Industrial Management:

17th International Conference, CISIM 2018, Olomouc,
Czech Republic, September 27-29, 2018, Proceedings
Khalid Saeed
https://textbookfull.com/product/computer-information-systems-
and-industrial-management-17th-international-conference-
cisim-2018-olomouc-czech-republic-
september-27-29-2018-proceedings-khalid-saeed/

Complex Systems Design Management Asia Smart

Transportation Proceedings of the Third Asia Pacific
Conference on Complex Systems Design Management CSD M
Asia 2018 Michel Alexandre Cardin
https://textbookfull.com/product/complex-systems-design-
management-asia-smart-transportation-proceedings-of-the-third-
asia-pacific-conference-on-complex-systems-design-management-csd-
m-asia-2018-michel-alexandre-cardin/

Information Retrieval: 24th China Conference, CCIR

2018, Guilin, China, September 27–29, 2018, Proceedings
Shichao Zhang

https://textbookfull.com/product/information-retrieval-24th-
china-conference-ccir-2018-guilin-china-
september-27-29-2018-proceedings-shichao-zhang/

Knowledge and Systems Sciences 19th International

Symposium KSS 2018 Tokyo Japan November 25 27 2018
Proceedings Jian Chen

https://textbookfull.com/product/knowledge-and-systems-
sciences-19th-international-symposium-kss-2018-tokyo-japan-
november-25-27-2018-proceedings-jian-chen/
Liang Li
Kyoko Hasegawa
Satoshi Tanaka (Eds.)

Communications in Computer and Information Science 946

Methods and Applications

for Modeling and Simulation
of Complex Systems
18th Asia Simulation Conference, AsiaSim 2018
Kyoto, Japan, October 27–29, 2018
Proceedings

123
Communications
in Computer and Information Science 946
Commenced Publication in 2007
Founding and Former Series Editors:
Phoebe Chen, Alfredo Cuzzocrea, Xiaoyong Du, Orhun Kara, Ting Liu,
Dominik Ślęzak, and Xiaokang Yang

Editorial Board
Simone Diniz Junqueira Barbosa
Pontifical Catholic University of Rio de Janeiro (PUC-Rio),
Rio de Janeiro, Brazil
Joaquim Filipe
Polytechnic Institute of Setúbal, Setúbal, Portugal
Igor Kotenko
St. Petersburg Institute for Informatics and Automation of the Russian
Academy of Sciences, St. Petersburg, Russia
Krishna M. Sivalingam
Indian Institute of Technology Madras, Chennai, India
Takashi Washio
Osaka University, Osaka, Japan
Junsong Yuan
University at Buffalo, The State University of New York, Buffalo, USA
Lizhu Zhou
Tsinghua University, Beijing, China
More information about this series at http://www.springer.com/series/7899
Liang Li Kyoko Hasegawa
•

Satoshi Tanaka (Eds.)

Methods and Applications

for Modeling and Simulation
of Complex Systems
18th Asia Simulation Conference, AsiaSim 2018
Kyoto, Japan, October 27–29, 2018
Proceedings

123
Editors
Liang Li Satoshi Tanaka
Ritsumeikan University Ritsumeikan University
Kusatsu, Shiga, Japan Kusatsu, Shiga, Japan
Kyoko Hasegawa
Ritsumeikan University
Kusatsu, Shiga, Japan

ISSN 1865-0929 ISSN 1865-0937 (electronic)

Communications in Computer and Information Science
ISBN 978-981-13-2852-7 ISBN 978-981-13-2853-4 (eBook)
https://doi.org/10.1007/978-981-13-2853-4

Library of Congress Control Number: 2018957409

© Springer Nature Singapore Pte Ltd. 2018

This work is subject to copyright. All rights are reserved 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, express 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
 Preface

The Asia Simulation Conference (AsiaSim) is the international conference that has the
longest history as a conference of modeling and simulation in Asia. It is an annual
conference organized by the Federation of Asia Simulation Societies (ASIASIM),
whose current member societies are: CSF (China Simulation Federation), JSST (Japan
Society for Simulation Technology), KSS (Korea Society for Simulation), SSAGsg
(Society of Simulation and Gaming of Singapore), and MSS (Malaysian Simulation
Society). This conference provides a forum for scientists and engineers from around the
world to promote the advancement of modeling and simulation in academic and
industrial communities.
AsiaSim 2018 was held in Kyoto, Japan. We received about 100 full papers.
Submissions came from China, Japan, Korea, Singapore, Malaysia, Columbia, and
Italy. After an intensive review process by the internationally assembled Program
Committee, where each paper was reviewed by multiple reviewers, we finally accepted
45 full papers. Due to the high quality of the submitted papers, the paper selection was
very difficult and we were forced to reject many interesting papers. The accepted papers
are now consolidated in this volume of the Communications in Computer and Infor-
mation Science (CCIS) series of Springer, and are divided into relevant topics. The
diversity of topics is a unique and important feature of the AsiaSim conference. Giving
researchers of different fields opportunities to get together and exchange ideas has
inspired many interesting research activities. We hope the publication of this volume
will further promote this nice feature of the AsiaSim conference.
We thank the members of the Program Committee for their valuable effort in
reviewing submitted papers. We also thank the Organizing Committee, which sup-
ported our editorial work in various aspects. We also express our special thanks to the
College of Information Science and Engineering and ICT Medical Healthcare Center of
Ritsumeikan University, the co-sponsor of the conference. Finally, we thank all the
authors and participants of AsiaSim 2018.

October 2018 Satoshi Tanaka

Kyoko Hasegawa
Liang Li
 AsiaSim 2018 Organization

General Chairs
Kazuo Furuta University of Tokyo, JSST President, Japan
Satoshi Tanaka Ritsumeikan University, Japan

Steering Chair
Kyoko Hasegawa Ritsumeikan University, Japan

Program Chairs
Liang Li Ritsumeikan University, Japan
Naohisa Sakamoto Kobe University, Japan

Technical Co-sponsors

China Simulation Federation (CSF)

Japanese Society for Simulation Technology (JSST)
Korea Society for Simulation (KSS)
Society of Simulation and Gaming of Singapore (SSAGsg)
Malaysian Simulation Society (MSS)
Society for Modeling and Simulation International (SCS)

Co-organizers

Japanese Society for Simulation Technology (JSST)

Federation of Asia Simulation Societies (ASIASIM)

Co-sponsors

College of Information Science and Engineering, Ritsumeikan University

ICT Medical Healthcare Center, Ritsumeikan University

International Program Committee (ASIASIM)

ASIASIM President
Gary Tan (SSAGsg President) School of Computing, National University
of Singapore
VIII AsiaSim 2018 Organization

ASIASIM Board Members

Bo Hu Li CSF board member
Zhang Lin CSF board member, Beihang University, China
Xiao Song CSF board member, Beihang University, China
Satoshi Tanaka JSST board member, Ritsumeikan University,
Japan
Kyoko Hasegawa JSST board member, Ritsumeikan University,
Japan
Liang Li JSST board member, Ritsumeikan University,
Japan
Yun-Bae Kim KSS board member, Sungkyun Kwan University,
South Korea
Kang Sun Lee KSS board member, Myongji University, South
Korea
Doo-Kwon Baik KSS board member, Korea University, South Korea
Gary Tan SSAGsg board member, NUS, Singapore
Teo Yong Meng SSAGsg board member, NUS, Singapore
Rubiyah Yusof Universiti Teknologi Malaysia
Yahaya Md.Sam Universiti Teknologi Malaysia
Axel Lehmann Honorary member of ASIASIM, Universität der
Bundeswehr München, Germany

Program Committee
Satoshi Tanaka Ritsumeikan University, Japan
Liang Li Ritsumeikan University, Japan
Kyoko Hasegawa Ritsumeikan University, Japan
Hiroshi Tamura Chuo University, Japan
Norifumi Yamada Fukui University, Japan
Akinori Kimura Ashikaga Institute of Technology, Japan
Hiroaki Nakamura National Institute for Fusion Science
Naohisa Sakamoto Kobe University, Japan
Yoshiyuki Miyamoto AIST, Japan
Yifa Tang Chinese Academy of Sciences, China
Taku Itoh Nihon University, Japan
Shigeru Shimamoto Waseda University, Japan
Soo-Hyun Park Kookmin University, South Korea
Shin Muroya Matsumoto University, Japan
Kazuo Furuta The University of Tokyo, Japan
Masami Iwase Tokyo Denki University, Japan
Katsuhisa Ozaki Shibaura Institute of Technology, Japan
Shafishuhaza Sahlan Universiti Teknologi Malaysia
Fumiaki Araki JAMSTEC, Japan
Yosuke Onoue Nihon University, Japan
Susumu Nakata Ritsumeikan University, Japan
 AsiaSim 2018 Organization IX

