-
坪井 昭
2001 年 121 巻 1 号 p.
1
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
-
大松 繁
2001 年 121 巻 1 号 p.
2
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
-
平田 廣則
2001 年 121 巻 1 号 p.
3
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
-
坪井 昭
2001 年 121 巻 1 号 p.
4
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
-
秋月 影雄
2001 年 121 巻 1 号 p.
5-6
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
-
舟橋 拓夫
2001 年 121 巻 1 号 p.
7-8
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
-
矢嶋 弘義
2001 年 121 巻 1 号 p.
9-10
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
-
茅 陽一
2001 年 121 巻 1 号 p.
11-14
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
-
甘利 俊一
2001 年 121 巻 1 号 p.
15-18
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
-
小田 俊理
2001 年 121 巻 1 号 p.
19-22
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
-
福田 宏
2001 年 121 巻 1 号 p.
23-26
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
-
中井 貞雄
2001 年 121 巻 1 号 p.
27-28
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
-
藤井 信生
2001 年 121 巻 1 号 p.
29-30
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
-
柿倉 正義
2001 年 121 巻 1 号 p.
31-32
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
-
三森 定道
2001 年 121 巻 1 号 p.
33-34
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
-
河合 洋一, 井上 節夫
2001 年 121 巻 1 号 p.
35-38
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
-
神成 文彦
2001 年 121 巻 1 号 p.
39
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
-
高木 茂孝
2001 年 121 巻 1 号 p.
40-41
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
-
小圷 成一
2001 年 121 巻 1 号 p.
42
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
-
医用生体工学の視点から
椎名 毅
2001 年 121 巻 1 号 p.
43-44
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
-
橋本 岳
2001 年 121 巻 1 号 p.
45-46
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
-
大川 剛直
2001 年 121 巻 1 号 p.
47
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
-
石黒 章夫
2001 年 121 巻 1 号 p.
48-49
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
-
黄瀬 浩一
2001 年 121 巻 1 号 p.
50-51
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
-
小林 真也
2001 年 121 巻 1 号 p.
52
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
-
原 元司
2001 年 121 巻 1 号 p.
53
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
-
山本 透
2001 年 121 巻 1 号 p.
54
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
-
鈴木 茂行
2001 年 121 巻 1 号 p.
55
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
-
錦織 昭峰
2001 年 121 巻 1 号 p.
56-57
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
-
村井 純
2001 年 121 巻 1 号 p.
58-60
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
-
小林 容子, 相吉 英太郎
2001 年 121 巻 1 号 p.
61-67
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
A search method based on genetic algorithms (GA) using the deterministic operators has been developed to generate optimized boiling water reactor (BWR) loading patterns (LP). The search method uses an improved GA operator, that is crossover, mutation, and selection. The handling of encoding technique and constraint conditions is designed so that the GA reflects the peculiar characteristics of the BWR. In addition, some strategies such as
elitism and self-reproduction are effectively used to improve the search speed. LP evaluations were performed with a three-dimensional diffusion code that coupled neutronic and thermal hydraulic models. Strong axial heterogeneities and three-dimensional-dependent constraints have always necessitated the use of three-dimensional core simulators for BWRs, so that an optimization method is required for computational efficiency.
The proposed algorithm is demonstrated by successfully generating LPs for an actual BWR plant applying the Haling technique. In test calculations, candidates that shuffled fresh and burned fuel assemblies within a reasonable computation time were obtained.
抄録全体を表示
-
明日 香昌, 駒谷 喜代俊
2001 年 121 巻 1 号 p.
68-75
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
For train simulations, there is a well-known train movement model. By using the model, the precise running status of a train at the designated values for the velocity, position, and time, must be obtained to analyze the train movement in detail. However, conventional simulation methods cannot calculate the precise status at the designated value, because they only calculate the running status discretely at fixed interval. To resolve this problem, we propose a simulation method that has two features. First, the method utilizes integral formulas. An expression of velocity is formulated that directly gives the precise running status at the designated velocity. With the expression, we can also formulate expressions for position and time. Second, since the designated values for the three variables are given at one time, we build a framework that contains
the above three expressions and performs the calculations in the appropriate order to get the running status corresponding to a designated value firstly fulfilled. The proposed simulation method calculates the required running status without considering errors introduced by the method itself. It also uses a smaller number of calculations than conventional methods. The proposed simulation has various uses from planning to control of railways to which the conventional method has been applied.
抄録全体を表示
-
村野 亘, 相吉 英太郎
2001 年 121 巻 1 号 p.
