TW201033933A - Scheduling method - Google Patents

Scheduling method Download PDF

Info

Publication number
TW201033933A
TW201033933A TW98107291A TW98107291A TW201033933A TW 201033933 A TW201033933 A TW 201033933A TW 98107291 A TW98107291 A TW 98107291A TW 98107291 A TW98107291 A TW 98107291A TW 201033933 A TW201033933 A TW 201033933A
Authority
TW
Taiwan
Prior art keywords
rule
factors
production
scheduling method
scheduling
Prior art date
Application number
TW98107291A
Other languages
Chinese (zh)
Inventor
Chung-Pei Chao
Chun-Chi Chen
Yun-Zong Tian
Yi-Yuan Cheng
Tusng-Wei Lin
Chih-Cheng Liao
Yi-Feng Lee
Shin-Chang Kao
Original Assignee
Inotera Memories Inc
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Inotera Memories Inc filed Critical Inotera Memories Inc
Priority to TW98107291A priority Critical patent/TW201033933A/en
Publication of TW201033933A publication Critical patent/TW201033933A/en

Links

Landscapes

  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)
  • General Factory Administration (AREA)

Abstract

A scheduling method for manufacturing semiconductor is shown below. First step is choosing a plurality of rules. Next step is classifying the rules into some rule classes and defining a weight for a single rule. Next step is scheduling the manufacturing processes depending on the weights of the rules. At last, the rules can be updated and the weights can be efficiently adjusted. Accordingly, the experience rules and the rules of tool and processes are considered in order to scheduling the steps smartly. Furthermore, the rules are easily updated and adjusted in this dynamic rule selection method.