Zhongkui Wang Ritsumeikan University, Japan

Katsumi Konishi Hosei University, Japan
Gary Tan National University of Singapore
Muhammad Shalihin Othman National University of Singapore
Chengxin Wang National University of Singapore
Kouta Sekine Toyo University, Japan
Sicheng Liu Beihang University, China
Rui Xu Dalian University of Technology, China
Xinchen Ye Dalian University of Technology, China
Kazuaki Tanaka Waseda University, Japan
Fei Wang Beihang University, China
Taro Kanno The University of Tokyo, Japan
Tomonori Yamada The University of Tokyo, Japan
Kazuya Shibata The University of Tokyo, Japan
Herman Wahid Universiti Teknologi Malaysia
Nurul Adilla Mohd Subha Universiti Teknologi Malaysia
Zaharuddin Mohamed Universiti Teknologi Malaysia
Hiroaki Natsukawa The University of Tokyo, Japan
Malcolm Low Singapore Institute of Technology, Singapore
 Contents

Modeling and Simulation Technology

A Novel Method to Build a Factor Space for Model Validation . . . . . . . . . . 3

Ke Fang, Ming Yang, and Yuchen Zhou

Simulation Credibility Evaluation Based on Multi-source Data Fusion . . . . . . 18

Yuchen Zhou, Ke Fang, Ping Ma, and Ming Yang

A Method of Parameter Calibration with Hybrid Uncertainty . . . . . . . . . . . . 32

Liu Bo, Shang XiaoBing, Wang Songyan, and Chao Tao

Reinforcement Learning Testbed for Power-Consumption Optimization . . . . . 45

Takao Moriyama, Giovanni De Magistris, Michiaki Tatsubori,
Tu-Hoa Pham, Asim Munawar, and Ryuki Tachibana

A DEVS Visual Front-End Interface for Model Reusability

and Maintainability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
Jiyong Yang, Moongi Seok, San Jeong, and Changbeom Choi

HLA-Based Federation Development Framework Supporting Model Reuse. . . . 72

Hang Ji, Xiang Zhai, Xiao Song, Xiaoliang Liu, Yazhou Liang,
and Zhengxuan Jia

Soft Computing and Machine Learning

Automatic Performance Simulation for Microservice Based Applications . . . . 85

Yao Sun, Lun Meng, Peng Liu, Yan Zhang, and Haopeng Chan

Predictive Simulation of Public Transportation Using Deep Learning. . . . . . . 96

Muhammad Shalihin Bin Othman and Gary Tan

An Ensemble Modeling for Thermal Error of CNC Machine Tools . . . . . . . . 107

Xuemei Jiang, PanPan Zhu, Ping Lou, Xiaomei Zhang, and Quan Liu

Gait Classification and Identity Authentication Using CNN . . . . . . . . . . . . . 119

Wei Yuan and Linxuan Zhang

Deep Dissimilarity Measure for Trajectory Analysis . . . . . . . . . . . . . . . . . . 129

Reza Arfa, Rubiyah Yusof, and Parvaneh Shabanzadeh
XII Contents

High Performance Computing and Cloud Computing

Performance Comparison of Eulerian Kinetic Vlasov Code Between Xeon

Phi KNL and Xeon Broadwell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
Takayuki Umeda and Keiichiro Fukazawa

Heterogeneous Scalable Multi-languages Optimization via Simulation . . . . . . 151

Gennaro Cordasco, Matteo D’Auria, Carmine Spagnuolo,
and Vittorio Scarano

Smart Simulation Cloud (Simulation Cloud 2.0)—The Newly Development

of Simulation Cloud . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
Bohu Li, Guoqiang Shi, Tingyu Lin, Yingxi Zhang, Xudong Chai,
Lin Zhang, Duzheng Qing, Liqin Guo, Chi Xing, Yingying Xiao,
Zhengxuan Jia, Xiao Song, and Rong Dai

A Semantic Composition Framework for Simulation Model Service . . . . . . . 186

Tian Bai, Lin Zhang, Fei Wang, Tingyu Lin, and Yingying Xiao

Simulation Technology for Industry

Dynamic Optimization of Two-Coil Power-Transfer System

Using L-Section Matching Network for Magnetically Coupled
Intrabody Communication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
Kenichi Ito

Demand and Supply Model for the Natural Gas Supply Chain in Colombia . . . 220
Mauricio Becerra Fernández, Elsa Cristina González La Rotta,
Federico Cosenz, and Isaac Dyner Rezonzew

Deep-Learning-Based Storage-Allocation Approach to Improve the AMHS

Throughput Capacity in a Semiconductor Fabrication Facility . . . . . . . . . . . . 232
Haejoong Kim and Dae-Eun Lim

Research on the Cooperative Behavior in Cloud Manufacturing . . . . . . . . . . 241

Ping Lou, Cui Zhu, Xiaomei Zhang, Xuemei Jiang, and Zhengying Li

Particle in Cell Simulation to Study the Charging and Evolution

of Wake Structure of LEO Spacecraft . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255
Nizam Ahmad, Hideyuki Usui, and Yohei Miyake

Simulation Technology for Intelligent Society

Wise-Use of Sediment for River Restoration: Numerical Approach

via HJBQVI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271
Hidekazu Yoshioka, Yuta Yaegashi, Yumi Yoshioka,
Kunihiko Hamagami, and Masayuki Fujihara
 Contents XIII

Calculation of Extreme Precipitation Threshold by Percentile Method

Based on Box-Cox Transformation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 286
Chi Zhang, Pu-wen Lei, and Koji Koyamada

OpenPTDS Dataset: Pedestrian Trajectories in Crowded Scenarios . . . . . . . . 296

Xiao Song, Jinghan Sun, Jing Liu, Kai Chen, and Hongnan Xie

Description and Analysis of Cognitive Processes in Ground Control

Using a Mutual Belief-Based Team Cognitive Model . . . . . . . . . . . . . . . . . 306
Sakiko Ogawa, Taro Kanno, and Kazuo Furuta

A Credibility Assessment Method for Training Simulations from the View

of Training Effectiveness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 316
Shenglin Lin, Wei Li, Shuai Niu, Ping Ma, and Ming Yang

Simulation of Instrumentation and Control Application

Digital Twin-Based Energy Modeling of Industrial Robots . . . . . . . . . . . . . . 333

Ke Yan, Wenjun Xu, Bitao Yao, Zude Zhou, and Duc Truong Pham

Dyna-Q Algorithm for Path Planning of Quadrotor UAVs . . . . . . . . . . . . . . 349

Xin Huo, Tianze Zhang, Yuzhu Wang, and Weizhen Liu

Boarding Stations Inferring Based on Bus GPS and IC Data . . . . . . . . . . . . 361

Xiang Yu, Fengjing Shao, Rencheng Sun, and Yi Sui

Acoustic Properties of Resonators Using Deployable Cylinders. . . . . . . . . . . 372

Sachiko Ishida and Ryo Matsuura

Iterative Unbiased Conversion Measurement Kalman Filter with Interactive

Multi-model Algorithm for Target Tracking . . . . . . . . . . . . . . . . . . . . . . . . 382
Da Li, Xiangyu Zou, Ping Lou, Ruifang Li, and Qin Wei

Computational Mathematics and Computational Science

On Convergence Speed of Parallel Variants of BiCGSTAB for Solving

Linear Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 401
Kuniyoshi Abe

Study on Chaotic Cipher with Robustness and Its Characteristics . . . . . . . . . 414

Takashi Arai, Yuta Kase, and Hiroyuki Kamata

A Stochastic Impulse Control Model for Population Management

of Fish-Eating Bird Phalacrocorax Carbo and Its Numerical Computation . . . 425
Yuta Yaegashi, Hidekazu Yoshioka, Koichi Unami,
and Masayuki Fujihara
XIV Contents

A Dialect of Modern Fortran for Computer Simulations . . . . . . . . . . . . . . . . 439

Shin’Ya Hosoyamada and Akira Kageyama

Flow Simulation

Performance Comparison of the Three Numerical Methods to Discretize

the Local Inertial Equation for Stable Shallow Water Computation . . . . . . . . 451
Tomohiro Tanaka, Hidekazu Yoshioka, Sokly Siev, Hideto Fujii,
Ly Sarann, and Chihiro Yoshimura

Development of the DRowning hUman Model (DRUM) Toward

Evaluation of Performance of Lifejackets in Tsunamis . . . . . . . . . . . . . . . . . 466
Daiki Ajima, Tatsuto Araki, and Takashi Nakamura

Illumination Recovery for Realistic Fluid Re-simulation. . . . . . . . . . . . . . . . 477

Hongyan Quan, Zilong Song, Xinquan Zhou, Shishan Xue,
and Changbo Wang

Ocean Analysis by Tsunami Simulation of the Nankai

Trough Massive Earthquake . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 488
Yuto Sakae, Ikuya Morimoto, Takuya Ozaki, Ryo Kurimoto, Liang Li,
Kyoko Hasegawa, Satoshi Nakada, and Satoshi Tanaka

