| DC 欄位 |
值 |
語言 |
| DC.contributor | 電機工程學系 | zh_TW |
| DC.creator | 林敬儒 | zh_TW |
| DC.creator | Ching-Ju Lin | en_US |
| dc.date.accessioned | 2018-4-12T07:39:07Z | |
| dc.date.available | 2018-4-12T07:39:07Z | |
| dc.date.issued | 2018 | |
| dc.identifier.uri | http://ir.lib.ncu.edu.tw:88/thesis/view_etd.asp?URN=104521030 | |
| dc.contributor.department | 電機工程學系 | zh_TW |
| DC.description | 國立中央大學 | zh_TW |
| DC.description | National Central University | en_US |
| dc.description.abstract | 本篇論文為以隨機晶圓為基礎使用統計檢定假說，找出不同晶圓尺寸間迴力棒圖的關係，在設計出演算法方便使用者只須要給出尺寸就能快速取得隨機瑕疵所造成的晶圓圖特性分界，以達到快速判讀晶圓圖異常的目地。
首先，我們先選擇出八種尺寸進行較縝密及精確的邊界模擬，對此八種尺寸以四種分佈兩次線性回歸及不同回歸階數做不同的判別率比較，最終取得較方便及優良的方法。經由此實驗結果發現，不同尺寸的迴力棒圖有中心線共線及寬度與尺寸有關係，再以這兩項特點進行模型建造，擴展成由使用者決定晶圓尺寸後快速產生出高判別度的迴力棒邊界。
在本實驗中，兩種方法各有優缺點，第一種模擬方法較為精確但是繁瑣及耗時，大量的時間花費在產生大量隨機些疵造成的晶圓上且尺寸越大所需時間越多，第二種方法雖然準確度較低但也有達到九成五以上的準確性且不論何種尺寸完成速度均極快。 | zh_TW |
| dc.description.abstract | In this paper, we find the relationship between diesize and Boomerang Chart base on wafer map which random distribution of defects. And building model by the relationship that let user just provided diesize to get bound of wafer cause by random defects. To achieve the aim of discrimination abnormal wafers fast.
At first, we choose eight kinds of size of wafers to simulat bound accurately and carefully. We compared 4 kinds of distribution and several kinds of methods and linear regression order. To get the fitter and more convenient method. In Boomerang Chart, we discover that the centerline of all diesizes are almost collinear and the relationship between width and diesize in this experiment. Modeling base on the two factors to let user get bound of Boomerang Chart fast and accurately just provided diesize.
In this experiment, the merits and demerits of two kind of methods are different. The simulation-base is more accurate but cumbersome and time-consuming. We spend lot of time generating random defect wafer and the time increases as the diesize increases. The rule-base is generating fast whatever diesize. Accuracy of bound is up to 95% although it is more imprecise. | en_US |
| DC.subject | 晶圓 | zh_TW |
| DC.subject | 隨機瑕疵 | zh_TW |
| DC.subject | 統計檢定假說 | zh_TW |
| DC.subject | 線性回歸 | zh_TW |
| DC.subject | 標準差 | zh_TW |
| DC.subject | 信賴區間 | zh_TW |
| DC.subject | wafer | en_US |
| DC.subject | random defect | en_US |
| DC.subject | testing statistical hypothesis | en_US |
| DC.subject | linear regression | en_US |
| DC.subject | standard deviation | en_US |
| DC.subject | confidence interval | en_US |
| DC.title | 空間隨機樣態分類器的識別 | zh_TW |
| dc.language.iso | zh-TW | zh-TW |
| DC.title | Identification of the Classifier for the Pattern of Spatial Randomness | en_US |
| DC.type | 博碩士論文 | zh_TW |
| DC.type | thesis | en_US |
| DC.publisher | National Central University | en_US |