Description

201033933 六、發明說明: 【發明所屬之技術領域】 方法本於—種應用於半導體產線的生產排程 ^法《日-種可將主客觀條件均納人考量之生產排程方 【先前技術】 •的功=2製程技術能力不斷向上提升,半導體晶片 曰片的針射κ M致半導體晶片的製程步料漸增加, =i:=增加。當半導體的製造產業進入自動 ,5、生產系統均依照著不同的㈣、製程將待 的原料或半成品送入相對應的機台或工具中進行f =隨著生產量的增加,不同的產品線可能會遇到:i = :Γί的狀況’或是必須針對同-生產線的的流 程進盯規劃,才能達成最佳生產效率的製造模式。 參般來說’製造資源是生產作業的要素之一,如生產 Ϊ是人員的管控’它可以靈活的移轉配置在某 ==產線上執行作業,但在多條產線的生產過程遇 =衝犬的情況,便會使得雜無法順利· =使用w取得錯的核,就需懸過合理的ί • ^排,才不會成為生產管理中追求高效率的綷聊石。簡 ' 财在生產過財可能使㈣人力 (作業員/技術員)或設備(儀器/模具/治具);它是生產排 ㈣要素之-’必須使用有限的生產資源進行最有效率且 3 201033933 不會產生衝突的排程計畫。 生產排程難以程式化的原因是因為在現 滿了各種可能性,預先設定的排程條件卻常=充 的人員或設備的情況必須隨時調整因為產線上 解決方案,必須要能符合實務應變的需求。王系統提出的 ❹ 的主依賴管理者的經驗,通常資深 丄二積的排程經驗進行生產線上的 盤考,::t 在排程因子的考慮上無法做-全 效:某一個重要因子,其排程效率就無法有 =升,另-方面,不同的管理者也會有不同的排程規 1不同的排程也容易造成人員在溝通協調上的問題。 緣是,本發日狀錢上述缺失之可改善, 计合理且有效改善上述缺失之本發明。 禋又 【發明内容】 本制之主要目的,在於提供—種半導體製程甲的生 /排私方法,該方法可提供較佳的工作排程規劃,該方法 =將線上主管或資深人員的經驗與其他製程條件因子整 口:並共同加以考量,故該方法可考慮較全面性的因素, 使最後的排程計畫具有相當優良的生產效率。、 、為了達成上述之目的,本發明係提供一種生產排程方 法.,應用於半導體生產的排程管理,該方法包括以下步 驟·決定複數個規則因子,·分類該等規則因子於複數個規 則類別(rule Class),且進行該等規則因子的權重分配; 利用分類该等規則因子步驟_之權重分配進行模擬及線 4 201033933 上測試;以及確認效能及微調之步驟。 本發明具有以下有益的效果:本發明提出之排程規劃 方法’可將線上資深主管的排程經驗納入排程系統的考慮 之中’使本排程方法所規劃之生產流程能提供半導體產線 更佳之生產效率。 為使此更進一步瞭解本發明之特徵及技術内容,請參 閱以下有關本發明之詳細說明與附圖、然而所附圖式僅提 供參考與說_,並_來對本發明加以限制者。 【實施方式】 5月參閱第一圖 ’本發明係提供一種有關半導體產線的 ^產排程方法,雜財法可以有效的考慮多個生產的條 1 ’其中可能整合有主管的排程經驗或是生產機台的 oading考篁等條件’其排程方法包括以下步驟: A、f驟(3)决疋規則因子(rule):在此一步驟中, 早則因子,,選。由於生產排程的概念中’規則因 的考置早疋’故本發明係可就每一生產階段可 二ίΠ則因子加以選取’以利用該些規則因子的關 排序作業。在本發明中,生產線上的資 量掏“上生的考量’故在此步驟中會大 的排序條件納人本發_方=:考量二= 本Τ明也可以針對每—生產道次的機丄:面’ 程條件等各方㈣考量加以評估,目的就在 201033933 i:::=考慮各種條件因子,並分析該些因子以達到 由操作者s排序流程。總的來說’上述的規則因子係 .規則)及系絲規則(即有經驗的主管所提出的排序 •、系、,先内建的規則(即機台條件與出貨量停件等等) 所操取出的影響工作排_序財量因素條件等等) 步驟(Μ分類該等規則因子。在步 選出的規則因子在此牛驟士 a c )中所卒 子在此步驟中會被分類為多個規則類別 (ule ciass),故使本排程方法 一 ❿統進行排程作業。另一方而… 禋物件導向的糸 賦予m、. h‘被分義規_子會被 子的優弁趨即可利用該權重定義每-規則因 子〇優先權,而本發明提出的排程方 :子的優先權進行工作的排程順序。且,在此 括一確認(verify)該等規則因子的步驟,以再一次確切 ^取的每—㈣子都是影響排㈣考量时;而在定義^ =規顧子的權重之後,更包括—排除該等規則因 ⑩抵觸的步驟’以將可能造成規則因子的 排序優先順序抵觸的情況加以排除。 步驟(c)提供最佳化方法之步驟。在本發明中,此 一步驟係為—種選擇性—υ的步驟,本步驟的主 要目的係選擇各種適合_計方法確認上述的 -的權重分配以達到最佳化的效果,亦即可利用合適的最佳 化方法進一步提高該排程方法所能達 來說,本步驟中可以使用fuzzy(^= 舉例 葆蝴理論)等等的最伟 化方法以達更佳的生產效能。 ^1 6 201033933 步驟(d )確認效能及模型微調。最後的步驟在於將 上述的排序結果導入實際的生產線上,以確認該排程方法 所排出的規則因子的權重可以達成的實際效能;然而此步 ' 驟中更允許進行該排程模型的微調作業,例如更新規則因 子的項目或是進行規則因子的權重(優先權)的微調等 等,以使該排程作業在一個動態規則的環境下運作,使用 者(如主管)可以進行最佳的排程規劃。 故在上述步驟流程之後,即可完成一生產排程的模 ® 型,藉由該生產排程的規劃,半導體生產線上的工作就可 以被妥善地排程,更可以減少生產機台的閒置時間,以提 高生產半導體元件的產能。 以下本發明將針對半導體生產線上的ARF Tool做一 排程的規劃,以解決傳統上ARF Tool所遭遇到的排程問 題’例如層間轉換的頻率(level change frequency)比 較差,造成生產週期無法縮短的問題。 A 請參考第二圖,首先,先決定規則因子,例如在一般 半導體製程中,為因應產品的出貨,可能會有Month line (月線)、Layer priority(高優先權的Layer優先,或是 Lot priority (如 super hot lot、hot lot 等優先權子項 目)等資深主管會優先考慮的規則因子,或是一些必須轉 換工具、製程的因子,如Speed Loss (製程轉換所需耗費 的時間);亦即在本步驟中,會針對全面性的條件因子做 一考慮及選取。 接下來,再將上述的規則因子加以分類,在本具體實 7 201033933 施例中,可大致將規則因子分為以下數類,並加以排序形 成生產的優先次序: 1. Tool constrained layer WIP Assignment。針對機台限 制的待生產批次(Lot)會優先指派例如考慮目前製程的 限制,黃光區中主動區(active area,AA )、閘極導電層 (gate conductor,GC Layer)的批次必須回流至深溝層 (deep trench’ DT Layer)生產的機台。201033933 VI. Description of the invention: [Technical field to which the invention belongs] The method is applied to the production scheduling of the semiconductor production line, and the production schedule of the subjective and objective conditions can be considered. 】 • Work = 2 process technology capabilities continue to rise, the semiconductor wafer 的 针 κ 致 致 致 致 致 致 致 致 致 致 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体When the semiconductor manufacturing industry enters the automatic, 5, the production system according to different (four), the process of the raw materials or semi-finished products to be sent to the corresponding machine or tool for f = as the production volume increases, different product lines You may encounter: i = : Γί's condition' or you must plan for the same-line process to achieve the best production efficiency. In general, 'manufacturing resources are one of the elements of production operations, such as production control is the control of personnel'. It can be flexibly transferred