Improving Traffic Flow at a Highway Tollgate with ARENA: Focusing

on the Seoul Tollgate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 501
Seung-Min Noh, Ho-Seok Kang, and Seong-Yong Jang

Visualization and Computer Vision to Support Simulation

Pixel Convolutional Networks for Skeleton-Based Human

Action Recognition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 513
Zhichao Chang, Jiangyun Wang, and Liang Han

Feature-Highlighting Transparent Visualization of Laser-Scanned Point

Clouds Based on Curvature-Dependent Poisson Disk Sampling. . . . . . . . . . . 524
Yukihiro Noda, Shu Yanai, Liang Li, Kyoko Hasegawa,
Atsushi Okamoto, Hiroshi Yamaguchi, and Satoshi Tanaka

Image-Based 3D Shape Generation Used for 3D Printing . . . . . . . . . . . . . . . 539

Zemin Li, Lin Zhang, Yaqiang Sun, Lei Ren, and Yuanjun Laili

A Memory Efficient Parallel Particle-Based Volume Rendering for

Large-Scale Distributed Unstructured Volume Datasets in HPC
Environments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 552
Yoshiaki Yamaoka, Kengo Hayashi, Naohisa Sakamoto,
and Jorji Nonaka
 Contents XV

A Transfer Entropy Based Visual Analytics System for Identifying

Causality of Critical Hardware Failures Case Study: CPU Failures
in the K Computer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 563
Kazuki Koiso, Naohisa Sakamoto, Jorji Nonaka, and Fumiyoshi Shoji

Modeling the Spread of Epidemic Diseases on ElasticStack-Based

Simulation Output Analysis Environment . . . . . . . . . . . . . . . . . . . . . . . . . . 575
Kangsun Lee and Sungwoo Hwangbo

Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 585

Modeling and Simulation Technology
 A Novel Method to Build a Factor Space
for Model Validation

Ke Fang(&), Ming Yang, and Yuchen Zhou

Control and Simulation Center, Harbin Institute of Technology,

Harbin 150001, China
{fangke,myang}@hit.edu.cn, zhouyuchen-01@163.com

Abstract. Factor space is an indispensable part of the model validation. In

order to provide an advantageous method to build factor space for model val-
idation, this paper states the challenging problems of the factor space, and
proposes a mathematical model of it. Further based on the model, this paper
provides the graphic illustration, the factor decomposition, the credibility
aggregation and the model defect tracing of the factor space, which construct a
novel method to build a factor space for model validation. Finally, the paper
provides a case study of an electromagnetic rail gun model validation to explain
the usage of the method.

Keywords: Model validation
Factor space
Credibility aggregation

Model defect tracing

1 Introduction

Simulation model is a complex object, which usually has multiple inputs/outputs and
sophisticated behaviors. The credibility of a simulation model is influenced by many
indicators (factors) related to the nature and outputs of the model. It is necessary to
build a factor space [1, 2] to describe these factors and their relationship, and further
aggregate the total credibility of the model through the factors. Traditional way of
building a factor space is using an AHP (Analytic Hierarchy Process) tree [3] and
weighted average function to aggregate the grand credibility from leaf nodes. However,
the relation between factors is not always linear, and the influence of the factor is not
always transferred through layers one by one. Model validation needs a better method
to reveal the factors and their influence to the credibility of the model.
Many generic methodologies develop the requirements, planning, architecture,
process and recommended techniques to perform verification and validation for M&S
(Modeling and Simulation) [4, 5]. Unfortunately, these methodologies all suggest
hierarchical tree to present and organize credibility indicators of model validation. Even
the project exercises and VV&A software tools adopt the traditional hierarchy to
perform simulation validation [6]. This cannot satisfy the practical requirements of the
model validation, and often leads to a credibility assessment result lack of objectivity
and confidence.
The remainder of this paper is organized as follows. We state the three challenging
problems of the factor space which are not resolved well by the hierarchical tree in
© Springer Nature Singapore Pte Ltd. 2018
L. Li et al. (Eds.): AsiaSim 2018, CCIS 946, pp. 3–17, 2018.
https://doi.org/10.1007/978-981-13-2853-4_1
4 K. Fang et al.

Sect. 2. Section 3 proposes the network method for building a factor space of model
validation. Section 4 presents a case study of the electromagnetic rail gun model
validation to explain the usage of the network method. Concluding remarks are given
in Sect. 5.

2 Challenging Problems

The basic function of the factor space in model validation is to determine the total
credibility of the model. In order to achieve the goal, factor space needs to reveal the
credibility indicators and their relationship, and use them to aggregate the total cred-
ibility. There are some challenging problems in the function of the factor space,
especially the factor decomposition, credibility aggregation and the model defect
tracing.

2.1 Factor Decomposition

Factor decomposition is the driving force of a factor space building. Naturally, in
model validation the root factor should be the total credibility of the model. Further by
making decomposition downwards layer by layer, we can reveal all the factors which
influence the model credibility at different level. The relationship between factors is
related to the decomposition which is made, and this relationship may exist between
any two factors, no matter at which layer the factor is located. Obviously, the AHP
hierarchical tree is a special case of the factor space, which can only present factor
affiliation between adjacent layers. However, the credibility indicators work in a more
complicated way than the analytic hierarchy.
Take the factor space part shown in Fig. 1 as an example. A 6-DOF (Degree of
Freedom) flying vehicle model is divided into four parts, which are “motion of mass
center”, “motion of around mass center”, “motion of elasticity”, and “motion of
shaking”. The variants below are the physical quantities of the motions. Apparently, the
model credibility is determined by the four parts together, and further is influenced by
the output variants below.

6-DOF Flying Vehicle Model

Motion of Motion of around Motion of Motion of

mass center mass center elasticity shaking

x y z φ ψ γ qiy qiz qni yp zp

1

Fig. 1. Part of a 6-DOF flying vehicle model validation factor space.

Actually the indicators of four motions are decomposed from the root by the model
structure, and this is what the hierarchical tree does in AHP. However when the
 A Novel Method to Build a Factor Space for Model Validation 5

decomposition goes down further it is stuck, because the physical quantities are linked
by the model resolving process but not the structure. Obviously the outputs (both
intermediate and final ones) are not always irrelevant and the relationship is not always
linear, thus it is incorrect to make structural decomposition further.
Meanwhile, in case that the outputs x; y; z are so important that if they failed to meet
some acceptability criteria, no matter what the performance of other indicators are, the
model credibility will be unaccepted. This is an ultra-connection from x; y; z to the root
node, which a traditional hierarchy cannot deal with. So we need a better way to
perform factor decomposition.

2.2 Credibility Aggregation

The traditional way of credibility aggregation by AHP hierarchy is using weighted
average function to accumulate the influence from leaf nodes to root. When the hier-
archy and weight matrix are set, the input of the credibility aggregation is only the
value of leaf nodes. The aggregation can be expressed by the function below:

X
q Y
ki
v0 ¼ ðvi
wj Þ; ð1Þ
i¼1 j¼1

where v0 is the root value (model credibility), vi is the leaf value, wj is the weight and
should be summed to 1 among brother nodes, q is the number of leaf nodes, and ki is
the number of ancestor nodes which belong to the leaf node possessing vi . Apparently,
the weighted average function is linear, and demands the brother nodes are irrelevant
each other at different levels. Actually this is almost impossible to be satisfied in the
model validation. See the factor space part expanded from the “Motion of mass center”
node of Fig. 2 below.

Fig. 2. Factor space part expanded from the “Motion of mass center” node of Fig. 1.
6 K. Fang et al.

Take the node of Wx (the vehicle acceleration of x direction) as an example. Wx is

obtained by the following formula containing Fqx1 (aerodynamic force of x direction)
and m (the vehicle mass), and further the Fqx1 is expanded as Cx (aerodynamic coef-
ficient of x direction), q (aerodynamic) and SM (cross sectional area of the vehicle), as
the formula shows below:

Fqx1 Cx qSM

Wx ¼ ¼ : ð2Þ
m m

Because m changes during the simulation, Fqx1 is not linear to Wx . If use normal
AHP hierarchy and weighted average function to aggregate the partial credibility of Wx
from Fqx1 and m, it will be an incorrect result. Actually the relationship of Fqx1 and m to
Wx is a derivation, but not a composition. So the credibility aggregation needs better
solution than the traditional way.

2.3 Model Defect Tracing

Except for revealing the credibility indicators and their relationship, factor space has
another function of model defect tracing [7]. The traditional AHP hierarchy is lack of
precise defect tracing ability, because there is swamping effect. The calculation of
combined weight may lessen the influence of certain indicators to the model credibility.
When these indicators fail to obtain acceptable validation results, we cannot detect
them by the mere final credibility.
If we find a new way to build factor space and express the relationship between
indicators more than hierarchy, it is possible to use it for model defect tracing. Not only
the leaf nodes but also the branch nodes should possess partial credibility, which allows
path tracing to locate the indicators that cause the deficiency of model credibility.

