請用此 Handle URI 來引用此文件:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/83264
標題: | 智慧製造系統服務創新與擴展之AI模擬與評估模型建置:基於可靠度分析設計發展 Smart Manufacturing Service System Innovation and Expansion through AI Simulation Modeling Application Assessment - System Reliability Analysis Based Design and Development |
其他標題: | Smart Manufacturing Service System Innovation and Expansion through AI Simulation Modeling Application Assessment - System Reliability Analysis Based Design and Development |
作者: | 溫彥俊 Yen-Chun Wen |
指導教授: | 陳文華 Wun-Hwa Chen |
關鍵字: | 工業4.0,微服務系統,可靠度分析,服務系統創新,機器學習,服務主導邏輯, Industry 4.0,Microservice System,Reliability Analysis,Service System Innovation,Machine Learning,Service Dominant Logic, |
出版年 : | 2022 |
學位: | 博士 |
摘要: | 隨著資訊科技技術的發展,智慧製造所應用與涵蓋的概念與技術範疇有著多元的擴展。本論文旨在探討在工業4.0的架構下,企業如何將製造系統進行有效率地規劃,在提供各種服務予顧客時,引入本文所提出之評估機制並導入智慧化模型,從而與客戶進行服務創新與價值共創。此研究將整體框架分為三個部分,首先,在系統建置上,引入微服務系統概念將各種系統服務模組化,相較傳統集中式架構,微服務系統賦予企業更高的靈活性與自由度。其次,應用服務主導邏輯(Service-Dominant Logic),將各系統與服務的組成進行動態調整。在創新的同時,即時進行檢核與修正,從而改善企業服務建置效率。最終,在與相關利害關係者共同評估後,企業可依模型的建構與系統模擬結果進行相應決策。在不同的場景分析下提出多元模擬架構,定義服務創新。本文以NASDAQ上市上櫃跨國科技製造公司為例,基於可靠度分析進行模組化微服務系統架構的設計、建置與擴充,擷取相應數據並以領域知識分析。應用Kaplan-Meier估計法與多層感知網絡演算法(MLP),進行服務品質分析與預測,並經由大數據蒐集與彙整,從而進行模型檢驗與優化。在評估系統中,本論文共引入330,837筆資料集,基於案例公司的評估數據模型得出該算法之均方根誤差為0.00024。透過模擬模型與各分析情境的建置,服務品質依該系統建議共提升約12%。最終,透過探尋不同可能性之評估,在降低試誤成本的同時,達成幫助企業擴展自身價值主張與創新商業模型的目標。 This paper proposed a theoretical framework to induct and define the development procedure in various application environments and scenarios. The concept of microservice and service-dominant logic, correlated to the service process and quality specified in the framework, can assist higher managers in dealing with engineering and business development problems. Integrating multilayer perceptron (MLP) and Kaplan-Meier estimation assessment approaches with the dynamic model construction, the simulation system considered 8D, PFMEA/DFMEA and a large amount of relevant data (330,837 datasets) help to increase the accuracy of the analysing model. For innovation, the quality and test-driven design (TDD) accompanied by different ideas and insights make firms consider its alter cases for market expansions. This paper takes a NASDAQ-listed firm and its ecosystem as a case study, trying to define an innovation process with smart-manufacturing approaches based on reliability assessment design. After introducing the system framework into the original construction, the implemented microservice configuration helped the firm to increase by 12% of its service quality. For system validation, the MSE of the prediction model is 0.00024 and the simulation structure also verified by posterior data. The NASDAQ-listed firm can expand its value proposition by implementing the validated analysing system. Furthermore, the prosed framework and assessment methodology can not only help program development but also can assist in dealing with different small cases or scenarios. |
URI: | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/83264 |
DOI: | 10.6342/NTU202300139 |
全文授權: | 同意授權(限校園內公開) |
電子全文公開日期: | 2027-12-07 |
顯示於系所單位: | 商學研究所 |
文件中的檔案:
檔案 | 大小 | 格式 | |
---|---|---|---|
ntu-111-1.pdf 目前未授權公開取用 | 3.42 MB | Adobe PDF | 檢視/開啟 |
系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。