to perform operations on a certain == production line, but in the production process of multiple production lines = In the case of rushing dogs, it will make the miscellaneous things unsatisfactory. If you use w to get the wrong nucleus, you need to hang a reasonable ί ^ row, so that you will not become a high-efficiency slogan in production management. Jane's production in the past may result in (iv) manpower (workers/technicians) or equipment (instruments/tools/fixtures); it is the element of production line (four) - 'must use limited production resources for the most efficient and 3 201033933 There are no conflicting schedules. The reason why the production schedule is difficult to stylize is because the various schedules are full, and the preset scheduling conditions are often the same. The situation of the personnel or equipment must be adjusted at any time. Because the solution on the production line must be in line with the practical constraints. demand. The system of the 依赖 提出 依赖 王 王 , , , , , 王 王 王 王 王 王 王 王 王 王 王 王 王 王 王 王 王 王 王 王 王 王 王 王 王 王 王 王 王 王 王 王 王 王 王 王 王The scheduling efficiency can't be increased. On the other hand, different managers will have different schedules. 1 Different schedules are also likely to cause problems in communication and coordination. The reason is that the above-mentioned deficiency of the present day money can be improved, and the present invention which is plausible and effective is improved.禋 【 【Abstract】 The main purpose of this system is to provide a semiconductor process A production / private method, this method can provide better work scheduling, this method = the experience of online supervisors or senior staff Other process conditions factors are: and considered together, so the method can consider more comprehensive factors, so that the final scheduling project has quite good production efficiency. In order to achieve the above object, the present invention provides a production scheduling method for scheduling management of semiconductor production, the method comprising the following steps: determining a plurality of rule factors, classifying the rule factors into a plurality of rules Class (rule class), and perform weighting of the rule factors; use the weighting of the rule factors to perform simulation and line 4 test on 201033933; and confirm the steps of performance and fine-tuning. The invention has the following beneficial effects: the scheduling method proposed by the present invention can incorporate the scheduling experience of the online senior supervisor into the scheduling system considerations, so that the production process planned by the scheduling method can provide the semiconductor production line. Better production efficiency. For a better understanding of the features and technical aspects of the present invention, reference should be made to the accompanying drawings and the accompanying drawings. [Embodiment] Referring to the first figure in May, the present invention provides a method for scheduling production of a semiconductor production line, and the miscellaneous wealth method can effectively consider a plurality of production strips 1 'which may integrate supervisor's scheduling experience Or the conditions of the oading test of the production machine's scheduling method includes the following steps: A, f (3) rule rule factor (rule): In this step, the factor is early, select. Due to the fact that the 'rule of the rule is set early' in the concept of production scheduling, the present invention can select a factor for each production stage to take advantage of the sorting operations of the rule factors. In the present invention, the amount of resources on the production line is "the consideration of the upper life", so in this step, there will be a large sorting condition of the person's own hair _ party =: consideration two = Ben Ming also can be for each production pass The machine: the surface conditions and other parties (four) considerations are evaluated, the purpose is to consider various condition factors in 201033933 i:::=, and analyze the factors to achieve the sorting process by the operator s. In general, the above Rule factor system. Rules and rules (ie, the ordering, system, and pre-built rules proposed by experienced supervisors (ie, machine conditions and shipments, etc.) The order of the _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ Ciass), so this scheduling method is used for scheduling work. The other side... 