76-82
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
Asynchronous Neural Networks as optimizing machines for the 0-1 combinatorial optimization problems have the problem that state transition is trapped at one of local optimal solutions in a neighborhood with a radius of one hamming distance. In order to get off such local minima, the Boltzmann Machine which state transition is probabilistic was proposed. On the other hand, the Hysteresis Machine which state transition acceptes increase of the minimizing function value within a certain range deterministically was developed. Moreover, according to the schedule to search global optimal solution, the Annealed Hysteresis Machine, which hysteretic width decreases monotonously, and the Chaotic Hysteresis Machine, which hysteretic width fluctuates chaotically, were proposed, and effectiveness for a few 0-1 combinatorial optimization problems was demonstrated. In this article, the chaotic hysteresis is introduced in the Multi-Valued Neural Network for integer programming problems. Moreover, the simulation results for a few integer programming problems demonstarate effectiveness of the Multi-Valued Chaotic Hysteresis Machine.
抄録全体を表示
-
中村 政俊, 山中 慎司, 後藤 聡, 久良 修郭
2001 年 121 巻 1 号 p.
83-89
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
Dual mode synchronous positioning control with switching master and slave axes is proposed to realize accurate contour control of industrial robot arms. Disturbance contaminated axis is switched to the master axis, then each of the disturbance in the first axis or the second axis can be compensated. Effectiveness of the proposed method was assured by the experimental results of an actual articulated robot arm.
抄録全体を表示
-
中村 政俊, 小畑 昌之, 後藤 聡, 久良 修郭
2001 年 121 巻 1 号 p.
90-97
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
In this paper, multi-dimensional modified taught data method was designed by considering interference of links of articulated robot arm. Effective of the proposed method was assured by simulation study and experiments of an articulated robot arms. The proposed method can bring satisfactory control performance for industrial articulated robot arms with interference of each links.
抄録全体を表示
-
平澤 宏太郎, 衛藤 広宣, 胡敬 炉, 村田 純一, 熊 慶宇
2001 年 121 巻 1 号 p.
98-105
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
Universal Leaning Network (ULN) which is a super set of supervised learning networks has been already proposed. Parameters of ULNs are trained in order to optimize a criterion function as conventional neural networks, and after training they are fixed as constant parameters. In this paper a new method to alter the parameters, therefore in a special case, to control the branch connection depending on the network flows is presented to enhance flexibility of the networks. In the proposed method, there exist two kinds of networks, the first one is a basic network which includes varying parameters and the other one is a network which cal-culates the optimal varying parameters, therefore decides the branch connection in a special case depending on the network flows of the basic network.
From simulations where parameters of a neural network are altered and branch connection in the neu-ral network is determined by a fuzzy inference network, it is shown that the proposed network has higher representation abilities than the conventional networks.
抄録全体を表示
-
岸田 一也, 福元 伸也, 宮島 廣美
2001 年 121 巻 1 号 p.
106-111
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
We propose a learning method of fuzzy inference rules using vector quantization, neural gas network. Some models using self-organization or vector quantization by neural networks are proposed in recent stud-ies. These models show good results as for the number of fuzzy inference rules in high dimensional problems. However, most of these models determine a distribution of initial fuzzy inference rules by considering only input data. In this paper, so as to make a more proper distribution of the initial fuzzy inference rules in input space, we propose a method considering not only input data but output data. Further, the number of fuzzy inference rules is determined to an objective value(threshold of inference error) by a constructive way. In order to demonstrate the validity of the proposed method, some numerical examples are performed.
抄録全体を表示
-
満倉 靖恵, 福見 稔, 赤松 則男
2001 年 121 巻 1 号 p.
112-117
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
In this paper, we propose a new method to examine whether or not human faces are included in color images by using a lip detection neural network (LDNN) and a skin distinction neural network (SDNN). In conventional methods, if there are the same color as scenes, the domain which is accepted as not only the skin color but any other color can be searched. However, first, the lip are detected by the LDNN in the pro-posed method. Next, SDNN is utilized to distinguish the skin color from the other. The proposed method can obtain relatively high recognition accuracy, since it have the double recognition structure of LDNN and SDNN.
Finally, in order to demonstrate the effectiveness of the proposed scheme, computer simulations were performed. First, 100 lip color, 100 skin color and 100 background pictures, which are picture changed into 10×10 pixels, are prepared for training. The validity was verified by testing images containing several faces.
抄録全体を表示
-
小松 利彰, 土方 嘉徳, 才脇 直樹, 西田 正吾
2001 年 121 巻 1 号 p.