3 The Network Method for Building a Factor Space

The hierarchical tree has a limitation of connections inside, that is, it only allows
relationship between factors in adjacent layers. In order to express more complicated
connections between credibility indicators, we propose a network method to build the
factor space in model validation.

3.1 The Mathematical Definition of the Factor Space Network

Define the factor space network as a directional graph composed by radially distributed
nodes. The network can be expressed by the quadruple below:

F ¼ f\N; V [ ; \L; A [ g; ð3Þ

where F is a factor space, N and V are the node set and value set, L and A are the link
set and attribute set mapped with each other. Set the link direction as from attribute-
holding node to attribute-receiving node. For example, a structural link of the
 A Novel Method to Build a Factor Space for Model Validation 7

traditional hierarchy has the direction from child node to parent node. According to the
requirements of model validation, further develop element definitions of the network as
the following:
Definition 1: Define N ¼ fn1 ; n2 ; . . .; nk g as the node set. N ¼ N  [N~ where N  is the
certain node set and N~ is the uncertain node set. If ni 2 N,
 use tðni Þ to indicate the node
~
type. tðni Þ 2 fregular; sufficient; inherited g. If ni 2 N, use cðni Þ as the transit condi-
tion, and cðni Þ 2 f0; 1g.
Definition 2: Define V ¼ fv1 ; v2 ; . . .; vk g as the value set. If tðNÞ 2 fregular;
sufficientg, V and N are mapped with each other. Use vi ¼ vðni Þ to indicate the value of
ni which is mapped with vi .
[L
Definition 3: Define L ¼ fl1 ; l2 ; . . .; lk g as the link set. L ¼ L ~ where L  is the certain
link set and L ~ is the uncertain link set. Use tðli Þ to indicate the link type, and
tðli Þ 2 fregular; sufficient; ultra;equivalent; contraditory; inherited; tracedg. Use to l ¼
ðni ; nj Þ express a link from node ni to node nj .
Definition 4: Define A ¼ fa1 ; a2 ; . . .; ak g as the attribute set, and A is one-one mapped
with L. Use ai ¼ aðli Þ to indicate the attribute of link li which is mapped with ai . If
tðli Þ 2 fequivalent; inherited; traced g, ai 2 £.
~ and cðnÞ ¼ 1, then n 2 N.
Definition 5: If n 2 N  If cðnÞ ¼ 0, then n 2 ;, and define n
as rubbish node, which needs to be deleted from the network.
 tðnÞ ¼ sufficient and vðnÞ ¼ 1, then the sufficient link from n
Definition 6: If n 2 N,
breaks. If vðnÞ ¼ 0, then n needs to supplement additional brother node, which is
defined as shadow node and values 0.
Definition 7: If l 2 L, tðlÞ ¼ regular, and l ¼ ðno ; nd Þ, define aðlÞ 2 ½0; 1 as the weight
 and tðLc Þ ¼ regular, then
distributed from nd to no . If Nc ¼ fnjðn; nd Þ 2 Lc g, Lc 2 L,
Pk
ak ðlk Þ ¼ 1, lk 2 Lc and k ¼ dðLc Þ.
i¼1

Definition 8: If l 2 L, l ¼ ðno ; nd Þ, and tðlÞ ¼ ultra, define aðlÞ 2 ½0; 1 as the

acceptability threshold. If vðno Þ \ aðlÞ, then vðnd Þ ¼ 0, and define no as the key node
of nd , nd as the super conduct node of no .
Definition 9: If l 2 L, l ¼ ðno ; nd Þ, and tðlÞ ¼ inherited, then the sub-network under
nd has to be replicated to no , and the new nodes are defined as inherited nodes, whose
physical meaning will be given by no .
Definition 10: If l 2 L, l ¼ ðno ; nd Þ, and tðlÞ ¼ contradictory, define
aðlÞ2 ½0; 1 as the contradiction percentage of no to nd . vðnd Þ
vðnd Þ vðnd Þ  1  aðlÞ
vðn0 Þ
¼ , where vðno Þ is the source node value,
1  aðlÞ
vðn0 Þ vðnd Þ [ 1  aðlÞ
vðn0 Þ
vðnd Þ is the destination node value.
Definition 11: If l 2 L, l ¼ ðno ; nd Þ, and tðlÞ ¼ equivalent, define no & nd as mirror
node, and vðnd Þ ¼ vðno Þ.
8 K. Fang et al.

Definition 12: If l 2 L, l ¼ ðno ; nd Þ, and tðlÞ ¼ traced, define nd as traced node. vðnd Þ
is irrelevant to vðno Þ.
Definition 13: If  n ¼ fxjx 2 N ^ ðx; nÞ 2 Lg, define  n as the pre-set of n. If
n ¼ fxjx 2 N ^ ðn; xÞ 2 Lg, then define n as the post-set of n.
Definition 14: If ðni ; nj Þ 2 L, define ni as the child node of nj , and nj as the father node
of ni . If ðni ; np Þ 2 L and ðnj ; np Þ 2 L, define ni and nj are brother nodes. If n ¼ ;,
define n as root node. If n 6¼ ; and  n 6¼ ;, define n as branch node. If  n ¼ ;, define n
as leaf node.
Definition 15: Set power operator satisfies ðn Þ0 ¼ n, ðn Þ1 ¼ n , ðn Þ2 ¼ ðn Þ … If
nj 2 ðni Þs1 , nj 2 ðni Þs2 ,…, nj 2 ðni Þsk , define Sðni ! nj Þ ¼ fs1 ; s2 ; . . .; sk g as the dis-
tance set from ni to nj . Non-negative integers s1  sk are all distances from ni to nj . If n0
is a root node, abbreviate Sðni ! n0 Þ as Sni .
Definition 16: If NL ¼ fn1 ; n2 ; . . .; nk g, 8n 2 NL makes  n \ NL ¼ ; and n \ NL ¼ ;,
and 8ni ; nj 2 NL , 8sk 2 Sni , 8sm 2 Snj makes Maxðsk Þ ¼ Maxðsm Þ, define NL as a layer
(order) of the factor space network, which is the Maxðsk Þ–th layer.
Definition 17: If 9s; nj 2 ðni Þs , define ni to nj as reachable. If 8s; nj 62 ðni Þs , define ni
to nj as unreachable, and set sðni ! nj Þ ¼ 1.
Definition 18: If sðni ! nj Þ [ 1 and sðni ! nj Þ 6¼ 1, define ni as the offspring node
of nj , nj is the ancestor node of ni , and ni & nj are lineal relative nodes. If sðni !
nj Þ ¼ 1 and ni & nj are not brother nodes, define ni & nj are collateral relative nodes.

3.2 The Rules of the Factor Space Network

In order to use the factor space network to perform model validation, we must define
rules to regulate its structure and operation. The rules below should be followed when
use the factor space network to validate simulation models.
Rule 1: If n 2 N, then  n [ n 6¼ ;. If l 2 L and l ¼ ðn1 ; n2 Þ, then n1 6¼ ; and
n2 6¼ ;. (There is no isolated node or link in the network.)
Rule 2: If N0 ¼ fnj n ¼ ;g, then dðN0 Þ ¼ 1 and N0  N. (There is only one root
node in the factor space.)
Rule 3: If n 2 N, ~ then l ¼ ðn; ni Þ 2 L. ~ (The link which has a source of uncertain
node is an uncertain link.)
Rule 4: If n 2 N and tðnÞ ¼ sufficient, then l ¼ ðn; ni Þ 2 L and tðlÞ ¼ sufficient.
(The link which has a source of sufficient node is a sufficient link.)
~ 8nc 2 Nc makes nc to np is reachable, and
Rule 5: If Nc ¼ fn1 ; n2 ; . . .; nk g, np 2 N,
cðnp Þ ¼ 0, then 8nc 2 Nc makes cðnc Þ ¼ 0. (The offspring nodes of a rubbish node
are all rubbish nodes.)
Rule 6: If nr 2 N, ~ cðnr Þ ¼ 0, lrb ¼ ðnr ; ni Þ 2 L~ and lre ¼ ðni ; nr Þ 2 L,
~ then
cðlrb Þ ¼ cðlre Þ ¼ 0. (The uncertain link which starts from or ends with a rubbish
node is a rubbish link.)
 A Novel Method to Build a Factor Space for Model Validation 9

3.3 The Graphic Illustration of the Factor Space Network

In order to express the factor space network visually, we define necessary graph
elements to provide a graphic illustration. The graph element set is mapped with all
definitions and follows the rules. The figure below shows an example of the graphic
illustration of the factor space (Fig. 3).
The graphic illustration is explained below:
(1) Use single-lined figure (circle or rectangle) to present regular node, double-lined
figure to present sufficient node, and double-lined round-cornered rectangle to
present inherited node.
(2) Mark the node name or number in the node figure, and mark the node value and
the transit condition outside nearby.
(3) Use directional line segment to present link, and use line end to present the link
type. Solid arrow end presents regular link, circle end presents sufficient link,
double-arrow end presents ultra link, equality-sign end presents equivalent link,
slash-sign end presents contradictory link, hollow arrow end presents traced link,
and slash-signed circle end presents inherited link.
(4) Mark the attribute by the link. To avoid the intersection of the links, fold the link
and mark the source and target node at the folded link.
(5) Use solid-lined figure to present certain element, and use dot-lined figure to
present uncertain element.