禋 object-oriented 糸 gives m, .h' is divided into stipulations _ sub-segment quilt can use this weight to define each - rule factor 〇 priority, and the scheduler proposed by the present invention: child The order of priority for the work to be performed. Also, the step of verifying the rule factors is included here, so that each (four) child that is exactly taken again is affected by the row (four) consideration; = After the weights of the rules, the steps of excluding the rules from the 10 conflicts are excluded from the situation in which the ordering priorities of the rule factors may be violated. Step (c) provides the steps of the optimization method. In the present invention, this step is a step of selective-υ, and the main purpose of this step is to select various suitable methods to confirm the above-mentioned weight distribution to achieve an optimized effect, that is, to use appropriate The optimization method can further improve the scheduling method. In this step, the most advanced method such as fuzzy (^= example butterfly theory) can be used to achieve better production efficiency. ^1 6 201033933 Step (d) confirms the performance and model fine-tuning. The final step is to import the above sorting result into the actual production line to confirm the actual weight of the rule factor discharged by the scheduling method can be achieved. However, in this step, the fine-tuning of the scheduling model is allowed, such as updating the rule factor item or fine-tuning the weight of the rule factor (priority), etc., so that the scheduling job is in a dynamic Operating in a regular environment, users (such as supervisors) can perform optimal scheduling planning. Therefore, after the above steps, a production scheduling model can be completed, and the planning of the production scheduling, semiconductor The work on the production line can be properly scheduled, and the idle time of the production machine can be reduced to increase the production capacity of the semiconductor components. The following invention will be planned for the ARF Tool on the semiconductor production line to solve the problem. Traditionally, the scheduling problems encountered by the ARF Tool, such as the level change frequency, are relatively poor, resulting in a problem that the production cycle cannot be shortened. A Please refer to the second figure. First, determine the rule factor. For example, in the general semiconductor process, in order to respond to the shipment of the product, there may be a Month line, a Layer priority, or a high priority. Lot priority (such as super hot lot, hot lot and other priority sub-projects) and other senior managers will give priority to the rule factors, or some factors that must be converted tools, processes, such as Speed Loss (time required for process conversion); That is to say, in this step, a comprehensive condition factor is considered and selected. Next, the above rule factors are classified, and in the embodiment of this embodiment 7 201033933, the rule factors can be roughly divided into the following The classes are sorted to prioritize production: 1. Tool constrained layer WIP Assignment. The batch to be produced (Lot) for the machine limit will be assigned preferentially, for example, considering the current process limit, active zone in the yellow zone (active) Area, AA), gate conductor (GC Layer) batch must be reflowed to the deep trench ' DT Layer production machine

