权益分类	功能权益	普通用户	{{item.name}}会员
{{category.name}}	{{benefitItem.name}}

創発型計算環境Artificial Cellular Machine:ACMの構築

应急计算环境Artificial Cellular Machine:ACM的构建

基本信息

批准号：
13780296
负责人：
遠藤聡志
金额：
$ 1.34万
依托单位：
University of the Ryukyus
依托单位国家：
日本
项目类别：
Grant-in-Aid for Young Scientists (B)
财政年份：
2001
资助国家：
日本
起止时间：
2001 至 2002
项目状态：
已结题

项目摘要

平成14年度は,次の研究を行い,以下の成果を得た.課題1:セルオートマトン(CA)の遷移規則学習アルゴリズムの改良課題2:セルオートマトンモデルの計算環境の一部である,CA専用インタプリタの実装課題3:セルオートマトンモデルの計算環境の一部である,CA専用コンパイラの開発課題4:CAを用いた応用研究課題1の成果:13年度に開発した共進化型学習アルゴリズムに,種の概念を導入し,クラスタリングによる種分類から,先見知識を必要としない汎用的遷移則学習アルゴリズムを開発した.また,ルール分析法を提案し,「密度分類タスク」について,得られたルールを分析した.この結果,人による遷移ルールに比べて,きわめて複雑な構造をもつルールを獲得していることが分かった.課題2の成果:13年度の試作システムに,GUI部,誤り検出部を加え実装した.CAモデルの遷移則を試行錯誤的に開発する際の,有効な支援ツールとして利用できる.課題3の成果:MMXテクノロジによる並列化コードを生成するCA用コンパイラの開発を行った.PCに標準的に用意されているSIMDを,アセンブラの知識を必要とせずに利用可能である.課題4:開発した計算環境により,交通網システムをCA上でモデル化し,交差点設計に関する応用研究を行った.この研究では,交差点の制御系として十字路型とロータリー型が自動生成され,従来の交通網で採用されている交差点の形式を獲得できた.本アプローチの妥当性を検証することができた.

Heisei 14 years ago, the second time the study was carried out, the following results were obtained. Topic 1: Improvement of migration rule learning for CA Topic 2: Implementation of migration rule learning for CA Topic 3: Development of migration rule learning for CA Topic 4: Development of migration rule learning for CA Topic 1: Development of migration rule learning for CA Topic 2: Implementation of migration rule learning for CA Topic 3: Development of migration rule learning for CA Topic 4: Development of migration rule learning for CA Topic 1: Development of migration rule learning for CA Topic 3: Development of migration rule learning for CA Topic 4: Development of migration rule learning for CA Topic 1: Development of migration rule learning for CA Topic 3: Development of migration rule learning for CA Topic 4: Development of migration rule learning for CA Topic 3: Development of migration rule learning for CA Topic 4: Development of migration rule learning for CA Topic 1: Development of co-evolution learning for CA Topic 3: Introduction of migration rule learning for CA Topic 4: Development of migration rule learning for CA Topic 3: Development of co-evolution learning for CA Topic 4: Development of migration rule learning for CA Topic 4: Development of migration rule learning for CA Topic 3: Development of migration rule learning for CA Topic 4: Development of migration rule learning for CA Topic 4: Development of migration rule learning For example, the density classification method is proposed, and the density classification method is used to analyze the density classification method. As a result, the migration of human beings is more complicated than that of human beings, and the structure of human beings is more complicated than that of human beings. Results of Task 2: During the trial operation in Y2013, GUI unit, error detection unit were added to the installation. CA unit was transferred to the trial error detection unit, and there was support for the application. Results of Task 3: MMX technology development, generation, development, PC technology development, SIMD technology development, PC technology development, and PC technology development. Topic 4: Research on the application of traffic network design in developing computing environment This research is to control the system of intersection point and to obtain the intersection point form by means of the automatic generation of the intersection type and the traffic network. This is the first time I've ever seen you.