利用類神經網路與田口方法參數優化與模型預測—以CNC銑切製程為例

Integration of Neural Network and Taguchi Method for Parameter Optimization and Model Prediction : Case Study in Cutting Process of CNC Machine

吳秋霖(Ciou-Lin Wu)

指導教授 ： 江瑞清

中原大學/電機資訊學院/工業與系統工程研究所/碩士(2019年)

https://doi.org/10.6840/cycu201900253

摘要

目前在工業界上的人工智慧應用發展得越來越好，數據量也變得越來越多，需要提高生產效率及降低成本才能帶來競爭性，因此，本研究透過CNC機台的數據為背景，使用三種不同的方式，分別為神經網路、田口方法以及田口結合神經網路，來探討預測目標值的準確度，透過實際切削出的值做為神經網路訓練的數據，再使用田口方法進行參數優化，透過優化後的參數組合所測量的值當作實際值，比較三種不同方式何種準確度高，最後實驗結果為田口與類神經網路結合所預測出的值最為精準，誤差值為0.044，小於5%，屬於一般業界可接受之範圍。

關鍵字

類神經網路 ； 田口方法 ； 量測系統分析 ； 切削

並列摘要

At present, the application of artificial intelligence in the industry is getting better and better, the amount of data is becoming more and more, and it is necessary to improve production efficiency and reduce costs to bring competitiveness. Therefore, this study uses the data of the CNC machine as the background, using three different methods, namely neural network, Taguchi method and Taguchi combined neural network, to explore the accuracy of the predicted target value. The actual cut value is used as the data of the neural network training, and then the parameters are used in the Taguchi method. The Taguchi method is used for parameter optimization, and the measured value is used as the actual value through the optimized parameter combination, and the accuracy is compared in three different ways. The final experimental result is the most accurate value predicted by the combination of Taguchi and the neural network. A value of 0.044, less than 5%, which is acceptable to the general industry.

並列關鍵字

Neural Network ； Taguchi Method ； Measurement System Analysis ； Cutting

參考文獻

【英文文獻】

Chern, G. L., & Liang, J. M., (2007). “Study on boring and drilling with vibration cutting”, International Journal of Machine Tools & Manufacture, 47, 133–140.

Ekanayake, R.A., & Mathew, P., (2007). “An Experimental Investigation of High Speed End Milling” 5th Australasian Congress on Applied Mechanics, ACAM 2007.

Frank, R., (1958). “The perceptron: a probabilistic model for information storage and organization in the brain”, Psychological review, 65.6, 386.

Gaitonde, V. N., Karnik, S. R., Achyutha, B.T., & Siddeswarappa, B., (2005). “GA applications to RSM based models for burr size reduction in drilling”, Journal of scientific and industrial research, 64(5), 347-353.