WEKO3
アイテム
スマートマニュファクチャリング用ビッグデータアナリティクスの開発
https://doi.org/10.19000/0002000555
https://doi.org/10.19000/000200055504b38f57-b557-4fe7-9f42-9a48acc3e61a
| 名前 / ファイル | ライセンス | アクション |
|---|---|---|
Thesis- Saman Fattahi-March 2023-10.pdf (3.3 MB)
|
|
| Item type | 学位論文 / Thesis or Dissertation(1) | |||||||
|---|---|---|---|---|---|---|---|---|
| 公開日 | 2024-01-09 | |||||||
| タイトル | ||||||||
| タイトル | Developing Big Data Analytics for Smart Manufacturing | |||||||
| 言語 | en | |||||||
| タイトル | ||||||||
| タイトル | スマートマニュファクチャリング用ビッグデータアナリティクスの開発 | |||||||
| 言語 | ja | |||||||
| 言語 | ||||||||
| 言語 | eng | |||||||
| 資源タイプ | ||||||||
| 資源 | http://purl.org/coar/resource_type/c_db06 | |||||||
| タイプ | doctoral thesis | |||||||
| ID登録 | ||||||||
| ID登録 | 10.19000/0002000555 | |||||||
| ID登録タイプ | JaLC | |||||||
| アクセス権 | ||||||||
| アクセス権 | open access | |||||||
| アクセス権URI | http://purl.org/coar/access_right/c_abf2 | |||||||
| 著者 |
Saman Fattahi
× Saman Fattahi
|
|||||||
| 抄録 | ||||||||
| 内容記述タイプ | Abstract | |||||||
| 内容記述 | Big data means horizontally networked yet independent data systems containing a vast number of structured and unstructured datasets. Statistical and logical computational arrangements (referred to as big data analytics) must be installed to make sense of big data. Like human-cyber-physical systems, digital twins, artificial intelligence, the internet of things, and sustainability, big data and big data analytics are essential elements of the fourth industrial revolution or smart manufacturing. In this study, big data of manufacturing processes are categorized into three main issues: 1) datasets for digital twin; 2) control variable-evaluation variable-centric datasets; 3) graphical datasets. This thesis considers the problem of developing big data and big data analytics for smart manufacturing, focusing on three related issues. Issue 1: digital twins of manufacturing phenomena are supposed to machine-learn the required knowledge using relevant datasets available in big data. Therefore, a research question is how to preprocess manufacturing phenomena-relevant datasets for using them directly in digital twins. Issue 2: Big data and analytics require expensive resources and sophisticated computation arrangements. Thus, big data hardly benefits small and medium-sized manufacturing organizations, resulting in “big data inequality.” Consequently, a research question is how to eliminate big data inequality. Issue 3: big data is often visualized using several two-dimensional plots (graphical dataset). These plots are then used to make a decision informally. Consequently, a research question is how to make formal decisions by computing two-dimensional plots, not numerical data. Thus, this thesis is organized as follows. Chapter 1 presents this thesis's background, objective, scope, and limitations. It also presents a comprehensive literature review on big data relevant to smart manufacturing. Chapter 2 describes the proposed big data analytics framework showing all the subsystems. In this chapter, the functional requirements of the subsystems are explained. big data analytics is developed for manufacturing process-relevant decision-making. The proposed analytics consists of five integrated systems: 1) big data preparation system, 2) big data analytics exploration system, 3) data visualization system, 4) data analysis system, and 5) knowledge extraction system. The big data analytics preparation system prepares contents that exhibit the characteristics of digital manufacturing commons. Chapter 3 deals with Issue 1. A digital twin consists of five modules (input, modeling, simulation, validation, and output modules), and big data must supply datasets for building these modules. This chapter presents a manufacturing phenomenon-related datasets preprocessing method considering the four modules of digital twins (input, modeling, simulation, and validation modules). As an example, the preprocessing of surface roughness-relevant datasets is considered. Chapter 4 deals with Issue 2. This chapter described the developed big data analytics framework for the control variable-evaluation variable-centric dataset. This system can support user-defined ontology and automatically produces Extensible Markup Language-based datasets. The big data exploration system can extract relevant datasets prepared by the first system. The system uses keywords derived from the names of manufacturing processes, materials, and analyses- or experiments-relevant phrases (e.g., design of experiment). The third system can help visualize relevant datasets extracted by the second system using suitable methods (e.g., scatter plots and possibility distribution). The fourth system establishes relationships among the relevant control variables (variables that can be adjusted as needed) and evaluation variables (variables that measure the performance) combinations for a given situation. In addition, it quantifies the uncertainty in the relationships. The last system can extract knowledge from the outcomes of the fourth system using user-defined criteria (e.g., minimize surface roughness and maximize material removal rate). The efficacy of the proposed big data analytics is demonstrated using a case study where the goal is to determine the right states of control variables of dry electrical discharge machining for maximizing material removal rate. It is found that the proposed big data analytics is transparent and free from big data inequality. Chapter 5 deals with Issue 3. Big data analytics is developed to compute two-dimensional plots (graphical datasets) generated from big data. The efficacy of the tool is demonstrated by applying it to assess sustainability in terms of Sustainable Development Goal 12 (responsible consumption and production). Regarding that, engineering materials' functional, economic, and environmental issues play a vital role. Accordingly, three two-dimensional plots generated from big data of engineering materials are computed using the proposed analytics. The plots refer to six criteria (strength, modulus of elasticity, cost, density, CO2 footprint, and water usage). The proposed analytics correctly rank the given materials (mild steel, aluminum alloys, and magnesium alloys). Chapter 6 describes future research directions and discusses the implication of this study from the viewpoint of smart manufacturing. Finally, Chapter 7 provides the concluding remarks of this thesis. |
|||||||
| 言語 | en | |||||||
| 書誌情報 |
p. 1, 発行日 2023-03 |
|||||||
| 著者版フラグ | ||||||||
| 言語 | en | |||||||
| 値 | ETD | |||||||
| 学位名 | ||||||||
| 言語 | ja | |||||||
| 学位名 | 博士 （工学） | |||||||
| 学位授与機関 | ||||||||
| 学位授与機関識別子Scheme | kakenhi | |||||||
| 学位授与機関識別子 | 10106 | |||||||
| 言語 | ja | |||||||
| 学位授与機関名 | 北見工業大学 | |||||||
| 学位授与番号 | ||||||||
| 学位授与番号 | 甲第207号 | |||||||
| 研究科・専攻名 | ||||||||
| 研究科・専攻名 | 生産基盤工学専攻 | |||||||
| 学位授与年月日 | ||||||||
| 学位授与年月日 | 2023-03-17 | |||||||