118-126
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
User support by a computer is desired for creating computer graphics. Automatic generation of moving human crowd is one of such demands. The feature of human crowd is that the behavior of each figure is independent, but the crowd, as a huge mass of it, has a pattern as a whole. It is too troublesome if the user inputs the all coordinates data of this behavior. This paper describes the modelling method using chaos and electric charge model. Chaos is convenient to represent such features. Electric charge model can represent the detail of each figure's movement.
抄録全体を表示
-
村田 純一
2001 年 121 巻 1 号 p.
127-133
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
Artificial neural networks with input gates are proposed which can select relevant input variables to the networks. An input is relevant, if it contains inherent and significant information which is not shared with the other inputs and indispensable in determining the output values. By selecting the relevant inputs, or equivalently by removing irrelevant inputs, one can obtain neural networks with higher generalization abilities and interpretabilities. The proposed network first removes redundant inputs that can be replaced by the other inputs, then its input gates placed at its input channels determine whether each of the remained inputs significantly contribute to the output or not. The gates open and close depending on their trained parameters and current input values to the network. Thus, they can determine the relevant inputs dynam-ically (depending on the current input values) and automatically (based on learning). Numerical examples are provided to demonstrate that the network can successfully select the relevant inputs.
抄録全体を表示
-
遺伝的アルゴリズムによる製造誤差への適応
村川 正宏, 河西 勇二, 安達 敏男, 高須賀 馨, 吉澤 修治, 樋口 哲也
2001 年 121 巻 1 号 p.
134-141
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
We propose a concept of evolvable analog LSIs and apply it to Intermediate Frequency (IF) filters, which are widely used in cellular phones. The developed chip can correct discrepancies in the values of analog circuit components by genetic algorithms (GAs), which have attained (1) yield rate of 97% (2) 60% reduction of the filter area (3) 38% reduction of power dissipation, compared with AKM commercial products. When analog Integrated Circuits (ICs) and Large-Scale Integrated Circuits (LSIs) are manufactured, the values of the analog circuit components, such as resistors or capacitors, often vary from the precise design specifications. Analog LSIs with such defective components can not perform at required levels and thus have to be discarded. However, using our technique can improve yield rates of analog LSIs and merit from two more advantages, namely: smaller circuits and less power dissipation, which can lead to cost reductions and efficient implementation of analog LSIs. This approach could be applied to a wide variety of analogue LSIs.
抄録全体を表示
-
池本 和生, 野澤 博, 仲 篤起, 才脇 直樹, 西田 正吾
2001 年 121 巻 1 号 p.
142-149
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
This paper deals with data management structures for spatio-temporal walkthrough in virtual 3D environment, in which large scale spatio-temporal data are processed. In the spatio-temporal walkthrough, it is very important to realize real-time response in query process from large scale spatio-temporal database. For this purpose, we propose Adaptive Tree structure(AT structure), in which either spatial data structure or temporal data structure is selected adaptively depending on the demand of search. This paper describes an outline of the proposed data structure and evaluation results by computer simulation.
抄録全体を表示
-
飯間 等, 三宮 信夫
2001 年 121 巻 1 号 p.
150-156
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
In this paper, an effective design of genetic algorithm is proposed for a large scale flow shop scheduling problem with a long scheduling period. The objective of this problem is to minimize the sum of the tardiness for each product. Because the convergence speed in genetic algorithm is slow for large scale problems, a good solution can not be obtained under the limitation of the computation time. In such a case the size of the search space in genetic algorithm should be decreased. Thus three methods are proposed by introducing a decomposition procedure for solving the large scale scheduling problem. In all of the methods the set of products is decomposed into several groups in the order of the due date. This decomposition procedure is based on an idea that a product with an earlier due date should be processed earlier. As a case study a scheduling problem for an electric wire production process is considered. The effectiveness of the proposed methods is examined by a numerical computation carried out on the basis of real operation data.
抄録全体を表示
-
バックワード形式によるアプローチ
一見 宜央, 飯間 等, 三宮 信夫, 小林 靖宜
2001 年 121 巻 1 号 p.
157-166
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
This paper deals with a scheduling problem for a large scale job-shop process. This problem has complicated constraints such as the precedence conditions among jobs, the existence of parallel machines and multi-function machines, and the working hours. For solving this problem, we propose an autonomous decentralized scheduling algorithm in which the schedule of each operation is determined backwards. In this algorithm, a number of decision makers called modules cooperate with one another in order to attain the goal of the overall system. They determine the scheduling plan on the basis of their cooperation and the sat-isfaction of their own objective function levels. Furthermore, a priority rule is proposed for the case where a product is given priority. The effectiveness of the proposed algorithm is examined through the computational result and the evaluation from a production manager in the job site.
