以經驗模態分解法與機器學習運用於產線可靠度之研究

Research on the reliability of production lines based on Empirical Mode Decomposition and machine learning

鄭清瑋(CHING-WEI CHENG)

指導教授 ： 楊康宏

中原大學/電機資訊學院/工業與系統工程學系/碩士(2022年)

https://doi.org/10.6840/cycu202201342

摘要

隨著近年來工業的蓬勃發展和企業的高度競爭下，產品生產的工廠扮演了重要的角色，而產能一直是重要的指標之一。產能不足和未達成客戶的要求會使企業承受大量的風險成本，而造成產能不足的部份原因是因為生產線上機台發生故障或是例行保養造成的停機損失。如果我們能知道停機時間和平均無故障時間(Mean Time Between Failure, MTBF)以及平均修復時間(Mean Time To Repair, MTTR)之間的關係，就可提供生產線上人員多一種評估的參考。
而本研究會依據已經建立的情境並利用Flexsim進行生產線上的生產狀況模擬，再把模擬出來的機台數據以經驗模態分解法進行分解。分解出來的分量訊號再以機器學習中的卷積神經網路來做分析，以用來探討不同MTBF/ MTTR運行時間之長短關係，讓我們可以評估生產線上的故障情況。

關鍵字

平均無故障時間 ； 平均修復時間 ； 系統模擬 ； 卷積神經網路 ； 經驗模態分解法

並列摘要

With the vigorous development of industry and the high competition of enterprises in recent years, factories that produce products have played an important role, and production capacity has always been one of the important indicators. Insufficient capacity and failure to meet customer requirements expose businesses to substantial risk costs. Part of the reason for the lack of production capacity is due to machine failures on the production line or downtime losses caused by routine maintenance. If we can know the relationship between MTBF and MTTR, it can provide an additional reference for people on the production line to evaluate.
This research will use Flexsim to simulate the production status of the production line based on the established situation, and then decompose the simulated machine data by the empirical modal decomposition method. The decomposed component signals are then analyzed by the convolutional neural network in machine learning to explore the relationship between the running time of different MTBF/MTTR, so that we can evaluate the failure situation on the production line.

並列關鍵字

Mean time between failure ； Mean time to repair ； System simulation ； Convolution neural network ； Empirical modal decomposition

參考文獻

Borith, T., Bakhit, S., Nasridinov, A., & Yoo, K. H. (2020). Prediction of Machine Inactivation Status Using Statistical Feature Extraction and Machine Learning. Applied Sciences-Basel, 10(21). https://doi.org/ARTN 7413

10.3390/app10217413

Dumoulin, V., & Visin, F. (2016). A guide to convolution arithmetic for deep learning. arXiv:1603.07285. Retrieved March 01, 2016, from https://ui.adsabs.harvard.edu/abs/2016arXiv160307285D

Fawcett, T. (2006). An introduction to ROC analysis. Pattern Recognition Letters, 27(8), 861-874. https://doi.org/https://doi.org/10.1016/j.patrec.2005.10.010

Gong, W., Chen, H., Zhang, Z., Zhang, M., Wang, R., Guan, C., & Wang, Q. (2019). A Novel Deep Learning Method for Intelligent Fault Diagnosis of Rotating Machinery Based on Improved CNN-SVM and Multichannel Data Fusion. Sensors, 19(7), 1693. https://www.mdpi.com/1424-8220/19/7/1693