| DC 欄位 |
值 |
語言 |
| DC.contributor | 工業管理研究所 | zh_TW |
| DC.creator | 陳泰成 | zh_TW |
| DC.creator | Tai-Cheng Chen | en_US |
| dc.date.accessioned | 2021-7-6T07:39:07Z | |
| dc.date.available | 2021-7-6T07:39:07Z | |
| dc.date.issued | 2021 | |
| dc.identifier.uri | http://ir.lib.ncu.edu.tw:88/thesis/view_etd.asp?URN=108426022 | |
| dc.contributor.department | 工業管理研究所 | zh_TW |
| DC.description | 國立中央大學 | zh_TW |
| DC.description | National Central University | en_US |
| dc.description.abstract | 自德國提出工業4.0後，各個國家包括美國先進製造夥伴聯盟（Advanced Manufacturing Partnership, AMP）再工業化合作策略，以及大陸、韓國和日本等新興國家積極發展物聯網和智慧製造等技術。台灣也於2015年推動生產力4.0計畫。全球國家在這樣的潮流下，進入另一個階段的製造競爭環境。「智慧工廠」則是未來生產系統的定位，透過設備感測器、物聯網、人工運算與機器學習，利用收集的大量數據加以分析進而做出相關決策，來實現設備高稼動率和製造流程的穩定性。
本研究所收集之數據為A公司所提供的塗佈機台感測器數據，以機器學習（Machine Learning, ML）中的半監督式學習（Semi-supervised learning）來進行塗佈機異常狀況分類，應用異常標籤來確立塗佈機的預測模型，以新增的alarm code數據集來挑選出異常的狀況，選取數據集建立出LSTM模型，透過模型參數設定，將驗證結果繪製出來可以觀察約82.5秒前偵測異常。本研究應用於塗佈機台感測器在時間序列數據中，能夠獲得異常的預測，以供後續研究參考。 | zh_TW |
| dc.description.abstract | Since Germany proposed Industry 4.0, various countries including the United States Advanced Manufacturing Partnership (AMP) reindustrialization cooperation strategy, and emerging countries such as the mainland, South Korea, and Japan are actively developing technologies such as the Internet of Things and smart manufacturing. Taiwan also promoted the productivity 4.0 program in 2015. Under this trend, global countries have entered another stage of manufacturing competition environment. The ＂Smart Factory＂ is the positioning of the future production system. Through device sensors, the Internet of Things, manual calculations and machine learning, a large amount of collected data is used to analyze and make relevant decisions to achieve high equipment utilization and manufacturing processes. stability.
The data collected in this research is the sensor data of the coating machine provided by Company A. The abnormal condition of the coating machine is classified by Semi-supervised learning in Machine Learning (ML). , Use error labels to establish the predictive model of the coating machine, classify the position of the parts of the machine with anomalies for dimensionality reduction, and use the Run Chart to find out the important conditions of the machine′s operation. Cooperate with the method research of failure prediction and health management (Prognostics and Health Management, PHM), use the Long short-term memory network (Long short-term memory, LSTM) model to establish abnormal predictions, provide advance maintenance decisions, and control the stable operation of the machine This reduces the cost of downtime and other abnormal occurrences. | en_US |
| DC.subject | 機器學習 | zh_TW |
| DC.subject | 智慧製造 | zh_TW |
| DC.subject | 工業4.0 | zh_TW |
| DC.subject | 故障預測與健康管理 | zh_TW |
| DC.subject | 塗佈機 | zh_TW |
| DC.subject | Smart Manufacturing | en_US |
| DC.subject | Predictive Maintenance | en_US |
| DC.subject | Prognostic and Health Management | en_US |
| DC.subject | Long short-term memory | en_US |
| DC.subject | Deep Learning | en_US |
| DC.title | 運用異常標籤分類與LSTM方法於停機預測-以A公司塗佈機為例 | zh_TW |
| dc.language.iso | zh-TW | zh-TW |
| DC.title | Using Label Classification And LSTM Approach To Predict – A Case of Coating Machine | en_US |
| DC.type | 博碩士論文 | zh_TW |
| DC.type | thesis | en_US |
| DC.publisher | National Central University | en_US |