Fig. 3. Example of the graphic illustration of the factor space.

3.4 Credibility Aggregation

Dynamic elements (uncertain and sufficient node/link) affect the structure of the factor
space. The credibility aggregation cannot be performed unless the network is static,
which means the dynamic elements have to be analyzed first.
10 K. Fang et al.

When the dynamic analysis is done, the credibility aggregation can be achieved by
the following procedure:
(1) Go through the factor space downwards, and stop at the destinations of traced
links. Perform similarity analysis by comparing the simulation outputs and real
world outputs, and get partial credibility on these nodes after result transformation
[8]. Generally, the partial credibility can be achieved by:

jjOðnÞ  Oðn0 Þjj

CðnÞ ¼ 1  jIðnÞ¼Iðn0 Þ ; ð4Þ
jjOðn0 Þjj

where IðnÞ and OðnÞ are the simulation input and output of the node, Iðn0 Þ and
Oðn0 Þ are the corresponding real world input and output, and jj
jj represents the
norm of the variant. According to the technique used, CðnÞ can be achieved by
statistics analysis [9], time domain analysis [10] or frequency domain analysis
[11] methods together with their result transformation formulas.
(2) If the destination of the traced link is lack of real world output, then go down
along the path further to find a node whose real world output is valid, and get the
value of the node by similarity analysis. The partial credibility of the upper node
which has no real world output can be achieved by error analysis via the com-
putational process of the model:

jjf ðð2  Cðn1 ÞÞ
Oðn1 ÞÞ  f ðOðn1 ÞÞjj

Cðn2 Þ ¼ 1  ; ð5Þ
jjOðn2 Þjj

where n1 is the node which has real world output, Cðn1 Þ is partial credibility, and
Oðn1 Þ is simulation output. n2 is the node which has no real world output, Cðn2 Þ is
partial credibility, and Oðn2 Þ is simulation output. f is the computational function
of the model from n1 to n2 .
(3) Use appropriate algorithm to aggregate the total credibility on the root node, by
gathering the partial credibility on the destinations of the traced links along the
decomposition paths of the factor space:

C ¼ f ðvi ; vi þ 1 ; . . .; vi þ k Þ; ð6Þ

where C is the total credibility, vi  vi þ k are the partial credibility on the

destination nodes of the traced links, f is the aggregation function. f can be the
method of taking the minimum, and weighted average etc., and can be different
across the layers in the factor space.

3.5 Defect Tracing

When the total credibility of the model is unsatisfactory, we can perform defect tracing
via the factor space network by following the procedure below:
(1) Use orthogonal design and Sobol’ method [7] to locate the defect factors other
than the destination of traced links.
 A Novel Method to Build a Factor Space for Model Validation 11

(2) Along the validation path in the factor space which contains the defect factors
already located, make further validation to get partial credibility on the destination
of traced links, and determine if it is a defect node by comparing with the
acceptability criteria.
(3) If there is a destination of iteration paths, make validation of its initial input, related
constants, and other variants which are irrelevant to the iteration variant [12].
(4) When the validation path reaches the leaf node, the defect tracing is over.
(5) Collect all the defect nodes in the tracing, and take the nodes with lowest order as
the origin that induces the deficiency of the model credibility.

4 Case Study

The electromagnetic rail gun model is composed by the sub-models of power supplies,
paralleled rails, armature, wave modulation inductor, projectile etc. The armature is
located between the two rails, and conducts the current and delivers Lorentz force. The
projectile is located in front of the armature, which is forced to move forward together
with the armature. The model reveals the whole physical process of the real world
object.

4.1 The Factor Space Network

According to the model’s resolving process, use the network method to build the factor
space for model validation. Figure 4 shows the factor space network.

4.2 The Validation of the Projectile Displacement

According to the factor space network in Fig. 4, the destination of the traced links is the
projectile displacement x. We use MGRA (Modified Grey Relational Analysis) to
validate the node. The analysis formula is shown below:
8
>
> cm ðX1 ; X2 Þ ¼ cðX1 ; X2 Þ  RRMSE
>
> P n
>
> cðX1 ; X2 Þ ¼ 1n cðx1 ðkÞ; x2 ðkÞÞ
>
>
>
> k¼1
>
>
< minjx1 ðkÞx2 ðkÞj þ q maxjx1 ðkÞx2 ðkÞj
cðx1 ðkÞ; x2 ðkÞÞ ¼ kjx1 ðkÞx2 ðkÞj þ q maxkjx1 ðkÞx2 ðkÞj ; ð7Þ
>
>  0:5 k
>
> P
>
n
>
>
1
ðx1 ðkÞx2 ðkÞÞ2
>
>
n
>
> RRMSE ¼ k¼1
P
>
: 1
n
jx2 ðkÞj
n
k¼1

where X1 is the simulation data series, X2 is the observed data series,

minjx1 ðkÞ  x2 ðkÞj is the minimum difference, maxjx1 ðkÞ  x2 ðkÞj is the maximum
k k
difference, RRMSE is the relative root mean square error of the simulation and observed
data. The result transformation from MGRA to simulation credibility can be defined as:
12 K. Fang et al.

Fig. 4. The factor space network of the electromagnetic rail gun model.

8
< 1Cth ðc  c Þ þ C ; c 2 ½c ; 1
1c m th th m th
Cðcm Þ ¼ th
Cth cm ; ð8Þ
: ; cm 2 ½0; cth 
cth

where is cm is the analysis result of MGRA, cth is the acceptability threshold of

similarity analysis, Cth is the acceptability threshold of simulation credibility. The
 A Novel Method to Build a Factor Space for Model Validation 13

simulation data and observed data are shown in Table 1. The data curves of the
projectile displacement x are shown in Fig. 5.
Set cth ¼ 0:5 & Cth ¼ 0:8, and use the function in Formula (7, 8), we can obtain the
partial credibility of the projectile displacement x is Cx ¼ 0:7049. Because Cth ¼ 0:8,
the node is not accepted. Meanwhile we can see the root node is directly linked by the
projectile displacement x only, so the model’s total credibility is also 0:7049 and
unaccepted. We need to perform defect tracing to find which part of the model causes
the credibility deficiency.

Table 1. The simulation data and observed data of the projectile displacement x.
Time (ms) Simulation data (m) Observed data (m)
0.1000 0.0001 0.0001
0.2000 0.0009 0.0009
0.3000 0.0048 0.0050
0.4000 0.0148 0.0155
0.5000 0.0335 0.0356
…… …… ……
1.1000 0.3404 0.3796
1.2000 0.4246 0.4746
…… …… ……
4.0000 4.6349 5.2051
4.1000 4.8064 5.3973
4.2000 4.9782 5.5899
4.3000 5.1503 5.7828
4.4000 5.3227 5.9760

Fig. 5. The curves of the projectile displacement x.

14 K. Fang et al.

4.3 The Validation of the Individual Circuit Current

According to the defect tracing procedure of the factor space, we make further vali-
dation along the paths of traced links. Because of the limited observed data and length
of the paper, we pick out two representative nodes individual circuit current and rail
voltage to explain the defect tracing work. The model contains 100 power capacitors
grouped by every 10 packs, which discharge in chronological order. Take the
No. 9 group of the capacitor as an example, the individual circuit current data is shown
in Table 2. The data curves of the No. 9 individual circuit current i9 are shown in
Fig. 6.

Table 2. The simulation data and observed data of the No. 9 individual circuit current i9 .
Time (ms) Simulation data (kA) Observed data (kA)
0.8000 0.0000 0.0000
0.9000 63.4997 63.3014
1.0000 118.0873 118.6114
1.1000 156.2808 159.1810
1.2000 172.6993 179.9446
…… …… ……
2.6000 71.6777 75.8113
2.7000 67.5994 71.4827
…… …… ……
4.0000 33.8028 35.8419
4.1000 32.1987 34.1577
4.2000 30.6890 32.5729
4.3000 29.2671 31.0806
4.4000 27.9271 29.6744

Fig. 6. The curves of the No. 9 individual circuit current i9 .

 A Novel Method to Build a Factor Space for Model Validation 15

Use the function in Formula (7, 8), we can obtain the partial credibility of the
No. 9 individual circuit current i9 is Ci9 ¼ 0:7393. The node is not accepted and we
need further defect tracing along the traced links downwards.

