透過您的圖書館登入
IP:3.94.102.228
  • 學位論文

柔性解算應用於品質機能展開之研究

The Application of Soft Computing in the Areas of Quality Function Deployment

指導教授 : 鄭春生
若您是本文的作者,可授權文章由華藝線上圖書館中協助推廣。

摘要


在一個製造系統的品質機能展開中, 由企劃、設計、以至於製造各階段的顧客需求以及技術需求, 均可以用品質屋中之關連矩陣來表示, 並逐層展開分析以得到最後製程規格及品質標準。 而關連矩陣中之關連性指標或符號, 必須用一客觀評估的指標值, 以利系統展開後之正確性及適用性。 傳統上,關連矩陣中指標值之決定, 均以品質機能展開小組人員評估決定, 再根據經驗法則,主觀給予一對應關係指標值, 這樣的方法會有一些缺點, 例如設計的人不懂製造以致指標及指標值偏差; 關連程度指標為階段之整數值,不夠精細準確等, 而導致品質機能展開至最後階段結果無法代表顧客之心聲。 本研究是結合類神經網路、 模糊理論應用於品質機能展開品質展開階段之顧客需求對應於品質特性之關係矩陣。 同時整合設計與製造各階段專業人員於品質展開階段之關係矩陣指標值之意見, 輔以類神經網路找出較客觀看法,並以語言變數表示關連性, 跳出以往有限階級式的整數指標值。 此時之關係矩陣是集合各階段專家之意見, 並已將少數異常意見剔除,故其較傳統關係矩陣更具客觀性與代表性, 使顧客需求能正確,並且客觀的導入品質機能展開中。

並列摘要


Within a quality function deployment system of manufacturing industries,the relationship matrix of quality house can not only be used to represent customer and technology requirements of every manufacturing phase,but also can be used to obtain the final manufacturing process specifications and quality standards. To keep the correctness and adaptability,obtaining an objective pointer of relationship matrix is necessary. The choosing of pointer of relationship matrix,traditionally,is evaluated and chosen by members of quality function develop staff. In this situation,all the selection is made by personal subjective experiment,so that the disadvantage of variance or mistakes induced by personal misunderstanding may occur. Besides this,the inaccuracy induced by personal misunderstanding will also cause the results of quality function deployment phases do not even match the customers' requirements. In this research, the application of neural net and fuzzy logic to the quality function deployment to achieve the customers' requirements is presented. At the mean time, the integration of different opinions of professional people in both design and manufacturing phases with the help of neural net to get more objective result is considered. After all, the final relationship matrix in this research is obtained by gathering and analyzing the opinions of professionals and then filtering out the extremes. From the comparison between our method and traditional relationship matrix, it can be proven that the relationship matrix presented from this research is more objective and more accurate to reflect customers' requirements.

參考文獻


15. Kaufmann, A. and M. M. Gupta, Introduction to Fuzzy Arithmatic , New York: Van Nostrand Reinhold, 1985.
1. Bossert, J. L., Quality Function Deployment: A Practitioner's Approach, Mil Waukee : Asqc Quality Press, 1990.
2. Carpenter, G. A. and S. Grossberg, "A massively parallel architecture for a self-organizing neural pattern recognition machine," Computer Vision, Graphics and Image Processing, 37, 54-115 , 1987.
3. Carpenter, G. A. and S. Grossberg, "ART2: self-organization of stable category recognition codes for analog input patterns," Applied Optics, 26, 4919-4930, 1987.
4. Carpenter, G. A. and S. Grossberg, "ART3: Hierarchical search using chemical transmitters in self-organizing pattern recognition architectures," Neural Networks, 3, 129-152, 1990.

被引用紀錄


林佳樺(2003)。退貨授權(RMA)資訊在品質保證系統上之探討─以某通信公司為例〔碩士論文,元智大學〕。華藝線上圖書館。https://www.airitilibrary.com/Article/Detail?DocID=U0009-0112200611314399

延伸閱讀