抄録全体を表示
-
松本 心, 顔 玉玲, 金城 寛, 山本 哲彦
2001 年 121 巻 1 号 p.
167-176
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
In this paper, a diagnosis method for machine faults using a neural network based on autocorrelation coefficients of wavelet transformed signals is presented. It is important for factory engineers to accurately estimate machine faults. In conventional diagnosis methods, frequency analysis using the fast Fourier trans-form (FFT) has often been employed. Recently, wavelet transforms have been studied and applied to many signal-processing applications. Wavelet transforms are very useful because of characteristics of time-frequency analysis. In this paper, we propose an application of wavelet transforms to machine fault diagnosis. In order to apply wavelet transforms to machine fault diagnosis, we use autocorrelation coefficients of the wavelet transformed signal. In this research, it becomes clear that the autocorrelation coefficients represent the dif-ferent classes of machine states. For the automatic diagnosis, we trained a neural network to recognize three classes of machine states based on the autocorrelation coefficients of wavelet transformed signals. Simulation and experimental results show that the trained neural network could successfully estimate machine faults.
抄録全体を表示
-
平澤 宏太郎, 吉田 英正, 胡 敬炉
2001 年 121 巻 1 号 p.
177-186
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
Recently, numerous attempts have been made by researchers to understand the essence of complex phe-nomena (complex systems). In this paper, we paid attention to biological systems in the nature as it is one of the most complex systems. The purpose of this paper is to propose a method to realize symbiotic phe-nomena such as mutual benefit, competition and exploitation more generally than Lotka-Volterra equation using neural networks.
抄録全体を表示
-
白石 優旗, 平澤 宏太郎, 胡 敬炉, 村田 純一
2001 年 121 巻 1 号 p.
187-195
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
Stuart Kauffman explored the law of self-organization in random Boolean networks, and Kosaku Inagaki also did it in neural networks partially. The aim of this paper is to show that probabilistic neural networks (PNNs), which are recurrent networks and are controlled by a probabilistic transision rule based on Boltzh-mann machine, hold the order, even though we determine the weights, the thresholds, and the connections between neurons randomly. And, we also studied the deterministic transient neural networks which are the special networks of PNNs.
抄録全体を表示
-
竹田 史章, 西蔭 紀洋
2001 年 121 巻 1 号 p.
196-205
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
Up to now, Authors' research have been focussed on applications for banknotes recognition on neural networks (NNs). Especially, we have clarified its effectiveness of neuro-recognition system on industrial use by implementing the neuro-recognition system to various banking machines in spite of unfocussed sensing data and various noised data. Recently, another demand on banknotes recognition will be appeared. Namely, it is needed the more number of banknotes to be recognized at the same time on the banking machines according to appearance of Euro banknotes. In order to solve this problem, we propose a neuro-templates matching method whose templates are constructed by plural NNs. We show its effectiveness by experiments for various foreign banknotes and Euro dummy. Furthermore, we also refer extension of the proposed method to human security using bio-parametric information such as palm prints, finger prints, face image and so on.
抄録全体を表示
-
池上 輝哉, 王 旗, 才脇 直樹, 西田 正吾
2001 年 121 巻 1 号 p.
206-211
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
In this paper we describe how to acquire sequential motions of characters modeled as 3D rnass-Spring Mod-el to combine global deformations of the lattice in the Controller Synthesis approach. A motion controller, which can represent the whole process of deformation, determines a, sequence of several simple motions. We use Genetic Algorithm to generate and sort motion controllers based on calculated evaluation values. Those high valued ones will be picked up. Futhermore, we describe the environment of computer aided motion acquisition in which we can acquire the motion beyond our imagination.
抄録全体を表示
-
福見 稔, 満倉 靖恵, 赤松 則男
2001 年 121 巻 1 号 p.
212-217
発行日: 2001/01/01
公開日: 2008/12/19
ジャーナル
フリー
In this paper a method to generate universal rukes by using a genetic algorithm with virus infection is presented. This method utilizes a modular structure in GA. Each module learns a different neural network architecture, such as a sigmoid and a higher order (GMDH-type) neural networks. Those information is communicated to the other modules by the virus infection. This GA can generate different kinds of rules for different problems by using the same algorithm. The results of computer simulations show that this approach can generate obvious network structures and as a result simple rules for a pattern classification and a chaotic identification problems.
抄録全体を表示