200800459 (1) 九、發明說明 【發明所屬之技術領域】 本發明是有關雷射加工裝置，特別是關於具有奈 求以上的精度之雷射加工裝置。 【先前技術】 一般，雷射加工裝置是具備： Φ 照射至被加工物的雷射光源；及 配備載置被加工物的可動台之具有耐震性的精密 且具有固定自雷射光源輸出的雷射光的基準點， 述可動台移動於精密台上的χ·γ平面上，而來加工 工物之步驟。 在該加工步驟中，可想像有檢測出精密台上的可 的位置之位置檢測器的的誤差、及從雷射光源到被加 的導光路所引起之雷射光的光軸的誤差，但可動台的 • 檢測器的精度可實現1 Onm以下的精度，從雷射光源 的雷射光的光軸的安定精度可實現到微弧度（i . radian)水準的精度。 在專利文獻1中揭示有可測定雷射光的照射位置 射加工裝置的集光光學系的集光位置檢測裝置。 該集光位置檢測裝置是設有： 複數個受光元件會被配置成矩陣狀，可裝卸配 X_ Y台的特定位置之位置感測器；及 朝該位置感測器照射雷射光時，檢測出構成位置 米要 台， 使上 被加 動台 工物 位置 輸出 n i c r 〇 之雷 置於 感測 -4- (2) 200800459 器的各受光元件的輸出，判別雷射光的集光位置之判別裝 置； 根據該判別結果，算出對上述χ·γ台的座標軸之集 光位置的偏差量之運算裝置。 在該裝置中，以X-Y台的χ-γ方向的座標軸與位置 感測器的X-Y方向的座標軸能夠一致之方式，將位置感 測器安裝於X-Y台，使X-Y台移動至特定的位置，將雷 φ 射光集中照射至位置感測器。然後，檢測出構成位置感測 器的各受光元件的輸出，判別哪個受光元件接受雷射光， 求取來自受光元件的位置感測器的原點的座標位置，藉此 求取雷射光的集光位置。 〔專利文獻1〕特開平6-23 5 77號公報 【發明內容】 (發明所欲解決的課題） # 專利文獻1的雷射光的位置檢測裝置是配置於裝卸可 能的可動台上，若爲具有微米單位的精度之加工技術，則 雷射光的光軸的安定性即使爲微弧度水準還是能夠容許， 因此可適切地控制可動台的位置。 但，當加工精度爲奈米單位以下時，雷射加工裝置的 誤差，雖可動台的位置檢測器之誤差可容許，但雷射光的 光軸的搖晃不被容許，雷射光的照射點的安定性所引起因 的誤差會形成問題。 本發明的目的是在於提供一種具備用以監視雷射光的 -5- 200800459 (3) 光軸的位置之光軸位置檢測器，校正照射至被加工物的雷 射光的光軸偏差，藉此可對應於奈米單位以下的加工精度 之雷射加工裝置。 (用以解決課題的手段） (1 )爲了達成上述目的，本發明之雷射加工裝置係 具備： 光源，其係輸出雷射光； 光路調整部，其係調整該雷射光的光路； 第1分光器，其係將雷射光分光； 可動台，其係具有特定的可動域，載置被加工物； 可動台位置檢測器，其係檢測出可動台的位置； 第1光軸位置檢測器，其係檢測出上述所被分光之雷 射光的光軸的位置； 平台，其係載置上述可動台、可動台位置檢測器、及 第1光軸位置檢測器；及 光軸控制部，其係接受來自上述第1光軸位置檢測器 的輸出，而控制上述光路調整部，調整上述分光後的雷射 光的光軸。 (2)又，本發明之雷射加工裝置的光軸控制部，較 理想是進行模糊控制（F u z z y C ο n t r ο 1 )。 (3 )又，本發明之雷射加工裝置，較理想是更具備 第2分光器，其係將上述分光後的雷射光再分光；及 -6 - 200800459 (4) 第2光軸位置檢測器，其係用以檢測出藉由第2分光 器所分光之光軸的位置， 以第1光軸位置檢測器及第2光軸位置撿測器中雷射 光能夠在特定的位置被檢測出之方式來控制上述光軸控制 部。 ' (4)又，本發明之雷射加工裝置的第1光軸位置檢 測器或第2光軸位置檢測器的至少一方，較理想是藉由在 • 第1分光器所被分光的雷射光或在第2分光器所被分光的 雷射光反射於球狀的反射面而以4象限感測器來受光，進 行各雷射光的光軸的位置檢測。 (5) 又，本發明之雷射加工裝置，較理想是在上述 第1分光器與第1光軸位置檢測器之間，或上述第2分光 器與第2光軸位置檢測器之間的至少一方配置具有用以雷 射光通過的特定口徑之針孔。 Φ 〔發明的效果〕 (1 )請求項1的發明是在將自雷射光源輸出的雷射 _ 光照射至被加工物爲止的導光路中，由於雷射光的光軸會 產生微弧度程度的偏動，照射點的位置會產生誤差，因此 ' 例如將由配置成矩陣狀的受光元件所構成的第1光軸位置 檢測器配置於可動台的附近，且將位置檢測用的雷射光予 以分光，使分光後的雷射光的光軸受光於該第1光軸位置 檢測器，檢測出受光後的雷射光的光軸的基準位置，藉此 算出自基準位置的偏差，根據該算出後的値來調整光路調 -1 · (5) (5)200800459 整部，而修正導光路，藉此可實現對應於奈米單位以下的 精度之雷射加工裝置。 (2)請求項2的發明，因爲進行模糊控制，所以即 使雷射光的光軸產生偏差，還是可以漸漸地校正光軸的偏 差，以最小的誤差來進行雷射加工。 (3 )請求項3的發明是在第1光軸位置檢測器及第 2光軸位置檢測器中以雷射光能夠在特定的位置被檢測出 之方式來控制上述光軸控制部，因此雷射光的光軸的基準 點會被設置兩處，例如以能夠將該兩個基準點維持成等間 隔的方式來調整光路調整部，校正光路，使能夠提高校正 精度。 (4 )請求項4的發明是藉由反射於球狀的反射面後 受光於4象限感測器來進行各雷射光的光軸的位置檢測， 因此可使具有特定光徑的雷射光照射至各受光元件，以受 光強度（受光面積）能夠均等維持的方式來校正光軸。並 且，可將雷射光的受光面設定於任意處，雷射加工裝置的 設計具有彈性。 (5)請求項5的發明是配置具有雷射光通過用的特 定口徑之針孔，因此可將雷射光集中於所望的光徑。 【實施方式】 舉實施例參照圖面來說明有關本發明的較佳實施形態 。另外，在各圖中對相同的要素賦予同樣的符號，且有時 適當省略其説明。 -8- 200800459 ⑹ 〔實施例1〕 圖1是表示具有毫微（nano)單位精度的本發明之雷 射加工裝置1的全體構成槪略圖。該雷射加工裝置1是由 :光源2、光路調整部4、半透鏡（第1分光器）8、全反 射鏡13、可動台6、可動台位置檢測器1 2、第1光軸位 置檢測器（光檢測器）1 〇、平台7、光軸控制部5、集光 透鏡9，9,所構成。並且，所有的構成要素1〜12是配置 於防振台3上。 雷射光源2是以1000兆分之1秒雷射、或UV雷射 光作爲基本波輸出。光路調整部4(詳細參照圖8)具有 使自雷射光源2輸入的雷射光反射之複數個全反射鏡（圖 8中爲2個：22及24)，該全反射鏡中具有用以變更雷 射光的反射角之複數個馬達（圖8中爲4個：Μ 1〜M4 ) 。該馬達如後述，接受來自光軸控制部5的信號，調整全 反射鏡的反射面的角度，將雷射光定位於特定的光軸位置 。半透鏡8是配置於自光路調整部4輸出的雷射光的光軸 上，而將該雷射光予以分光。可動台6是在載置固定被照 射雷射光的被加工物1 1的狀態下，可移動於特定的可動 域内，該雷射光是被反射於全反射鏡13，且被集中於集 光透鏡9 ’。可動台位置檢測器1 2能以可動台6的移動誤 差爲1 Onm的精度來進行位置檢測。第1光軸位置檢測器 (光檢測器）1 〇，爲了在加工用雷射光的附近精確地調整 雷射光的光軸，而配置於可動台位置檢測器1 2的附近， -9- 200800459 (7) 例如由配置成矩陣狀的CCD之畫素等的複數個受光元件 所構成。平台7是以集中於集光透鏡9的雷射光的光軸能 夠照射於第1光軸位置檢測器1 0的受光面，集中於集光 透鏡9 ’的雷射光的光軸能夠對被加工物1 1照射之方式， 將可動台6、可動台位置檢測器1 2、第1光軸位置檢測器 ' 10配置於同一平面上。光軸控制部5具備調整値算出手 段2 6及馬達控制部2 8 (參照圖8 )，如後述，根據第1 φ 光軸位置檢測器1 〇所接受後的雷射光的受光強度的輸出 來模糊控制上述光路調整部4，使上述分光後的雷射光的 光軸固定於一定的基準位置。 其次，說明有關從本發明的雷射加工裝置1的雷射光 源2輸出的雷射光的導光路。 首先，從光源2往光路調整部4輸出雷射光。其次， 在光路調整部4，爲了照射至被加工物11，將雷射光的光 路調整成能夠固定於特定位置。而且，通過光路調整部4 • 後的雷射光是在半透鏡8中分光成通過此的雷射光、及從 半透鏡8分岐的雷射光。通過半透鏡8的雷射光是被反射 於全反射鏡1 3而經由集光透鏡9 5來照射被加工物1 1，施 以所望的加工。又，從半透鏡8分光的雷射光是經由集光 ' 透鏡9來照射第1光軸位置檢測器1 0。 如上述，形成本發明之雷射加工裝置的導光路。 其次，參照圖8來詳述有關調整從圖1的雷射光源2 到第1光軸位置檢測器1 0的光路之機構。 光路調整部4是具備2個的全反射鏡22、24，各全 -10- 200800459 ⑹ 反射鏡22、24是分別具備用以調整反射面的角度之馬達 Ml、M2及M3、M4，該反射面是使從雷射光源2輸出的 雷射光反射。並且，分光器44是具備使射入的雷射光分 光之半透鏡8、及集光透鏡9。 自雷射光源2輸出的雷射光，如前述，會被射入光路 調整部4，然後藉由全反射鏡22、24來定位於特定的光 軸位置而輸出。所被輸出的雷射光會被輸入分光器‘ 44的 半透鏡8。半透鏡8會從射入被加工物（未圖示）的雷射 光來分光射入第1光軸位置檢測器1 0的雷射光。所被分 光的雷射光會被射入集光透鏡9,集中於第1位置檢測器 1〇的特定受光元件（參照圖2〜圖4)後照射。第1光軸 位置檢測器1 0的受光元件會將受光後的雷射光變換成與 其強度成比例的電氣信號輸出。由於該被輸出的電氣信號 其強度一般較弱，因此會藉由放大器34來放大後輸入至 光軸控制部5内的調整値算出手段26。調整値算出手段 26會根據藉由放大器所放大的電氣信號，利用模糊推論 來分別算出上述馬達Ml〜M4的旋轉量。有關利用模糊推 論來算出馬達Ml〜M4的旋轉量（調整量）的構成會在往 後敘述。在調整値算出手段26所被算出之馬達Ml〜M4 的旋轉量會傳送至控制馬達Ml〜M4的馬達控制部28。 其次，從馬達控制部28來將顯示各旋轉量的信號傳送至 各馬達Ml〜M4，驅動馬達Ml〜M4。藉此，全反射鏡22 及2 4的各個反射面的方向會被調整，而調整雷射光的光 路或光軸。其次，參照圖2〜圖4來說明有關第1光軸位 -11 - 200800459 Ο) 置檢測器1 〇的受光元件。 在圖2及圖3中，各畫素21是被配置成矩陣狀，畫 素間隔A約爲2 μηι。圖2是表示雷射光的照射點會僅以} 晝素作爲基準畫素來完全照射的實施例（符號2 〇 )，根 據該基準畫素來測定雷射光的受光強度。然後，當雷射光 ' 的光軸產生搖晃，而有未照射至畫素的部分發生時，受光 元件會感測到受光強度的減少，完全照射基準畫素，以受 φ 光強度能夠形成極大的方式來藉由光軸控制部5進行光路 調整部4的調整。又，圖3是表示雷射光的照射點會均等 地照射形成4象限感測器的4個畫素的一部份之實施例（ 符號3 0 )，設定應均等地照射各畫素2 1的照射面積s 1〜 S 4之雷射光的基準照射點。然後，當雷射光的光軸產生 搖晃，各畫素21的照射面積S1〜S4發生不均衡時，以光 軸控制部5能夠使各畫素21的受光強度形成均等的方式 來調整光路調整部4。 ^ 圖4是揭示有關第1光軸位置檢測器1 〇的受光元件 的配置之變形實施例。該實施例是在集光透鏡9與第1光 軸位置檢測器之間配置用以使雷射光反射的反射球41 ，使自半透鏡8分光的雷射光藉由反射球41的反射面來 ’ 反射於垂直方向（對該雷射光的光軸而言）’例如在形成 配置於與該光軸平行的面40的4象限感測器之4個畫素 中設定與圖3同樣的照射基準點。然後’當雷射光的光軸 產生搖晃，各晝素的照射面積發生不均衡時’以光軸控制 部5能夠使各晝素的受光強度形成均等的方式來調整光路 -12- 200800459 (10) 調整部4。 其次’參照圖5說明有關利用第1光軸位置檢測器及 第2光軸位置檢測器來檢測出雷射光的光軸的位置之變形 實施例。該實施例是在被分光於半透鏡8的雷射光的導光 路上，於半透鏡8與第1光軸位置檢測器1 〇之間配置第 ‘ 2半透鏡（第2分光器）16。第2半透鏡16會再將雷射 光分光，該被分光的雷射光會被反射於全反射鏡.1 8，而 φ 照射至第2光軸位置檢測器1 5。第2半透鏡1 6及第2全 反射鏡1 8是一起被支持於反射鏡支持台5 0。第2光軸位 置檢測器1 5可與第1光軸位置檢測器1 〇相同型態或相異 型態。如上述，任何的光軸位置檢測器皆是在受光元件中 設定照射雷射光的基準點，以2個基準點的距離能夠形成 相等之方式來使光軸控制部5運算，藉此控制光路調整部 4 〇 其次，參照圖6來說明有關圖4的變形實施例。圖6 • 中，在反射球4 1與配置形成4象限感測器的4個畫素的 面40之間配置針孔42的點是與圖4的實施形態不同。針 孔42可將雷射光的光徑縮小成所望的徑，藉由縮小形成 上述4象限感測器的4個畫素的相互間隔，可對應於徑'的 * 縮小來確保各畫素的照射面積。 圖7是表示在圖1的集光透鏡9與第1光軸位置檢測 器1 0之間配置以支持台7 1所支持的針孔72的狀態擴大 圖。在本實施形態中亦可縮小雷射光的光徑，照射至第1 光軸位置檢測器。 -13- 200800459 (11) 圖9是藉由在光路調整部4中調整雷射光的光軸，以 第1光軸位置檢測器1 〇的受光強度能夠安定地成極大値 之方式來進行控制的結果，定性地描繪其受光強度爲時間 性變化的狀態。