2· New Product (新產品優先):針對新的產品給予較高的 生產優先權。 ° ' 3. Month line (月線):為確保每月的總產能,生產線會 設計一動態的月線分布,當某一產品月線到達特定製程 時,負責該製程的機台則必須優先處理該產品、或 程的生產批次。 4. Layer priority:高優先權的Uyer優先,例如生產 制部門為降低生產週期(Cycle Time),故會依據線上= 品的分布較不同產品、不同Layer-個權重值,權重ί 則具有較高優先權。 μ 5· Lens heating (透鏡熱效應):由於部分[邱打之 線寬有不同限制與要求,當機台轉換㈣⑶時能 ^ 異’會導致生產品質的差異,例如 DT Layer ’必須先經過其他 行緩衝 造j 因子的分類。 此―分類即為製程工具上的優先考量 〜千丨常仵的限制) 8 201033933 以生產,另外考慮到生產品質,部份機台在生產時有搭配 特定光罩的限制。 7. Single score value :除了 Layer priority 外,生產單 位會依據生產批次的等待時間、權重、產品等級…等因 素,額外計算一個權重指標Single score value作為參 考,若Single score value越小,則越優先處理。 8. Train size :考慮到轉換Layer時需要更換光罩,甚至 面臨Lens heat i ng的風險,所以應儘可能讓同一個產品 參 且同一個層(Layer)的生產作業一起集中處理,故有相 同產品時,其執行的權重就會較高。 經過上述的分類與權重的定義後,並確認其中的排程 規劃不會出現互相衝突的因子權重,且利用fuzzy等的統 計方法進行最佳化的分析,確認因子的權重分配均已達到 最適化的要求,即可將該些權重導入實際的生產端進行生 產的排程計晝,以使其可以達到最佳的生產效能。請參考 表1,其為利用本發明與傳統的排程方法所得到的層間轉 ® 換的頻率(level change frequency)的比較表: 表1 生產工具 層間轉換次數 (layer change) 批次數量 (lot count) 頻率 tool 1 傳統方法 4 77 19. 25 本發明 6 75 12. 50 tool 2 傳統方法 20 76 3.80 本發明 13 79 6. 08 9 201033933 tool 3 傳統方法 10 66 6. 60 本發明 6 57 9. 50 tool 4 傳統方法 4 33 8.25 本發明 7 68 9. 71 tool 5 傳統方法 14 68 4. 86 本發明 9 77 8. 56 tool 6 傳統方法 24 87 3. 63 本發明 9 82 9. 11 tool 7 傳統方法 10 38 3. 80 本發明 13 80 6. 15 tool 8 傳統方法 19 83 4.37 本發明 16 86 5.38 tool 9 傳統方法 24 86 3. 58 本發明 8 64 8. 00 上表中,該頻率係為批次數量與層間轉換次數的比 值,該比值越高則代表系統進行層間更換的次數越低,生 產端即可以進行更有效率的作業。故除了第一個工具 (tool 1)之外,其他的工具則具有較傳統排程方法較低 的轉換次數,亦即本排程方法可以有效提升ARF Tool的 生產效能。 表2則顯示每一工具應用傳統排程方法與本發明之 排程方法在每一製程,如DT( deep trench)段或是GC( gate contact)段的生產數量。表3則是根據表2做一整合計 201033933 算,以顯示出每一製程所產出的產品總量: 表2 DT AA GC CS M0 Total tool 1 傳統 0 0 0 18 59 77 本發明 0 0 0 24 51 75 tool 2 傳統 18 5 28 6 19 76 本發明 16 17 30 8 8 79 tool 3 傳統 0 13 4 7 42 66 本發明 0 11 4 18 24 57 tool 4 傳統 4 29 0 0 0 33 本發明 0 36 0 8 24 68 tool 5 傳統 17 16 21 14 0 68 本發明 16 15 22 8 16 77 tool 6 傳統 11 36 28 12 0 87 本發明 16 40 26 0 0 82 tool 7 傳統 0 27 2 9 0 38 本發明 16 28 4 8 24 80 tool 8 傳統 6 23 25 13 16 83 本發明 16 22 32 8 8 86 tool 9 傳統 27 10 23 3 23 86 本發明 24 10 22 0 8 64 表 3___ 傳統方法 本發明 11 201033933 DT製程數量 83 104 AA製程數量 159 179 GC製程數量 131 140 CS製程數量 82 82 M0製程數量 19 163 總量 614 668 由表3中可知,利用本發明的排程方法可以提高製程 ® 的產出量,從表3的分析來看,除了 CS製程以外的其他 製程,應用本發明的權重排序方法均可以較傳統的排序方 法達成更佳的產出效率(即產出的總量較大)。 綜上所述,本發明具有下列諸項優點: 1、 提供較佳的工作排程,由於本發明可以將線上主管或 資深人員的經驗與其他工作因子結合並共同加以考 量,進而可考慮較全面性的因素,使最後的排程計畫 A 具有相當優良的生產效率。 2、 另一方面,本發明屬於動態的排程管理,故本發明可 針對新產品或是新的規則因子列入評估的範圍,且本 排程方法更可隨時調整規則因子的權重,以符合生產 線上的情況,使排程計晝具有相當高的適用性。 惟以上所述僅為本發明之較佳實施例,非意欲侷限本 發明之專利保護範圍,故舉凡運用本發明說明書及圖式内 容所為之等效變化,均同理皆包含於本發明之權利保護範 圍内,合予陳明。 12 201033933 【圖式簡單說明】 第一圖係本發明之生產排程方法之流程圖。 第二圖係本發明之生產排程方法應用於ARF Tool之流程 圖。 【主要元件符號說明】 無2· New Product: Gives higher production priority to new products. ° ' 3. Month line: To ensure the monthly total capacity, the production line will design a dynamic monthly distribution. When a product's monthly line reaches a specific process, the machine responsible for the process must be prioritized. The production batch of the product, or process. 4. Layer priority: High priority Uyer priority. For example, the production system reduces the production cycle (Cyce Time). Therefore, the distribution of the online product is different according to different products and different Layer-weight values. The weight ί is higher. priority. μ 5· Lens heating: Due to the different restrictions and requirements of the section [Qiu Dao's line width, when the machine converts (4) (3) can be different, it will lead to differences in production quality, such as DT Layer 'must pass other lines first The classification of the buffering factor. This classification is a priority consideration on the process tool. ~ 1000丨常仵) 8 201033933 In order to produce, and in addition to the production quality, some machines are produced with a specific mask limitation. 7. Single score value: In addition to the Layer priority, the production unit will calculate a single score value as a reference based on the waiting time, weight, product grade, etc. of the production batch. If the single score value is smaller, the more Priority processing. 