2022 Fiscal Year Final Research Report
Developing Fast Algorithms based on Advanced Evolutionary Algorithms with Machine Learning and Applying to Real-time based Manufacturing and Logistics Systems
Project/Area Number |
19K12148
|
Research Category |
Grant-in-Aid for Scientific Research (C)
|
Allocation Type | Multi-year Fund |
Section | 一般 |
Review Section |
Basic Section 61040:Soft computing-related
|
Research Institution | Fuzzy Logic Systems Institute |
Principal Investigator |
Gen Mitsuo 一般財団法人ファジィシステム研究所, 研究部, 特別研究員 (20095003)
|
Co-Investigator(Kenkyū-buntansha) |
川上 浩司 京都先端科学大学, 工学部, 教授 (90214600)
|
Project Period (FY) |
2019-04-01 – 2023-03-31
|
Keywords | 進化算法 / 遺伝的アルゴリズム / 粒子群最適化 / 機械学習 / 生産スケジューリング / 持続可能なサプライチェイン |
Outline of Final Research Achievements |
Various devices of semiconductor products are required to be processed within a specification processing time in the real-time based manufacturing system, and if they cannot be processed within the specified time to them, they become defective products and affect production efficiency. The purpose of this research is to develop a high-speed algorithm that utilizes advanced evolutionary computation and machine learning, and apply it to various optimization problems in production and logistics systems. We published the proposed algorithms and experimental results with the effectiveness in the several international journals. Application research to flexicible jobshop scheduling problem and sustainable closed-loop supply chain model are NP-hard problems, and we developed advanced hybrid evolutionary algorithms for FJSP in the production / distribution system, time scheduling of high-speed train, and VRP problems and demonstrated the effectiveness with experiments in the several journals.
|
Free Research Field |
ソフトコンピューティングと生産・物流問題への応用
|
Academic Significance and Societal Importance of the Research Achievements |
一般に半導体製品は各種素子から構成され,生産工程は実時間スケジューリング問題であり,特に各種素子は所定の処理時間内の要求仕様の必然性から,制約時間内で処理できない時は不良品となり生産効率に影響を及ぼす。本研究の目的は先端的進化計算と機械学習を活用した高速スケジューリング算法を開発し,生産・物流システムや高速列車の時刻管理の最適化問題等に応用することである。ハイブリッド型多目的進化計算ベースのアルゴリズム開発研究:ハイブリッド協調進化算法(CoEA+PSO)を開発し,共有リソースのスケジューリング問題解法に応用し,更に差分進化(DE)を組み込んだ手法の有効性を明らかにし,国際誌に掲載した。
|