橫軸爲時間，縱軸爲受光強度，分別爲任 意刻度。 • 如上述説明，實行調整從圖8的雷射光源2到第1光 軸位置檢測器1 〇爲止的光路或光軸的過程之結果，從往 Φ 第1光軸位置檢測裝置、〇的入射開始時刻經過時間t’後 入射光的強度會達到極大値，然後若在入射光的強度變化 所被容許設定的容許範圍（圖9中爲箭號所夾著的P範圍 )内變動，則可謀求雷射加工裝置的第1光軸位置檢測裝 置的輸出強度的安定化。藉此，持續該雷射加工裝置的作 動中之該光軸調整步驟，可謀求雷射光的光軸的安定化。 其次，參照圖1 〇所示的流程圖來更詳述從第1光軸 位置檢測器之雷射光的受光到光路調整部4的光軸調整爲 • 止的過程。 步驟S-1 0 :此步驟是控制開始步驟。根據來自該雷 射加工裝置的操作者或個人電腦等的指示，開始進行供以 使第1光軸位置檢測裝置的受光強度安定化成極大値的控 制。 步驟S-12 :此步驟是調整値算出手段26取得來自第 1光軸位置檢測器1〇的輸出之步驟。但，在設置放大器 34時，是調整値算出手段26取得來自放大器34的輸出 之步驟。往後爲了簡單說明，表記爲「來自第1光軸位置 -14- 200800459 (12) 檢測裝置1 0的輸出」，在設置放大器24時，是意指來自 放大器24的輸出。在此步驟中，在雷射光的光路的控制 開始之後緊接著測定在第1光軸位置檢測裝置10所取得 的受光強度。 步驟S-14 :此步驟是依序驅動馬達Ml〜M4的步驟 ▼ 。在馬達Ml〜M4中選定任意的馬達後開始。最初所被選 擇的馬達（在此爲Μ 1 )是針對圖8的全反射鏡22來使反 φ 射面的方向變化，在受光強度極大的位置固定馬達盤的旋 轉。其次所被選擇的馬達（在此爲M2 )同樣是針對全反 射鏡22來使反射面的方向變化，在受光強度極大的位置 固定馬達盤的旋轉。同樣地，馬達M3及Μ4是針對第2 全反射鏡24來使反射面的方向變化，在受光強度極大的 位置固定馬達盤，而針對第2全反射鏡24來固定反射面 的方向。 用以決定全反射鏡22及24的反射面的方向之馬達 φ Μ 1〜Μ4的旋轉量是根據後述的模糊推論來決定。在此所 用的模糊推論的算法是控制上述全反射鏡22及24的反射 面的方向而以馬達Μ 1〜Μ 4的旋轉量作爲參數來説明。 步驟S-16:此步驟是爲了確定馬達Ml〜Μ4的旋轉 ^ 方向，而進行旋轉驅動的試行驅動步驟。 步驟S-18:此步驟是取得與照射至第1光軸位置檢 測器1 〇的雷射光的受光強度成比例的信號之步驟。 在上述的步驟S-16及S-18中，若得知藉由使馬達旋 轉於特定的方向，在第1光軸位置檢測器1 0受光的受光 -15- 200800459 (13) 強度増加，則表示該馬達的旋轉是受光強度成極大的方向 。相反的，若得知第1光軸位置檢測器1 〇所受光的受光 強度減少，則表示該馬達的旋轉是與受光強度成極大的方 . 向相反的方向。 步驟S-20:此步驟是在於計算第1光軸位置檢測器 10的輸出信號的時間微分値、及離目標値（極大値）的 偏差量之步驟。就此步驟而言，是在模糊推論中，計算作 ® 爲輸入値利用之輸出信號的時間微分（差分）値，計算離 目標値（極大値）的偏差量。若來自第1光軸位置檢測器 1 〇的時刻t i之輸出信號的値爲s i，時刻12之輸出信號的 大小爲s2，則假定，輸出信號的時間差分値S’爲 S’WSrSd/A-tO。並且，當目標値（極大値）爲S()時， 計算AS = (Sl/S())-l之離目標値的偏差量（離目標値的偏差 比例）AS。利用S ’及AS來進行模糊推論。 步驟S - 2 2 :此步驟是在於計算模糊推論之馬達的驅 ^ 動量（旋轉量）的步驟。詳細會在往後敘述，此步驟是利 用上述· S ’及AS的値來進行模糊推論，計算馬達的驅動量 - (旋轉量）的絕對値Μ。 步驟S-24 :此步驟是在於求取馬達的驅動方向（旋 轉方向）之步驟。若在上述步驟S-20所求得的S，的値爲 負’必須使馬達的驅動方向（旋轉方向）反轉。另一方面 ’若S，的値爲正，則馬達的旋轉方向原封不動即可。此 步驟是以其次的程序來求取上述馬達的旋轉方向。亦即， 決定馬達的旋轉方向的參數爲α。α是取値1或値-1者。 -16- 200800459 (14) 並且，以其次的方式來決定參數δ。若在上述步驟S-20 所求得的S ’的値爲負，貝tj δ = -1，若S ’的値爲正，貝|j δ= 1 。然後，該馬達的下次旋轉方向爲α χ δ。亦即，藉由將該 α X δ値設定成下次新的參數α的値’來確定馬達的下次旋 轉方向。若亦包含馬達的旋轉方向來表示旋轉量，則以 αχ Μ表示 0 步驟S-26 :此步驟是驅動馬達的步驟，使馬達只旋 • 轉上述α χ Μ。 步驟S-28:此步驟是與上述步驟S-18同樣，取得與 照射至第1光軸位置檢測器1 〇的受光強度成比例的信號 之步驟。 步驟S-30:此步驟是根據與在上述步驟S-28所取得 的受光強度成比例的信號的値來判斷是否終了至目前爲止 的步驟所控制及調整後的馬達的調整作業，前進至控制其 次的馬達的步驟。與上述步驟S-28所取得的受光強度成 ® 比例的信號的値若在視爲目標値（極大値）大小的範圍內 (圖9中箭號所夾著的Ρ値範圍），則爲了進行其次的馬 • 達控制’而切換控制對象的馬達。然後，前進至下個步驟 ，亦即步驟S - 3 2。另一方面，若判定與步驟s - 2 8所取得 的受光強度成比例的信號的値未達目標値，則回到步驟 S-20。 步驟S-32 :此步驟是進行判定是否終了光路調整部4 的調整作業之步驟。若確定對馬達Μ 1〜Μ4的調整作業完 全終了，則前進至下個步驟S-34，使調整作業終了。另 -17- 200800459 (15) 一方面，若未終了持續進行控制的話，則回到上述步驟 S-14。即使被確定對上述馬達Ml〜M4的調整作業完全終 了，爲了對應於經時變化，在驅動此雷射加工裝置的期間 ，不使光路調整部4的調整作業終了的判斷亦有可能。 步驟S-34 :此步驟是使光路調整部4的調整作業終 _ 了之步驟。 _ <模糊推論〉 參照圖11(A1)〜（A4)及（B1)〜（B4)、圖12 (A1 )〜（A3 )及（B1 )〜（B3 )來說明有關使用於爲 了該雷射加工裝置的雷射光的光軸調整而實行的模糊推論 之歸屬函數（Membership function)。以下，在指圖 11 (A1 )〜（A4 )及（B1 )〜（B4 )的所有圖時，僅記爲 圖1 1。又，同樣在指圖12 ( A1 )〜（A3 )及（B1 )〜（ B3 )的所有圖時，亦僅記爲圖12。 # 圖11是表示對第1光軸位置檢測器1 〇所檢測出的輸 出信號的時間微分（差分）値S5之歸屬（Membership) 函數。圖1 2是表示對第1光軸位置檢測器1 0的輸出信號 値的目標輸出値接近極大輸出値時的輸出信號的値AS之 ’ 歸屬函數。圖1 1所示的（A1 )〜（A4 )是表示模糊推論 的前件部，（B1)〜（B4)是表示分別對應於前件部（ A1 )〜（A4 )的後件部。並且，在圖12中亦同樣，(A1 )〜（A3 )是表示模糊推論的前件部，（B1 )〜（B3 ) 是表示分別對應於前件部（A1)〜（A3)的後件部。 -18- 200800459 (16) 即使在光路調整部4中進行光路調整，雷射光的光軸 還是會不安定產生微弧度程度的偏差，因此第1光軸位置 檢測器的受光強度會時間性地變動。如上述，此受光強度 的時間變化的狀態是藉由第1光軸位置檢測器1 0來觀測 。藉由第1光軸位置檢測器1 0所觀測的受光強度的時間 * 變化的狀態是以上述輸出信號的時間差分値S，，亦即、 S’WsysJ/Gyti)來表現。 • 因此，以能夠按照以下的規則（rule )(以下亦稱爲 「模糊規則」）之方式來定義作爲模糊推論的基礎之歸屬 函數。 規則1 1 :若s ’爲取正的値，其値大，則馬達的旋轉 量的絕對値大。 規則1 2 :若S ’爲取正的値，其値小，則馬達的旋轉 量的絕對値小。 規則1 3 .右S ’爲取〇的値，則馬達的旋轉量的絕對 # 値爲〇。 規則1 4 ·右S ’爲取負的値，則馬達的旋轉量的絕對 値小。 參照圖1 1來視覺性地說明上述規則。圖1 1所示的（ A1 )〜（八4)是表示上述模糊規則之各規則11〜14的前 件部。在圖11(A1)〜（A4)中，橫軸是表示s，，縱軸 是表示合致的程度（取0〜1的値之範圍）。另一方面， 圖11所示的（B1)〜（B4)是表示上述模糊規則之各規 貝[J Η〜14的後件部。橫軸是表示馬達的驅動量（旋轉量 -19- 200800459 (17) )的絕對値Μ，縱軸是表示合致的程度。 其次，說明有關第1光軸位置檢測器1 〇的輸出信號 値的目標輸出値接近最大輸出値時，目標値（極大値）爲 s〇時，對AS (該△Slsi/sd-i )的歸屬函數。在此，31是 時刻t!的輸出信號的値。利用對a s的歸屬函數的理由是 ' 基於以下的2點。 首先，針對第1點來進行説明。從雷射光源輸出的雷 Φ 射光爲高斯光束。高斯光束的性質上，對光束的中心附近 的強度的動徑方向之微分値小。而且，充分離開光束的中 心之處的強度的動徑方向的微分値亦小。亦即，往半透鏡 8 (分光器44)之雷射光的入射角度的對準大致正確時、 及對準大幅度偏離時，無論哪個情況，在光路調整部4中 所被進行的雷射光的光路調整效果是形成同程度的大小， 其效果小。換言之，在光路調整部4中爲了調整雷射光的 光路而使變化之全反射鏡22及24的反射面的方向的各單 • 位變化量之第1光軸位置檢測器1 〇所檢測出的受光強度 的變化比例爲同程度小。 . 亦即，對準大幅度偏離時，應以馬達的旋轉角度的絕 對値能夠變大之方式來設定，但若僅利用上述規則丨丨〜 14來進行模糊推論’則馬達的旋轉角度會小幅度被計算 。因此，藉由對AS的歸屬函數設定新的規則，可使馬達 的旋轉角度的大小適正化。不過，即使不設定該新的規則 ’目的之光學系的調整還是可進行。只是所被計算之馬達 的旋轉角度的値小，所以將光學系調整至最佳狀態的時間 -20- 200800459 (18) 會花費長（控制的步驟多）。 其次，針對第2點來進行説明。光路調整機能可藉由 上述新規則的設定來提高對雜訊的耐久性，即使在第1光 軸位置檢測器1 0中所被檢測出的受光強度中有某雜訊混 入也不會有問題。但，假使只靠規則1 1〜1 4，除此以外 沒有設置新的規則時，若第1光軸位置檢測器10所檢測 後的受光強度的値中混入有雜訊，則S ’的値會特別異常 • 地形成大的値，有時馬達的旋轉角度的値會不適當地成較 大的値算出，會有無法適當控制的可能性。 因此，若事先設定以下所示的新規則，則即使在受光 強度的値中混入有雜訊的事態發生，還是能夠排除上述的 可能性。 因此，以能夠按照以下的模糊規則（新規則）之方式 來定義有關作爲模糊推論的基礎之AS的歸屬函數。 規則2 1 :若第1光軸位置檢測器1 〇所檢測出的受光 # 強度的信號比目標値（極大値）s〇還要非常小（AS的値 爲負的値，其絕對値大），則馬達的旋轉角度大。 規則22 :若第1光軸位置光檢測器1 〇所檢測出的受 光強度的信號對目標値（極大値）sG而言大致同程度（ △S的値爲負的値，其絕對値小），則馬達的旋轉角度小 〇 規則23:若第1光軸位置檢測器10所檢測出的受光 強度的信號達到目標値（極大値）so或之上（AS的値比 〇大），則馬達的旋轉角度爲0。 -21 - 200800459 (19) 參照圖1 2來視覺性地說明上述新的規則。圖1 2所示 的（A1 )〜（A3 )是表示上述模糊規則之各個規則21〜 23的前件部。在（A1 )〜（A3 )中，橫軸是表示AS，縱 軸是表示合致的程度（取0〜1的値之範圍）。另一方面 ，圖12所示的（B1)〜（B3)是表示上述模糊規則之各 ' 個規則21〜23的後件部。橫軸是表示馬達的驅動量（旋 轉量）的絕對値Μ，縱軸是表示合致的程度。 φ 藉由模糊推論來計算馬達的驅動量（旋轉量）的手法 ，在此可利用min-max合成重心法。若藉由第1光軸位置 檢測器來檢測出受光強度，則可根據其値來求取S ’及Δ8 。今，假設S’及AS的値，說明求取3、及AS!者。 圖1 3是供以說明根據規則1 1〜1 4之統合化的步驟。 在此圖1 3中，對應於規則1 1〜1 4的歸屬函數是再記錄與 圖1 2所示的歸屬函數相同者。 由於 AS^ASi，因此在表示對應於圖13所示的規則 # 11〜14之歸屬函數的前件部的圖中，藉由縱的點線來表 示對上顯示AS的橫軸之的位置。由此圖可知，因爲 _ 在上述規則1 3及規則14中，前件部的適合度爲0，所以 後件部亦爲0。在上述規則11及規則12中，由於前件部 的適合度不爲〇，因此會使對應於該適合度來進行後件部 的歸屬函數的去頭處理。其結果，規則11〜14的模糊推 論會被進行，求取圖1 3中作爲統合化1表示之後件部的 邏輯和（統合化1 )，作爲該等的結果。另外，顯示作爲 統合化1表示之後件部的邏輯和之函數是藉由合成進行規 -22- 200800459 (20) 則11及規則12的後件部的去頭處理後的歸屬函數來求取 〇 圖14是供以說明根據規則21〜23之統合化的步驟。 在此圖中，對應於規則2 1〜23的歸屬函數是再記錄與圖 13所示的歸屬函數相同者。 > 由於AS^ASi，因此在表示對應於圖14所示的規則 .21〜23之歸屬函數的前件部的圖中，藉由縱的點線來表 φ 示對上顯示AS的橫軸之AS!的位置。由此圖可知，因爲 上述規則2 1的適合度爲0，所以後件部亦爲0。在上述規 貝！