4.4 The Validation of the Rail Voltage

The rail voltage data is shown in Table 3. The data curves of the rail voltage Ur are
shown in Fig. 7.

Table 3. The simulation data and observed data of the rail voltage Ur .
Time (ms) Simulation data (V) Observed data (V)
0 0 0
0.1000 0.6055 0.6286
0.2000 6.6897 7.2459
0.3000 29.8628 33.4997
0.4000 72.0103 84.9831
0.5000 110.6497 134.7681
…… …… ……
1.1000 402.5833 493.5161
1.2000 410.8884 509.3004
…… …… ……
4.0000 −42.0534 −47.6174
4.1000 −43.5441 −49.2517
4.2000 −44.7558 −50.5756
4.3000 −45.7219 −51.6273
4.4000 −46.4721 −52.4406

Fig. 7. The curves of the rail voltage Ur .

16 K. Fang et al.

Use the function in Formula (7, 8), we can obtain the partial credibility of the rail
voltage Ur is CUr ¼ 0:6849. The node is not accepted.

4.5 Result Analysis

According to the theory of the factor space network, the credibility of the electro-
magnetic rail gun model is C ¼ 0:7049, which is not accepted. The defect tracing
shows that the node which causes the credibility deficiency is under the node of rail
voltage Ur . By further validation we find that the constant of inductor gradient L0r is
problematic. The constant in simulation is set as L0r ðsÞ ¼ 0:42 106 , but the observed
value in the real world is L0r ðrÞ ¼ 0:40 106 . Although the difference is tiny, since
the magnetic force F is in proportion to the total circuit current I, and I is at the level of
mega amps, the error is magnified dramatically. Meanwhile, the traced links in the
factor space network give chain reaction of the error, and induce the credibility defi-
ciency of the model.
Moreover, we can see that the total circuit current I, the individual circuit current i,
the projectile acceleration a, the projectile velocity v, and the projectile displacement x
have cross iteration links in the factor space network. These iterations intensify the
accumulation of the error. So as long as to correct the value of L0r in the simulation, the
electromagnetic rail gun model has the opportunity to gain an acceptable credibility.

5 Conclusion

The factor space plays an important role in model validation. It reveals all the indicators
that influence the total credibility of the model, and makes the work of model defect
tracing possible. However, the traditional way of factor space building only analyses
the model by structural decomposition, and relies on the linear aggregation function to
get the total credibility. This cannot satisfy the factor space requirements in the engi-
neering of model validation. The network based factor space overcomes the limitation
of the traditional hierarchy, and serves for the model validation better. Case studies
prove that the method is practical and effective.
Meanwhile we should notice that, although the factor space lists all the credibility
indicators and their relationship, it does not mean the validation can be performed
thoroughly without doubt. Some physical quantity is not observable or very hard to
measure in the real world, such as the aerodynamic force, the projectile mass etc. The
node possessing this kind of physical quantity is called “black node”. We cannot
perform similarity analysis from comparing the simulation data and observed data and
get the partial credibility, except for knowledge based expert evaluation. Therefore, if
there is black node in the factor space, the validation result may be less of objectivity.
The definition of the network based factor space is extendable. The types of each
quadruple element can be extended to express new indicators and relation-
ship. According to the practical requirements in validation engineering, the network
method for building a factor space should be more adaptive and accuracy.
Another random document with
no related content on Scribd:
Polybia, nests of, 81, 82, 83
Polyctenes fumarius, 560
Polyctenidae, 560
Polyergus lucidus, 151;
P. rufescens, 150 f.
Polymorpha (Coleoptera), 189, 190, 213 f.
Polymorphism, 139 f., 143
Polyphylla fullo, antenna, 191
Polyploca, 386
Polyplocidae, 386
Polyplocotes, 248
Polyrhachis, 155;
P. pandurus, 156;
P. spinigera, 138
Pompilidae, 93, 101 f.
Pompilus, 103;
P. polistoides, 104;
P. sericeus, 106
Ponera contracta, P. ergatandria, P. punctatissima, 172
Ponerides, 132, 170 f.
Porphyraspis tristis, 284
Porphyrophora polonica, 597
Porrorhynchus, 216
Portschinsky, on habits and development of Muscidae, 507, 512
Post-metamorphic growth, 141
Post-scutellum, 307, 312
Potamogeton pectinatus, beetle on, 280
Potamophilus acuminatus, 244
Potassium hydroxide, 328, 384
Potato-beetle, 278
Pouch, abdominal, 350, 362
Poulton, on colours, 336, 339
Praecostal nervures, 319
Praescutum, 312
Pratt, on imaginal discs, 453 n.;
on Melophagus, 519
Prepona, larva, 354
Prey, of Fossores, table, 92
Pria dulcamarae, 232
Primitive, beetles, 251, 252;
Diptera, 475
Priocnemis affinis, 5;
P. bicolor, 107
Priodont, 193
Prionides, 287
Proboscis, 13, 14, 16, 17, 304, 307, 309, 311, 443, 482, 485,
532
Processional maggots, 464
Processionary caterpillars, 376, 408
Prodoxidae, 432;
mouth, 309
Prodoxus, 433
Pro-legs, 323
Prolific, Aphids, 589;
Coccids, 594;
Hepialus, 397;
Lice, 601;
Meloe, 274;
Stylops, 301
Prominents, 383
Pronuba, 321;
P. yuccasella, 432;
P. synthetica, 432
Pronymph, 453
Propodeum, 131, 133
Propolis, 63
Propygidium, 187
Prosopis, 21, 22;
hair of, 11;
proboscis of, 17;
P. signata, 21
Protection, 43, 413;
of trees by ants, 158;
of plant by ants, 168
Proterhinidae, 298
Proterhinus lecontei, 298
Protolepidoptera, 336
Protoparce carolina, 309
Protopaussus, 214
Psammorycter vermileo, 481
Pselaphidae, 223;
and ants, 182
Psen atratus, P. concolor, 127
Psephenus, 244
Pseudholoptic, 440 n.
Pseudocorylophidae, 228
Pseudodicthadia, 180;
P. incerta, 177
Pseudodoxia limulus, 431
Pseudomeria graeca, 99
Pseudomorphides, 205, 206
Pseudomyrma bicolor, 168
Pseudomyrmini, 168
Pseudoneuroptera, 527
Pseudopaedogenesis, 303
Pseudopod, 188, 264, 267, 290, 449, 492
Pseudopontia paradoxa, 357
Pseudo-pupa, 271, 273
Pseudotetramera, 190
Pseudotrimera, 239
Pseudovespa, 88
Pseudovarium and Pseudovum, 584
Pseudovitellus, 588
Psilidae, 504
Psiliglossa, larva, 8
Psilocephala, 484
Psilura monacha, 407
Psithyrus, 53, 57, 59 f.;
P. campestris, 60;
P. vestalis, 60
Psocidae, 248
Psyche helix, 394
Psychidae, 369, 392
Psychina, 394, 395, 404
Psychoda, 466
Psychodidae, 470
Psylla pyricola, 579, 580;
P. succincta, 579;
P. buxi, 580
Psyllidae, 578
Pterocheilus, 76
Pterodecta, 400
Pterophoridae, 340, 371, 425, 426
Pterophorinae, 426
Pterostichus, 205
Pterothysanidae, 369, 406
Pterygodes, 312
Pterygogenea, 542
Ptilinum, 442, 503, 520
Ptilinus pectinicornis, 253
Ptilocnemus sidnicus, 557
Ptilomacra, 395
Ptilomera laticaudata, 553
Ptinidae, 246
Ptomaphila lacrymosa, 222
Ptychoptera, 466
Ptychopterinae, 472
Ptyelus goudoti, 577
Pugs, 411
Pulex avium, P. irritans, P. serraticeps, 525;
P. obtusiceps, 525
Pulicidae, 522 f.
Puliciphora lucifera, 495
Pulvillus, 446
Pupa, of beetles, 188 f.;
coarctata, 452;
obtecta, 227, 327, 451;
depositing eggs, 469;
 hairy, 407, 426;
of Lepidoptera, 326 f.;
with mandibles, 436
Puparium, 452
Pupation, of Badamia, 365;
of Parnassius and Thais, 363
Pupipara, 456, 513, 517
Purple emperor, 344
Pusci, 504
Puss-moth, 328, 383, 406
Pygidium, 187
Pyralidae, 340, 370, 420
Pyralidina, 420, 426
Pyrameis atalanta, 353;
P. cardui, 353;
P. tameamea, 353
Pyraustidae, 421
Pyrochroa rubens, 266
Pyrochroidae, 266
Pyrophorus, 258;
P. noctilucus, 259
Pyrrhocoridae, 549
Pyrrhocoris apterus, 535, 549
Pyrrhopyge, 364
Pythidae, 265
Pytho depressus, 266