8. Train size: Considering the need to change the mask when switching the Layer, even the risk of Lens heat i ng, so the same product should be processed together with the same layer (Layer) production work, so the same When the product is used, its execution weight will be higher. After the above classification and weight definition, and confirm that the schedule planning does not have conflicting factor weights, and use the statistical methods such as fuzzy to optimize the analysis, confirm that the weight distribution of the factors has been optimized. The requirements can be used to import the weights into the actual production side for the production schedule, so that it can achieve the best production performance. Please refer to Table 1, which is a comparison table of the level change frequency obtained by the present invention and the conventional scheduling method: Table 1 The number of batch changes of the production tool (lot change) Count) frequency tool 1 conventional method 4 77 19. 25 invention 6 75 12. 50 tool 2 conventional method 20 76 3.80 invention 13 79 6. 08 9 201033933 tool 3 conventional method 10 66 6. 60 invention 6 57 9. 50 tool 4 Conventional method 4 33 8.25 The present invention 7 68 9. 71 tool 5 Conventional method 14 68 4. 86 The present invention 9 77 8. 56 tool 6 Conventional method 24 87 3. 63 The present invention 9 82 9. 11 tool 7 Tradition Method 10 38 3. 80 The invention 13 80 6. 15 tool 8 Conventional method 19 83 4.37 The invention 16 86 5.38 tool 9 Conventional method 24 86 3. 58 The invention 8 64 8. 00 In the above table, the frequency is a batch The ratio of the number of times to the number of times of interlayer conversion. The higher the ratio, the lower the number of times the system performs interlayer replacement, and the production end can perform more efficient operations. Therefore, in addition to the first tool (tool 1), other tools have a lower number of conversions than the traditional scheduling method, that is, the scheduling method can effectively improve the production performance of the ARF Tool. Table 2 shows the number of productions for each tool using the conventional scheduling method and the scheduling method of the present invention in each process, such as a DT (deep trench) segment or a GC (gate contact) segment. Table 3 is an integration calculation according to Table 2, 201033933, to show the total amount of products produced by each process: Table 2 DT AA GC CS M0 Total tool 1 Traditional 0 0 0 18 59 77 The present invention 0 0 0 24 51 75 tool 2 Conventional 18 5 28 6 19 76 Invention 16 17 30 8 8 79 tool 3 Conventional 0 13 4 7 42 66 Invention 0 11 4 18 24 57 tool 4 Tradition 4 29 0 0 0 33 0 36 0 8 24 68 tool 5 Tradition 17 16 21 14 0 68 Invention 16 15 22 8 16 77 tool 6 Tradition 11 36 28 12 0 87 Invention 16 40 26 0 0 82 tool 7 Tradition 0 27 2 9 0 38 Inventive 16 28 4 8 24 80 tool 8 Conventional 6 23 25 13 16 83 The invention 16 22 32 8 8 86 tool 9 Tradition 27 10 23 3 23 86 The invention 24 10 22 0 8 64 Table 3___ Conventional method The present invention 11 201033933 DT Process number 83 104 AA Process Quantity 159 179 GC Process Number 131 140 CS Process Number 82 82 M0 Process Number 19 163 Total 614 668 As can be seen from Table 3, the throughput of Process ® can be increased by using the scheduling method of the present invention. According to the analysis of 3, in addition to the CS process, the weight of the invention is applied. The sorting method can achieve better output efficiency (that is, a larger total output) than the traditional sorting method. In summary, the present invention has the following advantages: 1. Providing a better work schedule, since the present invention can combine the experience of online supervisors or senior personnel with other work factors and consider them together, and thus can be considered comprehensively. The sex factor makes the final scheduling plan A very productive. 2. On the other hand, the present invention belongs to dynamic scheduling management, so the present invention can be included in the evaluation scope for new products or new rule factors, and the scheduling method can adjust the weight of the rule factors at any time to meet the requirements. The situation on the production line makes the schedule meter have a very high applicability. The above is only the preferred embodiment of the present invention, and is not intended to limit the scope of the present invention. Therefore, the equivalents of the present invention and the equivalents of the drawings are all included in the present invention. Within the scope of protection, it is given to Chen Ming. 12 201033933 [Simple description of the drawings] The first figure is a flow chart of the production scheduling method of the present invention. The second figure is a flow chart of the production scheduling method of the present invention applied to the ARF Tool. [Main component symbol description] None