I 22及規則23中，由於前件部的適合度不爲〇，因此會 使對應於該適合度來進行後件部的歸屬函數的去頭處理。 其結果，規則21〜23的模糊推論會被進行，求取圖14中 作爲統合化2表示之後件部的邏輯和（統合化2)，作爲 該等的結果。另外，顯示作爲統合化2表示之後件部的邏 輯和之函數是與上述統合化1時同樣，藉由合成進行規則 # 2 1及規則2 3的後件部的去頭處理後的歸屬函數來求取。 其次，進行對於規則11〜14(以下亦稱爲「第1規 . 則系列」）而言，重視規則21〜2 3 (以下亦稱爲「第2 規則系列」）多少，或均等地重視第1及第2系列之附加 重要性的處理。分別使作爲上述統合化1及統合化2而取 得的結果（圖1 3及圖1 4中分別作爲統合化丨及統合化2 來表示之作爲後件部的邏輯和而求取的合成歸屬函數）形 成r倍及（1-r)倍，藉此來對各個的函數進行附加重要 性，如圖15 ( A )〜（D)所示，使兩者統合化。 -23- 200800459 (21) 在此，r是取〇〜1的値範圍的實數値。例如，選擇 r- 1，疋對應於只取入第1規則系列，而無視第2規則系 列。又，選擇r = 0.5，是意指同等處理第i規則系列及第 2規則系列。又，選擇r = 〇，是對應於只取入第2規則系 列，而無視第1規則系列。 圖1 5 ( A )〜（D )是供以說明求取統合化3作爲上 述圖1 3及圖1 4中統合化丨及統合化2的歸屬函數的邏輯 φ 和的步驟。圖1 5 ( A)是作爲統合化i而求取的合成歸屬 函數的槪略形狀，圖15(B)是作爲統合化2而求取的合 成歸屬函數的槪略形狀。圖1 5 ( C)是使作爲統合化i而 求取的合成歸屬函數形成r倍，使作爲統合化2而求取的 合成歸屬函數形成（Ι-r )倍後合成的統合化3之歸屬函 數的槪略形狀。圖1 5 ( D)是用以說明求取在圖1 5 ( c ) 所被賦予的歸屬函數的合成重心的値，採用該合成重心的 値作爲馬達的驅動量（旋轉角度）的程序。在圖1 5 ( D ) φ 中，橫軸上以Μ箭頭所示之橫軸的値是從圖i 5 ( c )所 示的歸屬函數求取的合成重心的位置，該位置會顯示馬達 的旋轉角度。 亦即，可藉由進行上述模糊推論，調整雷射光的光路 ，求取爲了使光路調整部4的全反射鏡的反射面等的角度 變化而驅動之馬達的旋轉角度。 在上述説明中是針對第1規則系列的規則1 1〜14的 各規則或第2規則系列的規則2 1〜2 3的各規則來均等地 處理，但亦可在該等的規則間對重視的程度附上輕重。此 -24 - (22) 200800459 情況，只要針對對應於第1規則系列的規則11〜14的各 規則或第2規則系列的規則2 1〜23的各規則之歸屬函數 乘以相當於上述r的參數，然後進行統合化即可。 又，上述模糊推論中是利用min-max合成重心法來求 取馬達的旋轉角度的値，但並非限於此方法，亦可採用代 數積-加算重心法等模糊推論的方法之其他習知的方法。 只要根據經驗等，採用對模糊推論控制的對象之雷射加工 裝置最適合的方法即可。 其次，在表1及表2中，分別將針對使用於上述模糊 推論的第1規則系列及第2規則系列的參數彙整於一覧表 。由表1及表2所示的參數可明確得知，未設定特別複雜 的模糊規則。即使如此，只要實行根據上述模糊推論的控 制’便可確定雷射加工裝置的光學系的對準能夠簡單地實 現。200800459 (1) Description of the Invention [Technical Field] The present invention relates to a laser processing apparatus, and more particularly to a laser processing apparatus having an accuracy higher than that of the present invention. [Prior Art] Generally, a laser processing apparatus is provided with: Φ a laser light source that is irradiated onto a workpiece; and a radar having a shock-proof precision and a fixed lightning output from a movable table on which a workpiece is placed. The reference point of the light emission, the step of processing the workpiece by moving the movable table on the χ·γ plane on the precision stage. In this processing step, it is conceivable that there is an error in the position detector detecting the position on the precision stage, and an error in the optical axis of the laser light caused by the laser light source to the applied light guide path, but movable The accuracy of the detector can achieve an accuracy of 1 Onm or less, and the accuracy of the optical axis of the laser light from the laser source can be adjusted to the accuracy of the micro radian. Patent Document 1 discloses a light collecting position detecting device that can measure the irradiation position of a laser beam at a light collecting optical system. The concentrating position detecting device is provided with: a plurality of light receiving elements arranged in a matrix shape, and a position sensor capable of attaching and detaching a specific position of the X_Y stage; and detecting the laser light when the position sensor is irradiated with the laser light a device for determining the position of the light meter, and the device for determining the light collecting position of the laser light by the output of the light receiving element of the -4- (2) 200800459 device; Based on the result of the discrimination, a calculation device for calculating the amount of deviation of the collected position of the coordinate axis of the χ·γ table is calculated. In this device, the position sensor is attached to the XY stage so that the coordinate axis of the XY stage in the χ-γ direction and the coordinate axis of the position sensor in the XY direction can be moved, and the XY stage is moved to a specific position. The ray ray is concentrated to illuminate the position sensor. Then, the output of each of the light-receiving elements constituting the position sensor is detected, and it is determined which light-receiving element receives the laser light, and the coordinate position of the origin from the position sensor of the light-receiving element is obtained, thereby obtaining the light collection of the laser light. position. [Patent Document 1] Japanese Laid-Open Patent Publication No. Hei 6-23 No. 5 77. SUMMARY OF THE INVENTION (Problems to be Solved by the Invention) # The position detecting device for laser light of Patent Document 1 is disposed on a movable table that can be attached or detached, and has With the processing technology of the accuracy of the micrometer unit, the stability of the optical axis of the laser light can be tolerated even at the micro-radius level, so that the position of the movable table can be appropriately controlled. However, when the machining accuracy is less than or equal to the nanometer unit, the error of the laser processing device is acceptable, but the error of the position detector of the movable table is acceptable, but the shaking of the optical axis of the laser light is not allowed, and the irradiation point of the laser light is stabilized. Errors caused by sex can cause problems. An object of the present invention is to provide an optical axis position detector including a position of a -5 - 200800459 (3) optical axis for monitoring laser light, thereby correcting an optical axis deviation of laser light irradiated to a workpiece, thereby A laser processing device corresponding to the processing accuracy below the nano unit. (Means for Solving the Problem) (1) In order to achieve the above object, a laser processing apparatus according to the present invention includes: a light source that outputs laser light; and an optical path adjustment unit that adjusts an optical path of the laser light; The device is configured to split the laser light; the movable table