透過您的圖書館登入
IP:3.15.6.77
  • 學位論文

應用真圓度量測於精密塑膠球

Applying Roundness Measurement to Precise Plastics Ball

指導教授 : 田方治

摘要


精密塑膠球有著很高的硬度和鋼性,其精密度可達到正負0.2mm~0.001mm之間,具有高度抗變形能力和應力鬆馳能力,它具有良好的綜合性能和耐疲勞性,且耐多次重復衝擊,強度變化很小,加上高技術和高品質,提供給各種高精密的機械零件使用。廣泛的運用在生活上各種零件中,主要用於汽車,傳感器,閥門,泵,衛浴,電子,軸承,電器,等工業。 精密塑膠球最重要的就是其真圓度,真圓度對於精密零件的產品品質之重要性已不在話下,因此在產品製造、生產過程中就不能不考慮其真圓度的量測與管制。但生產現場裡常有些毛邊或異物附著在精密塑膠球表面上,而影響到利用AOI檢驗真圓度時的正確性。本研究比較傅立葉轉換加XY座標法、傅立葉轉換加複數座標法、傅立葉轉換加質心距離法和穩態小波轉換加XY座標法、穩態小波轉換加質心距離法,這五種方法應用在精密塑膠球毛邊濾除的效果,並以PSO結合全域學習之模式搜尋法(HJ-PSO) 計算精密塑膠球之真圓度,實驗結果中以穩態小波轉換加質心距離法的效果最好,能成功的濾除精密塑膠球的毛邊。

並列摘要


Precise plastic ball has very high hardness and rigidity, and it is precision can be achieved between 0.001mm and 0.2mm. It is having capability to resist deformation and stress relaxation, and good fatigue, also it can endure the impact repeatedly, and the change of intensity is very small. Add to high technique and high quality which provide all kinds of high precise machine to use. It applies to all kinds of parts in our life widely. The main purpose of use in the cars, sensor, valve, electron, bearing, electric equipment... etc. The most important thing is that the roundness of precise plastic ball. Therefore, when we manufacture the product must consider about roundness measure and control. But the dust or raw edge on the surface of precise plastic ball may influence the accuracy of roundness detection by Automatic Optical Inspection. In order to filter out the noise in the boundary of precise plastic ball via several shape descriptors, this study compares the Fourier Transform add XY Coordinates, the Fourier Transform add Complex Coordinates, the Fourier Transform add Centroid Distance, the Stationary Wavelet Transform add XY Coordinates and the Stationary Wavelet Transform add Centroid Distance, these five methods used in precise plastic ball filter out the effect of edges noise, and PSO combined with the global search Hooke-Jeeves pattern search method calculate roundness of precise plastic ball. In experimental results show the Stationary Wavelet Transform add Centroid Distance is better than other methods, able to success filter out the edges noise of precise plastic balls.

參考文獻


[30] 滕有揚,數位信號處理晶片應用於聲源訊號雜訊消除之研究,碩士論文,私立中原大學資訊工程學系,桃園,2004。
[32] 蔡宗翰,應用粒子群最佳化演算法於真圓度量測,碩士論文,國立台北科技大學工業工程與管理系,台北,2006。
[26] 陳玟伶,應用機器視覺於鉚釘電氣接點之表面檢測,碩士論文,國立台北科技大學工業工程與管理系,台北,2007。
[25] 陳志龍,智慧型高速真圓度測定儀模擬系統及演算法,碩士論文,國立台灣大學電機工程研究所,台北,2002。
[11] David L. Donoho and Iain M.Johnstone. Ideal spatial adaptation via wavelet shrinkage. Biometrika, 81:425-455, 1994. Also Stanford Statistics Dept.Report TR-400,July 1992.

延伸閱讀