1313

Claims (1)

201033933 七、申請專利範圍: 1、一種生產排程方法,其應用於半導體生產的排程管 理,該生產排程方法包括以下步驟: 決定複數個規則因子(rule); 分類該等規則因子於複數個規則類別(ruleclass), 且進行该等規則因子的權重分配; 矛J用为類該等規則因子步驟中之權重分配進行工作 排程、模擬及線上測試;以及 ί 確認效能及微調權重之步驟。 2 '如^專利範圍第!項所述之生產排程方法,其中在 決疋複數個規則因子之步驟中,該等規則因子係由系 統内建的規則及操作者定義的排序規則所擷取出。 3、如申請專利範圍第2項所述之生產排程方法,其中在 分類該等_因子之步财更包括—確認該等 因子的步驟。 • 4、如中請專圍第3項所述之生產排程方法,其中在 分類該等關时之步射更包括_排除該等規則 因子的權重分配造成互相抵觸的步驟。 5、如中請專利範圍第4項所述之生產排程方法,其 分類該等規則因子之步驟之後更包括一提供最佳 方法計算該等規則因子的步驟。 6如申請專利範圍第5項所述之生產排程方法,其中該 最佳化方法係用以修正該分類該等規則因子二 中之該等規則因子的權重分配。 乂 14 201033933 7、如申請專利範圍第6項所述之生產排程方法,其中在 確認效能及微調權重之步驟中更包括一更新規則因 子及調整該等規則因子的權重分配之步驟。201033933 VII. Patent application scope: 1. A production scheduling method, which is applied to scheduling management of semiconductor production. The production scheduling method comprises the following steps: determining a plurality of rule factors (rule); classifying the rule factors in plural Rule classes, and assign weights to the rule factors; Spear J uses the weight assignments in the rule factor steps for work scheduling, simulation, and online testing; and ί steps to confirm performance and fine-tune weights . 2 'such as ^ patent range! The production scheduling method described in the item, wherein in the step of determining a plurality of rule factors, the rule factors are extracted by a system built-in rule and an operator-defined ordering rule. 3. The production scheduling method of claim 2, wherein the step of classifying the _ factors further comprises the step of confirming the factors. • 4. In the case of the production scheduling method described in item 3, the step of categorizing the levels includes the step of excluding the weighting of the rule factors from causing conflicts. 5. The production scheduling method of claim 4, wherein the step of classifying the rule factors further comprises the step of providing an optimal method for calculating the rule factors. 6. The production scheduling method of claim 5, wherein the optimization method is for modifying a weighting of the rule factors of the classification factor 2 of the classification.乂 14 201033933 7. The production scheduling method according to claim 6, wherein the step of confirming the performance and fine-tuning the weight further comprises the step of updating the rule factor and adjusting the weight distribution of the rule factors. 1515
TW98107291A 2009-03-06 2009-03-06 Scheduling method TW201033933A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
TW98107291A TW201033933A (en) 2009-03-06 2009-03-06 Scheduling method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
TW98107291A TW201033933A (en) 2009-03-06 2009-03-06 Scheduling method