has a specific movable field and is placed on the workpiece; the movable table position detector detects the position of the movable table; and the first optical axis position detector And detecting a position of an optical axis of the laser beam to be split; and a platform for mounting the movable table, the movable table position detector, and the first optical axis position detector; and the optical axis control unit The output from the first optical axis position detector controls the optical path adjusting unit to adjust the optical axis of the spectrally separated laser light. (2) Further, the optical axis control unit of the laser processing apparatus of the present invention preferably performs fuzzy control (F z z y C ο n t r ο 1 ). (3) Further, the laser processing apparatus of the present invention preferably further includes a second beam splitter that splits the split laser light; and -6 - 200800459 (4) second optical axis position detector It is for detecting the position of the optical axis split by the second spectroscope, and the laser light can be detected at a specific position in the first optical axis position detector and the second optical axis position detector. The above-described optical axis control unit is controlled. (4) Further, at least one of the first optical axis position detector or the second optical axis position detector of the laser processing apparatus of the present invention is preferably laser light that is split by the first spectroscope The laser beam split by the second spectroscope is reflected by the spherical reflecting surface and received by the four-quadrant sensor, and the position of the optical axis of each of the laser beams is detected. (5) Further, the laser processing apparatus of the present invention is preferably between the first spectroscope and the first optical axis position detector or between the second optical splitter and the second optical axis position detector. At least one of the pins is provided with a pinhole having a specific aperture for the passage of the laser light. Φ [Effects of the Invention] (1) The invention of claim 1 is that the laser beam output from the laser light source is irradiated onto the light guide path until the workpiece is processed, and the optical axis of the laser light is slightly curved. Since the position of the irradiation spot is erroneous, the first optical axis position detector composed of the light receiving elements arranged in a matrix is disposed in the vicinity of the movable table, and the laser beam for position detection is split. The optical axis of the laser light after the splitting is received by the first optical axis position detector, and the reference position of the optical axis of the received laser light is detected, thereby calculating the deviation from the reference position, and based on the calculated 値Adjusting the optical path adjustment -1 (5) (5) 200800459 The entire section, and correcting the light guide path, can achieve the laser processing device corresponding to the precision below the nano unit. (2) In the invention of claim 2, since the blur control is performed, even if the optical axis of the laser light is deviated, the deviation of the optical axis can be gradually corrected, and the laser processing can be performed with a minimum error. (3) The invention of claim 3 is that the optical axis control unit is controlled such that the laser light can be detected at a specific position in the first optical axis position detector and the second optical axis position detector, so that the laser beam is emitted. The reference point of the optical axis is set to two. For example, the optical path adjusting unit can be adjusted so that the two reference points can be maintained at equal intervals, and the optical path can be corrected to improve the accuracy of the correction. (4) The invention of claim 4 is that the position of the optical axis of each of the laser beams is detected by being reflected by the spherical reflecting surface and then being received by the four-quadrant sensor, so that the laser light having the specific optical path can be irradiated to Each of the light receiving elements corrects the optical axis so that the received light intensity (light receiving area) can be uniformly maintained. Moreover, the light receiving surface of the laser light can be set to any position, and the design of the laser processing apparatus is flexible. (5) The invention of claim 5 is to arrange a pinhole having a specific aperture for the passage of laser light, so that the laser light can be concentrated on the desired optical path. BEST MODE FOR CARRYING OUT THE INVENTION A preferred embodiment of the present invention will be described with reference to the accompanying drawings. In the drawings, the same components are denoted by the same reference numerals, and the description thereof will be omitted as appropriate. -8- 200800459 (6) [Embodiment 1] Fig. 1 is a schematic diagram showing the overall configuration of a laser processing apparatus 1 of the present invention having nano unit precision. The laser processing apparatus 1 is composed of a light source 2, an optical path adjusting unit 4, a half mirror (first spectroscope) 8, a total reflection mirror 13, a movable table 6, a movable table position detector 1, and a first optical axis position detection. The device (photodetector) 1 , the stage 7, the optical axis control unit 5, and the collecting lenses 9 and 9 are formed. Further, all of the constituent elements 1 to 12 are disposed on the vibration isolating table 3. The laser light source 2 is a 1000 megahertz