應用資料探勘建構品質保證矩陣-以PCB產業為例

NCUIR > School of Management at National Central University > Executive Master of Industrial Management > Electronic Thesis & Dissertation > Item 987654321/83174

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/83174

Title:	應用資料探勘建構品質保證矩陣-以PCB產業為例
Authors:	葉晏豪;Ye, Yan-Hao
Contributors:	工業管理研究所在職專班
Keywords:	工業 4.0;智慧工廠;資料探勘;品質保證矩陣;QFD
Date:	2020-07-20
Issue Date:	2020-09-02 15:05:16 (UTC+8)
Publisher:	國立中央大學
Abstract:	製造業正面臨人口老化、人力不足、工資成本提高、生產產品多樣性、產品週期縮短、全球市場需求劇烈變動等等問題浮現，因此誰能靈敏反應市場變化、快速生產多樣性的商品，誰就是市場的贏家。為了克服此問題製造業積極導入工業4.0逐步往智慧(智能)工廠發展，工廠物(機)聯網、虛實整合、智慧(智能)設備與機器人的應用，將多類設備都能夠互相連網互相溝通，使生產流程更加靈活，以協助製造業解決經營困境並提昇競爭力。智慧工廠已成為現代製造工業趨勢，但在製造生產過程中大量數據的蒐集，這會導致製造管理方式與現況不同，已難以再透過人工方式，在短時間進行處理、分析並整理成為有效資訊，其難以管理之主因有資料量、多樣性、資料輸入輸出的速度以及真實性，管理者已難以即時下達有效對策。因此智慧工廠中擁有巨量數據建構多類資料庫，管理者如何應用資料探勘（Data Mining），快速獲得有價值資訊進行判斷下達對策，或是設備透過機器學習(Machine Learning)方式，快速找出最佳方法快速調整，已經是現代管理者必學課程。本研究針對印刷電路板（Printed Circuit Board，PCB）產業以U公司為例，探討如何透過機聯網建置智慧工廠，利用所收集的大量數據於資料庫中，透過巨量資料應用資料探勘技術，提供現場操作人員快速獲得有效資訊進行調整最佳生產參數，提高生產力與效率。 ;The manufacturing industry is facing the problems of aging population, insufficient manpower, rising wage costs, diversity of production products, shortening of product cycles, drastic changes in global market demand, etc., so who can sensitively respond to market changes and quickly produce diverse commodities, who is The winner of the market. In order to overcome this problem, the manufacturing industry actively introduces Industry 4.0 and gradually develops into smart factories. The factory′s physical networking, virtual and real integration, and the application of smart devices and robots will enable multiple types of devices to be connected to each other and communicate with each other. To make the production process more flexible, to assist the manufacturing industry to solve business difficulties and enhance competitiveness. Smart factories have become a modern manufacturing industry trend, but the collection of large amounts of data in the manufacturing process will cause the manufacturing management method to be different from the current situation. It is no longer possible to manually process, analyze and organize it into effective information in a short time. The main reason for the difficulty in managing is the amount of data, diversity, speed and authenticity of data input and output, and it is difficult for managers to issue effective countermeasures in real time. Therefore, there are huge amounts of data in smart factories to construct multiple types of databases. How can managers use Data Mining to quickly obtain valuable information to judge and issue countermeasures, or the equipment can quickly find the best method and quickly adjust through Machine Learning. Modern management must learn courses. This study is directed to Printed Circuit Board and PCB industry. Take U company as an example to discuss how to build a smart factory through machine networking, use the large amount of data collected in the database, and apply data exploration technology through huge amounts of data to provide on-site operators with rapid Obtain effective information to adjust the best production parameters and improve productivity and efficiency.
Appears in Collections:	[Executive Master of Industrial Management] Electronic Thesis & Dissertation

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	249	View/Open

社群 sharing

Loading...