Publications (1)

Publication Number Publication Date
TW201033933A true TW201033933A (en) 2010-09-16

Family

ID=44855350

Family Applications (1)

Application Number Title Priority Date Filing Date
TW98107291A TW201033933A (en) 2009-03-06 2009-03-06 Scheduling method

Country Status (1)

Country Link
TW (1) TW201033933A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
TWI701621B (en) * 2019-10-15 2020-08-11 薩摩亞商紘織國際有限公司 A system that combines dynamic production and knitting machine work management

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
TWI701621B (en) * 2019-10-15 2020-08-11 薩摩亞商紘織國際有限公司 A system that combines dynamic production and knitting machine work management

Similar Documents

Publication Publication Date Title
US20080195241A1 (en) Method and system for wafer lot order
Gilland A simulation study comparing performance of CONWIP and bottleneck-based release rules
CN104915742B (en) Method and apparatus based on user preference distribution resource
CN111356988B (en) Method and system for scheduling semiconductor processing
CN108053047A (en) Cloud resources of production scheduling methods, devices and systems
CN113554363B (en) Power customer service work order processing method and system based on grid system monitoring
Singh et al. Experimental investigation for performance assessment of scheduling policies in semiconductor wafer fabrication—a simulation approach
Thenarasu et al. Development and selection of hybrid dispatching rule for dynamic job shop scheduling using multi-criteria decision making analysis (MCDMA)
Needy et al. A cost model for the evaluation of alternative cellular manufacturing configurations
TW201033933A (en) Scheduling method
Joo et al. Dispatching rule-based algorithms for a dynamic flexible flow shop scheduling problem with time-dependent process defect rate and quality feedback
Park et al. Fab simulation with recipe arrangement of tools
CN1529261A (en) Regular scheduling method with state and performance feedback and its scheduling system
CN100383914C (en) Method for determining machinery processing sequence and manufacturing system therewith
TW573270B (en) Multiple constrains planning and scheduling system
Tang et al. Virtual production lines design for back-end semiconductor manufacturing systems
Aguirre et al. Applying a simulation-based tool to productivity management in an automotive-parts industry
Wong et al. Real-time GA-based rescheduling approach for the pre-sewing stage of an apparel manufacturing process
Lohmer et al. Order release methods in semiconductor manufacturing: State-of-the-art in science and lessons from industry
TWI663557B (en) Manufacture schedule variation minimizing system
Norozi et al. An optimization technique using hybrid GA-SA algorithm for multi-objective scheduling problem
TW440988B (en) Device and method for dispatching chips
Sun et al. DC-WIP—a new release rule of multi-orders for semiconductor manufacturing lines
TW200412518A (en) Front end dispatching method and system for long batch processing equipment in semiconductor manufacturing
Montoya-Torres et al. A consistent approach for vehicle planning and control in large unified automated material handling systems