laser or UV laser light as a fundamental wave output. The optical path adjusting unit 4 (see FIG. 8 in detail) has a plurality of total reflection mirrors (two in the FIG. 8 : 22 and 24) that reflect the laser light input from the laser light source 2, and the total reflection mirror has a change A plurality of motors of the reflection angle of the laser light (four in Fig. 8: Μ 1 to M4). As will be described later, the motor receives a signal from the optical axis control unit 5, adjusts the angle of the reflection surface of the total reflection mirror, and positions the laser light at a specific optical axis position. The half mirror 8 is disposed on the optical axis of the laser light output from the optical path adjusting unit 4, and splits the laser light. The movable table 6 is movable in a specific movable region in a state in which the workpiece 1 1 to which the irradiated laser light is irradiated is placed, and the laser light is reflected on the total reflection mirror 13 and concentrated on the collecting lens 9 '. The movable table position detector 1 2 can perform position detection with the accuracy of the movement error of the movable table 6 being 1 Onm. The first optical axis position detector (photodetector) 1 配置 is disposed in the vicinity of the movable table position detector 1 2 in order to accurately adjust the optical axis of the laser light in the vicinity of the processing laser light, -9-200800459 ( 7) For example, it consists of a plurality of light-receiving elements such as a pixel of a CCD arranged in a matrix. The stage 7 is capable of illuminating the light receiving surface of the first optical axis position detector 10 with the optical axis of the laser light concentrated on the collecting lens 9, and the optical axis of the laser light concentrated on the collecting lens 9' can be used for the workpiece. In the case of 1 1 irradiation, the movable table 6, the movable table position detector 1 2, and the first optical axis position detector '10 are disposed on the same plane. The optical axis control unit 5 includes an adjustment 値 calculation means 26 and a motor control unit 28 (see FIG. 8), and as will be described later, based on the output of the received light intensity of the laser light received by the first φ optical axis position detector 1 〇 The optical path adjusting unit 4 is fuzzy controlled to fix the optical axis of the spectrally separated laser light to a predetermined reference position. Next, a light guiding path of the laser light output from the laser light source 2 of the laser processing apparatus 1 of the present invention will be described. First, laser light is output from the light source 2 to the optical path adjusting unit 4. Next, in the optical path adjusting unit 4, the optical path of the laser light is adjusted so as to be fixed to a specific position in order to be irradiated onto the workpiece 11. Further, the laser light that has passed through the optical path adjusting portion 4 is divided into laser light that has passed through the half mirror 8 and laser light that is branched from the half mirror 8. The laser light that has passed through the half mirror 8 is reflected on the total reflection mirror 13 and irradiates the workpiece 1 through the concentrating lens 95 to perform the desired processing. Further, the laser light split from the half mirror 8 is irradiated with the first optical axis position detector 10 via the collecting light lens 9. As described above, the light guiding path of the laser processing apparatus of the present invention is formed. Next, a mechanism for adjusting the optical path from the laser light source 2 of Fig. 1 to the first optical axis position detector 10 will be described in detail with reference to Fig. 8 . The optical path adjusting unit 4 includes two total reflection mirrors 22 and 24, and each of the total -10-200800459 (6) mirrors 22 and 24 are respectively provided with motors M1, M2, M3, and M4 for adjusting the angle of the reflecting surface. The surface reflects the laser light output from the laser light source 2. Further, the spectroscope 44 is provided with a half mirror 8 for splitting the incident laser light and a collecting lens 9. The laser light output from the laser light source 2 is incident on the optical path adjusting portion 4 as described above, and is then output by the total reflection mirrors 22, 24 at a specific optical axis position. The laser light to be output is input to the half mirror 8 of the beam splitter '44. The half mirror 8 splits the laser light incident on the first optical axis position detector 10 from the laser beam incident on the workpiece (not shown). The laser light to be split is incident on the collecting lens 9, and is concentrated on the specific light receiving element of the first position detector 1 (see Figs. 2 to 4). The light receiving element of the first optical axis position detector 10 converts the received laser light into an electrical signal output proportional to its intensity. Since the output electrical signal is generally weak in intensity, it is amplified by the amplifier 34 and input to the adjustment 値 calculating means 26 in the optical axis control unit 5. The adjustment 値 calculation means 26 calculates the amount of rotation of the motors M1 to M4 by fuzzy inference based on the electric signal amplified by the amplifier. The configuration for calculating the amount of rotation (adjustment amount) of the motors M1 to M4 by the fuzzy inference will be described later. The amount of rotation of the motors M1 to M4 calculated by the adjustment 値 calculation means 26 is transmitted