Quartenia, 89
Queen, 66, 67, 69, 140

Races, of Apis, 68
Radial nervures, 319
Raffray, on Pogonostoma, 204;
on classification of Paussidae, 214;
on classification of Pselaphidae, 224
Railway-beetle, 251
Ranatra linearis, 563
Raptorial legs, 493, 554, 556
Réaumur, on Xylocopa, 33
Receptaculum seminis, 140
Rectal cauda, 538
Red admiral, 352
Red ant—see Formica rufa
Reduviidae, 555 f. 537
Reduvius personatus, 558
Régimbart, on Gyrinidae, 216
Resemblance, between ant, wasp, and spider, 169;
between Anthophora and Bombus, 33;
between Arctia villica and Eusemia villicoides, 410;
between beetle-larva and Termite, 206;
between Bombus and Eulema, 35;
between Bombyliidae and Hymenoptera, 489;
between bug and ant, 556, 557;
bug and Tipulid, 556, 558;
bug and fly, 547;
in butterflies, 348;
between Callidulidae and Lycaenidae, 400;
between Celyphus and Hemiptera, 505;
between Cleridae and insects they destroy, 254;
between Dioptinae and Ithomiides, 409;
between Diptera and Hymenoptera, 499, 500;
between Epicopeia and Papilio, 418;
between Flatides and Lepidoptera, 576;
of flies and bees, 502;
to galls, 403;
between host and parasite-bees, 30;
between Insects of different Orders, 339;
of Ithomiides to other butterflies, 346;
between lady-bird and Endomychid, 237;
of larva to a colony of larvae, 418;
of larva of Odynerus and of Chrysis, 4;
between larvae, 162;
in Limacodidae, 401;
of Lobster caterpillar, 385;
 between Longicorns and Hymenoptera, 287;
of moth to bird-excrement, 401;
of parasite and prey, 95;
of Pericopinae and Heliconiides, 409;
between protected butterflies, 345;
between Psithyrus and Bombus, 59;
of Reduviid and Pepsis, 558;
between Rhyphus- and Mycetobia-larvae, 463;
of Syntomids to other Insects, 388 f.;
of Tipulids and Hymenoptera, 475;
between two kinds of ants, 162
Resting-larva, 306
Retinaculum, 316, 319, 420
Retort-shaped bodies, in mouth of Hemiptera, 535
Reuter, E., on classification of butterflies, 343 n.
Rhachicerus, 480
Rhagovelia plumbea, 552
Rhaphiorhynchus, 483
Rhegmatophila alpina, 305, 386
Rheumatobates bergrothi, 553
Rhinomacerides, 291
Rhinopsis, 115;
R. ruficornis, 169
Rhinosimus, 266
Rhipicera mystacina, 256
Rhipiceridae, 256
Rhipidiini, 267
Rhipidioptera, 543
Rhipiphoridae, 267
Rhipiphorides, 268
Rhipiptera, 298 [in error for Rhiphiptera]
Rhizophagus, 232
Rhizotrogus, 191
Rhodoneura, 405
Rhogmus, 179, 180
Rhopalocera, 340, 341 f.
Rhopalomelus angusticollis, 206
Rhopalomeridae, 504
Rhopalosoma poeyi, 100
Rhopalosomides, 100
Rhygchium brunneum, R. carnaticum, R. nitidulum, R. oculatum,
77
Rhynchites betulae, 291, 292
Rhynchitides, 291
Rhynchophora, 190, 277, 288 f.
Rhynchophorous series, 240
Rhynchopsylla pulex, 526
Rhynchota, 532
Rhyphidae, 478
Rhyphus, 463
Rhysodidae, 201, 234
Riley, on Epicauta, 270, 271;
on spermathecal bodies, 321;
on Yucca-moth, 432
Ripidius pectinicornis, 269
Rippe, of Schäffer, 334
Robber-flies, 491
Rose-chafers, 200
Rosema, 401
Rostrum, 291, 472
Rothney, on Ampulex, 115;
on Sima, 169;
on Sphex, 110
Rothschild, N. C., on morphology of fleas, 523 n.
Royal jelly, 66
Rozites gongylophora, 167
Ruby-wasps, 1 f.
Rurales, 342
Rutelides, 195, 198