to the motor control unit 28 that controls the motors M1 to M4. Next, the motor control unit 28 transmits a signal indicating the amount of rotation to each of the motors M1 to M4, and drives the motors M1 to M4. Thereby, the directions of the respective reflecting surfaces of the total reflection mirrors 22 and 24 are adjusted, and the optical path or optical axis of the laser light is adjusted. Next, the light receiving element of the detector 1 有关 will be described with reference to Figs. 2 to 4 regarding the first optical axis position -11 - 200800459 Ο). In Figs. 2 and 3, each of the pixels 21 is arranged in a matrix, and the pixel interval A is about 2 μm. Fig. 2 is a view showing an embodiment (symbol 2 〇) in which the irradiation spot of the laser light is completely irradiated with only the pixel as the reference pixel, and the received light intensity of the laser light is measured based on the reference pixel. Then, when the optical axis of the laser light is shaken and a portion that is not irradiated to the pixel occurs, the light-receiving element senses a decrease in the received light intensity, and completely illuminates the reference pixel to be greatly affected by the φ light intensity. In this manner, the optical axis adjustment unit 5 performs adjustment of the optical path adjustment unit 4. 3 is an embodiment (symbol 30) in which a part of four pixels forming a four-quadrant sensor is uniformly irradiated with an irradiation spot of the laser light, and the setting should uniformly illuminate each pixel 2 1 . The reference irradiation point of the laser light of the irradiation area s 1 to S 4 . Then, when the optical axis of the laser beam is shaken and the irradiation areas S1 to S4 of the respective pixels 21 are unbalanced, the optical axis control unit 5 can adjust the optical path adjustment unit so that the received light intensity of each pixel 21 is equalized. 4. Fig. 4 is a view showing a modified embodiment of the arrangement of the light receiving elements of the first optical axis position detector 1 。. In this embodiment, a reflecting ball 41 for reflecting the laser light is disposed between the collecting lens 9 and the first optical axis position detector, and the laser beam split from the half lens 8 is reflected by the reflecting surface of the reflecting ball 41. Reflected in the vertical direction (for the optical axis of the laser light), the illumination reference point similar to that of FIG. 3 is set, for example, in four pixels forming a four-quadrant sensor disposed on the surface 40 parallel to the optical axis. . Then, when the optical axis of the laser beam is shaken and the irradiation area of each element is unbalanced, the optical axis control unit 5 can adjust the optical path by equalizing the received light intensity of each element (-12). Adjustment unit 4. Next, a modified embodiment in which the position of the optical axis of the laser light is detected by the first optical axis position detector and the second optical axis position detector will be described with reference to Fig. 5 . In this embodiment, the second "half-lens (second beam splitter) 16 is disposed between the half mirror 8 and the first optical axis position detector 1 导 on the light guide path of the laser beam split by the half mirror 8. The second half lens 16 splits the laser light again, and the split laser light is reflected by the total reflection mirror .18, and φ is irradiated to the second optical axis position detector 15. The second half mirror 16 and the second total reflection mirror 18 are supported together by the mirror support table 50. The second optical axis position detector 15 can be of the same type or different type as the first optical axis position detector 1 。. As described above, any of the optical axis position detectors is configured such that the reference point for irradiating the laser light is set in the light receiving element, and the optical axis control unit 5 is operated so that the distance between the two reference points can be equalized, thereby controlling the optical path adjustment. Section 4 Next, a modified embodiment related to Fig. 4 will be described with reference to Fig. 6 . In Fig. 6, the point at which the pinholes 42 are disposed between the reflecting spheres 41 and the faces 40 of the four pixels in which the four quadrant sensors are arranged is different from that of the embodiment of Fig. 4. The pinhole 42 can reduce the optical path of the laser light to a desired diameter, and by reducing the mutual spacing of the four pixels forming the four-quadrant sensor, it is possible to ensure the illumination of each pixel in accordance with the reduction of the diameter '*. area. Fig. 7 is an enlarged view showing a state in which a pinhole 72 supported by the support base 7 is placed between the collecting lens 9 of Fig. 1 and the first optical axis position detector 10. In the present embodiment, the optical path of the laser light can be reduced and irradiated to the first optical axis position detector. (13) In the optical path adjusting unit 4, the optical axis of the laser beam is adjusted so that the received light intensity of the first optical axis position detector 1 can be stably stabilized. As a result, a state in which the received light intensity is temporally changed is qualitatively depicted. The horizontal axis is time and the vertical axis is received light intensity, which is an arbitrary scale. • As described above, the result of the process of adjusting the optical path or optical axis from the laser light source 2 of Fig. 8 to the first optical axis position detector 1 , is performed, and the Φ first optical axis position detecting device and the incidence of 〇 The intensity of the incident light after the elapse of the time t' will be extremely large, and then if the allowable range of the allowable change in the intensity of the incident light (the range of P sandwiched by the arrow in Fig. 9) is changed, The stability of the output intensity of the first optical axis position detecting device of the laser processing apparatus is stabilized. Thereby, the optical axis adjustment step in the operation of the laser processing apparatus can be continued, and the optical axis of the laser light can be stabilized. Next, the process of adjusting the optical axis of the laser light from the first optical axis position detector to the optical path adjusting portion 4 as described above will be described in more detail with reference to the flowchart shown in Fig. 1A. Step S-1 0: This step is a control start step. In response to an instruction from an operator or a personal computer or the like from the laser processing apparatus, control for setting the received light intensity of the first optical axis position detecting device to a maximum is started. Step S-12: This step is a step of adjusting the output of the first optical axis position detector 1A by the adjustment means 26. However, when the amplifier 34 is provided, the adjustment 値 calculation means 26 takes the step of obtaining the output from the amplifier 34. For the sake of simplicity, the description will be made "from the first optical axis position -14-200800459 (12) Output of the detecting device 10", and when the amplifier 24 is provided, it means the output from the amplifier 24. In this step, the received light intensity obtained by the first optical axis position detecting device 10 is measured immediately after the start of the control of the optical path of the laser light. Step S-14: This step is a step ▼ of sequentially driving the motors M1 to M4. Start after selecting any motor among the motors M1 to M4. The motor (here, Μ 1 ) which is initially selected is directed to the total reflection mirror 22 of Fig. 8 to change the direction of the inverse φ plane, and the rotation of the motor disk is fixed at a position where the received light intensity is extremely large. Next, the selected motor (here, M2) is also directed to the total reflection mirror 22 to change the direction of the reflecting surface, and the rotation of the motor disk is fixed at a position where the received light intensity is extremely large. In the same manner, the motor M3 and the cymbal 4 fix the direction of the reflecting surface with respect to the second total reflection mirror 24, fix the motor disk at a position where the received light intensity is extremely large, and fix the direction of the reflecting surface with respect to the second total reflection mirror 24. The amount of rotation of the motors φ Μ 1 to Μ 4 for determining the directions of the reflection surfaces of the total reflection mirrors 22 and 24 is determined based on a fuzzy inference described later. The algorithm of the fuzzy inference used here is to control the direction of the reflection surfaces of the total reflection mirrors 22 and 24 and to describe the rotation amount of the motors Μ 1 to Μ 4 as parameters. Step S-16: This step is a trial driving step of performing rotational driving in order to determine the rotational direction of the motors M1 to Μ4. Step S-18: This step is a step of obtaining a signal proportional to the received light intensity of the laser light irradiated to the first optical axis position detector 1 。. In the above-described steps S-16 and S-18, when the motor is rotated in a specific direction, the intensity of the light receiving light -15-200800459 (13) received by the first optical axis position detector 10 is increased. It is indicated that the rotation of the motor is a direction in which the received light intensity is extremely large. On the other hand, if the received light intensity of the light received by the first optical axis position detector 1 is reduced, it means that the rotation of the motor is extremely large with respect to the received light intensity. Step S-20: This step is a step of calculating the time differential 値 of the output signal of the first optical axis position detector 10 and the amount of deviation from the target 値 (maximum 値). For this step, in the fuzzy inference, calculate the time derivative (differential) 输出 of the output signal used as input 値 and calculate the deviation from the target 値 (maximum 値). If 値 of the output signal from the first optical axis position detector 1 ti is si, and the magnitude of the output signal at time 12 is s2, it is assumed that the time difference 输出S' of the output signal is S'WSrSd/A- tO. Also, when the target 値 (maximum 値) is S(), calculate the deviation amount (the deviation ratio from the target )) AS of AS = (Sl/S())-l from the target 値. Fuzzy inferences are made using S' and AS. Step S - 2 2: This step is a step of calculating the amount of driving (rotation amount) of the motor of the fuzzy inference. The details will be described later. In this step, the fuzzy inference is performed using the above-mentioned S' and AS, and the absolute value of the driving amount - (rotation amount) of the motor is calculated. Step S-24: This step is a step of obtaining the driving direction (rotation direction) of the motor. If 値 of the S obtained in the above step S-20 is negative, the driving direction (rotation direction) of the motor must be reversed. On the other hand, if the 値 of S is positive, the direction of rotation of the motor may be intact. This step is based on the next procedure to determine the direction of rotation of the motor. That is, the parameter that determines the direction of rotation of the motor is α. α is taken as 1 or 値-1. -16- 200800459 (14) Also, the parameter δ is determined in the second way. If 値 of S ′ obtained in the above step S-20 is negative, Bj δ = -1, and if 値 of S ′ is positive, B | δ = 1 . Then, the next rotation direction of the motor is α χ δ. That is, the next rotation direction of the motor is determined by setting the α X δ 成 to 値' of the next new parameter α. If the direction of rotation of the motor is also included to indicate the amount of rotation, it is represented by αχ 0. 