Saccoderes, 558;
S. tuberculatus, 537
Saccus, 314
Sagra splendida, 279
Sagrides, 279
Salda, 560
Saldidae, 544, 562
Salivary, duct, 320;
gland, 320
Sand-flea, 525;
-flies, 477
Saperda populnea, 285, 286
Sapromyzidae, 504
Sapyga quinquepunctata, 100
Sapygides, 99
Sarcophaga, 510;
S. carnaria, 510
Sarcophagidae, 510
Sarcophila magnifica, S. wohlfahrti, 510
Sarcopsylla gallinacea, S. penetrans, 525
Sarginae, 479
Saropoda, 32
Sarrothripus, 410
Sasaki, on parasite of silkworm, 508
Saturnia, 307, 310;
S. pavonia, 313, 374
Saturniidae, 368, 372
Satyrides, 347
Sauba, or Sauva, ant, 137
Saunders, E., on feathered hairs, 11;
on proboscis of bees, 16
Saunders, Sir S., on Hymenopterous larvae, 8
Scale, 131, 133, 315, 467;
development of, 329 f.;
-formation, 333;
-holder, 331
Scale-Insects, 592 f.;
enemy of, 356, 357, 417, 425, 430
Scalpella, 443
Scape, 441
Scaphidiidae, 229
Scaphisoma agaricinum, 229
Scaphium, 314
Scapulae, 312
Scarabaeidae, 194 f.
Scarabaeini, 196
Scarabaeus sacer, 196, 197
Scatomyzidae, 504
Scatophagidae, 504
Scatopse, 477
Sceliphron, 112;
S. nigripes, 91
Scenopinidae, 484
Schäffer, on structure of wings and nervures, 330
Schiödte, on Anoplura, 599 f.;
on Hemiptera, 543
Schistocerca peregrina, enemy of, 506, 514
Schizocarpus mingaudi, 220
Schizometopa, 504
Schizoneura, 589
Schizophora, 455
Schmidt-Schwedt, on Donacia, 280
Schoenbergia, 360, 361
Schoenobiinae, 425
Sciapteron, 387
Sciara militaris, 464
Sciomyzidae, 504
Sciophila unimaculata, 462
Scolia bifasciata, S. haemorrhoidalis, 97
Scoliidae, 93, 94 f.
Scoliides, 97 f.
Scolytidae, 294
Scopariidae, 421
Scopariinae, 421
Scopelodes, 401
Scopulipedes, 20, 32 f.
Screw-worm, 512
Scudder, on butterflies, 543
Scutata or Scutati, 546
Scutellerides, 545
Scutelligera, 501
Scutellum, 307, 312, 537
Scutum, 307;
S. proboscidis, 443
Scydmaenidae, 223
Scymnites, 238
Scymnus minimus, 238
Seasonal dimorphism or variation, 335
Seitz, on Syntomid resemblances, 388, 389
Semi-loopers, 415
Seminal duct, 321
Semi-pupa, 271
Semper, on development of wing, 333
Sense-organ, 442, 448;
thoracico-abdominal, 414;
in Uraniidae, 419
Sepsidae, 504
Sericaria mori, 375
Sericteria, 325
Seroot-fly, 482
Serricornia, 189, 213, 255
Sesia, 387;
S. scoliaeformis, 321
Sesiidae, 386, 388, 370 (for Syntomidae), 389
Setae, 534 f.;
aërostatic, 408
Setina, 410
Seventeen-year Cicada, 569
Sex, differences, 92, 95;
production of, 32, 67;
in larva, 325
Sexuparous, 586
Sharp, on classification of Dytiscidae, 213 n.
Sheep, bot-fly, 517;
-tick, 518
Shield, 592
Shoulder, -lappets, 312;
-tufts, 312;
of wing, 316, 319
Siagona, 206
Siculodidae, 423
Siebold, von, on Strepsiptera, 301
Sigara minutissima, 568
Sight, of Pompilus, 104
Silk-glands, 325
Silkworm, 375;
affected by parasitic fly, 507, 508;
Madagascar-, 405
Silpha, 221;
Silpha atrata, S. laevigata, S. lapponica, 222;
S. obscura, 222;
S. opaca, S. thoracica, 222
Silphidae, 221 f., 252, 256
Sima, 168;
S. leviceps, stridulating organ, 169;
S. rufonigra, 169
Simuliidae, 477
Simulium columbaczense, 477
Sinodendron cylindricum, 194
Siphon, 563, 581, 589
Siphonaptera, 522 f.
Siphonophora, 239
Siphunculata, 600
Sitaris, 33;
S. humeralis, 272
Sitodrepa, 247
Skippers, 363
Slave-making ant, 149, 150, 163
Sloth, 430
Slug-worms, 402
Smallest Insect, 228
Smerinthini, 380
Smerinthus populi, 309, 381
Smith, F., on Mellinus arvensis, 123
Snails, enemies of, 205, 222, 510;
parasite of, 495
Social, bees, 35;
wasps, 78, 84
Sociales, 20, 53 f.
Solanum dulcamara, beetle on, 232
Soldiers, 132
Soldier-ant, 150
Solenobia, 395, 430
Solenopsis fugax, 137
Solitary wasps—see Fossores and Eumenidae
Song, of Cicada, 572
Soronia, 232
Sound-organs, 448;
of Ageronia, 354;
of Hecatesia, 371;
of Cicada, 573, 574—see also stridulating organs
Sound-production, 155, 156;
by Aegocera, 411;
by Arctiidae, 410;
by Sphingidae, 382—see also Stridulation
Spalacopsini, 288
Spatula, 459
Spencer, Herbert, on Weismann, 143
Spercheus emarginatus, 218, 219
Spermatheca, 320, 321
Spermophila, 506
Sphaeridiides, 219
Sphaeriidae, 227
Sphaerites, 223
Sphaeritides, 229
Sphaerius acaroides, 227
Sphaerocarides, 279
Sphecia, 387
Sphecius speciosus, 123
Sphecodes, 21, 22;
S. gibbus, 23, 23;
S. rubicundus, 22;
S. subquadratus, 23
Sphegidae, 93, 107 f.
Sphegides, 107 f.
Sphex coeruleus, 110;
S. flavipennis, 108;
S. lobatus, 110;
S. maxillosus, 108
Sphindidae, 245
Sphingidae, 309, 315, 316, 368, 380 f.
Sphinx ligustri, 380
Spider parasite, 490
Spilosoma, 408
Spinneret, 324, 325, 403, 417
Spirachtha, 227
Spiracles, 188, 191;
of Diptera, 449 f.;
of Hippoboscidae, 519;
of Lepidoptera, 313, 314;
of Lipara, 451;
of Nepa, 564;
of Thrips, 528
Spondyliaspis, 581
Spondylidae, 288
Springing plant-lice, 579
Spuler, on nervures, 317 n.
Squama, 448
Squeakers, 209
Staetherinia, 401
Stag-beetles, 193
Stalk, 317, 319
Staphylinidae, 223, 224 f.
Staudinger, Schatz and Röber, on butterflies, 343 n.
Stauronotus maroccanus, 489
Stauropus fagi, 385
Stelis minuta, 29;
S. nasuta, 30, 43;
S. signata, 30
Stelocyttares, 81
Stenamma westwoodi, 159
Stenopteryx hirundinis, 519
Stephostethus, 240
Sternorhyncha, 544
Sterrhopteryx, 394
Stethopathidae, 496
Stigmata—see Spiracles
Stigmatomma, 180
Stigmus pendulus, 128
Sting, 4, 5, 6, 58, 144;
development of, 8, 9
Stinging, 98;
by Calicurgus, 102;
by Pompilidae, 104;
by Sphex, 109
Stingless bees, 61
Stink-gland, 257, 533;
-vessel, 225
Stipes, 309
Stizinae, 123
Stomach, 320
Stomoxys calcitrans, 512
Stratiomyidae, 478
Stratiomys, 452
Straus-Durckheim, on Melolontha, 198
Strawberries, eaten by beetles, 205
Streblidae, 521
Strepsiptera, 189, 298 f.
Streptoperas, 401
Stridulating organ, of Myrmica, 133;
of Heterocerus, 243;
of Passalus-larva, 192;
of Sima leviceps, 169
Stridulation, of ants, 134;
of Corixa, 568;
of Criocerides, 281;
of Dynastides, 199;
of Geotrupes, 195;
of Ipides, 232;
of Lomaptera, 200;
of Longicorns, 287;
of larva of Lucanus cervus, 194;
of Megalopides, 282;
of Melolontha-larva, 198;
of Mutilla, 94;
of Pelobius, 208;
of Phonapate, 246;
of Phyllomorpha, 548;
of Praogena, 264;
of Siagona, 206;
of Trox, 195
Strigil, 568
Striphnopterygidae, 376
Strongylognathus huberi, S. testaceus, 162
Strumigenys, 170
Style, 442
Stylopidae, 298
Stylopised bees, 26
Stylops dalii, 299
Styx infernalis, 340, 358
Suana, 405
Subcostal nervure, 318
Submedian nervure, 318
Suboesophageal ganglion, 320
Sucking-stomach, 311, 449
Suction by Lepidoptera, 311
Suctoria, 526
Supericornia, 546
Swallow-flies, 519
Swarming of wasps, 70 n.
Swarms, 62, 65, 67, 80, 135, 467, 505, 584
Swift-moths, 396
Symbiosis, of ants and plants, 139;
of bee and Acarid, 70.
See also Ants'-nest Insects, and Association
Symbius blattarum, 269
Symphily, 183
Synecthry, 183
Synemon, 371
Synoeca cyanea, nest, 82
Syntelia westwoodi, 229
Synteliidae, 229
Syntomidae, 339 n., 369, 388
Syntomis phegea, 390
Syringe, 535, 536
Syrphidae, 439, 498 f.
Systoechus oreas, 489
Systropus crudelis, 489

Tabanidae, 481, 492

Tabanus, 482
Tachinidae, 507, 514
Tachysphex panzeri, 117
Tachytes, 116;
destroyer of, 275;
T. australis, 113, 117;
T. pectinipes, 117
Tachytides, 116
Taenia, fleas as hosts of, 526
Tajuria diaeus, pupa, 357
Taleporia, 395
Taleporiidae, 430
Tanypezidae, 504
Taphroderides, 296
Tapinoma erraticum, 157
Tarantula-killer, 105
Tarphius, 233
Tarsolepis, 383
Taschenberg, on anatomy of flea, 523 n.
Tascina, 372
Tea-plant bug, 562
Teara melanosticta, 408
Tegula, 71, 187, 307, 311, 312, 447
Tegmina, 539
Teleodont, 193
Telephorides, 248
Telmatophilus, 235
Temnochila coerulea, 232
Temnochilides, 233
Tenebrio molitor, 263
Tenebrionidae, 263
Tenebroides mauritanica, 232
Tentacle, maxillary, 309, 432
Tenthredinidae, parasite of, 4
Terebrantia, 531
Termites, 203, 206, 227, 231
Terrifying attitude, 384
Tesseratomides, 546
Testes, 321, 324, 400, 429;
in larva, 325
Tetanocera ferruginea, 504
Tetanoceridae, 504
Tetragona, 53, 61
Tetramera, 190
Tetramorium caespitum, 160, 163
Tettigometrides, 567
Teucrium, bug and galls on, 550
Thais, pupation of, 363
Thanaos, 342;
T. tages, androconia, 332
Therevidae, 484
Thiridopteryx, 420
Thomas, on androconia, 331
Thorictidae, 236
Thorictus, 236
Thrips, 526 f.;
Thrips lini, 531;
T. secalina, 530
Throscides, 260
Thyatira batis, T. derasa, 386
Thymaridae, 392
Thynnides, 96
Thyreophoridae, 504
Thyrididae, 370, 404
Thyridopteryx ephemeraeformis, 394
Thysanoptera, 526 f.
Tiger-beetles, 201 f.
Tiger-moths, 409
Tillus elongatus, 253, 254
Tinaegeriidae, 370, 387
Tinea, 305;
T. pellionella, 429, 430;
T. vastella, 430;
T. vivipara, 430
Tineidae, 340, 370, 394, 427, 428
Tineodidae, 423
Tineola biselliella, 430
Tingidae, 549
Tipula brobdignagia, 475
Tipulidae, 471 f.;
T. Brevipalpi, 472, 473;
T. Longipalpi, 472, 475
Tipulinae, 475
Tiresias serra, 241
Titanus giganteus, 287
Tithorea, 346
Tomicides, 295
Tomognathus sublaevis, 161
Tongue, 309
Tortoise-shell butterflies, 352
Tortricidae, 340, 395, 427, 432
Tortricina, 395
Toxorrhina, 472
Toxotrypana, 506
Trechus, 205
Trichiini, 200
Trichocera, 473
Trichodes alvearius, T. ammios, T. apiarius, 254
Trichophaga tapetzella, 430
Trichoptera, 306, 425
Trichopterygidae, 227
Trichopteryx fascicularis, 227
Trichroism, 351
Trichterwickler, 294
Trichura, 389, 390
Trictenotomidae, 275
Triecphora, 543
Trigona, 53, 61;
T. carbonaria, 63;
T. crassipes, 65;
T. mosquito, 62
Trimera, 238, 544
Trimeria, 89
Trineura aterrima, 494
Triodites mus, 489
Trioza rhamni, 580
Triphaena, 415
Triphleps, 530
Tritoma bipustulata, 236
Triungulin, 262, 268, 270, 271, 272, 299, 300
Trochalopoda, 543, 544
Trochanter, 307;
divided, 123
Trochilium, 387
Trogini, 195
Trogositidae, 232, 235
Trogosita mauritanica, 232
Tromoptera, 457