0 Step S-26: This step is a step of driving the motor so that the motor only rotates the above α χ Μ. Step S-28: This step is a step of obtaining a signal proportional to the received light intensity of the first optical axis position detector 1 同样, similarly to the above-described step S-18. Step S-30: In this step, based on the signal proportional to the received light intensity obtained in the above step S-28, it is determined whether or not the adjustment operation of the motor controlled and adjusted by the step up to the present is completed, and the control proceeds to control. The second step of the motor. The signal of the signal proportional to the received light intensity obtained in the above step S-28 is within the range of the target 値 (maximum 値) size (the Ρ値 range enclosed by the arrow in Fig. 9), The second horse control is switched to the motor that controls the object. Then, proceed to the next step, step S - 3 2 . On the other hand, if it is determined that the signal which is proportional to the received light intensity obtained in the step s - 28 is less than the target 値, the process returns to the step S-20. Step S-32: This step is a step of determining whether or not the adjustment operation of the optical path adjusting unit 4 is completed. If it is determined that the adjustment operation for the motors Μ 1 to Μ 4 is completed, the process proceeds to the next step S-34 to end the adjustment operation. Another -17- 200800459 (15) On the one hand, if the control is not continued, return to the above step S-14. Even if it is determined that the adjustment operation for the motors M1 to M4 is completely completed, it is possible to determine that the adjustment operation of the optical path adjustment unit 4 is not completed during the period in which the laser processing apparatus is driven in response to the change over time. Step S-34: This step is a step of ending the adjustment operation of the optical path adjusting unit 4. _ <Fuzzy Inference> Referring to Figs. 11(A1) to (A4) and (B1) to (B4), Figs. 12(A1) to (A3) and (B1) to (B3), the description will be made regarding the use of the laser for the laser. A membership function of the fuzzy inference performed by adjusting the optical axis of the laser light of the processing device. Hereinafter, when referring to all the figures of Figs. 11 (A1) to (A4) and (B1) to (B4), only FIG. Also, when referring to all the figures of Figs. 12 (A1) to (A3) and (B1) to (B3), only FIG. 12 is also referred to. # Figure 11 is a membership function showing the time differential (difference) 値 S5 of the output signal detected by the first optical axis position detector 1 。. Fig. 12 is a home function of 値AS of the output signal when the target output 値 of the output signal 値 of the first optical axis position detector 10 is close to the maximum output 値. (A1) to (A4) shown in Fig. 11 are the front parts of the fuzzy inference, and (B1) to (B4) are the rear parts corresponding to the front parts (A1) to (A4), respectively. Further, in Fig. 12, (A1) to (A3) are the front parts of the fuzzy inference, and (B1) to (B3) are the latter which correspond to the front parts (A1) to (A3), respectively. unit. -18- 200800459 (16) Even if the optical path adjustment is performed in the optical path adjustment unit 4, the optical axis of the laser beam is unstable and the degree of micro-radiation is not changed. Therefore, the received light intensity of the first optical axis position detector changes temporally. . As described above, the state of temporal change of the received light intensity is observed by the first optical axis position detector 10. The state in which the time of the received light intensity observed by the first optical axis position detector 10 is changed is represented by the time difference 値S of the output signal, that is, S'WsysJ/Gyti). • Therefore, the attribution function that is the basis of the fuzzy inference can be defined in such a way as to be able to follow the following rule (hereinafter also referred to as "fuzzy rule"). Rule 1 1 : If s ' is a positive 値 and its 値 is large, the amount of rotation of the motor is absolutely large. Rule 1 2: If S ′ is a positive 値 and its 値 is small, the amount of rotation of the motor is absolutely small. Rule 1 3 . The right S ′ is the 値 of the 値, and the absolute # 値 of the amount of rotation of the motor is 〇. Rule 1 4 · Right S ' is a negative 値, and the amount of rotation of the motor is absolutely small. The above rules are visually explained with reference to FIG. (A1) to (8) shown in Fig. 11 are the front parts of the respective rules 11 to 14 of the above fuzzy rule. In Figs. 11(A1) to (A4), the horizontal axis represents s, and the vertical axis represents the degree of convergence (the range of 値 from 0 to 1). On the other hand, (B1) to (B4) shown in Fig. 11 are the posterior parts of the respective rules of the fuzzy rule [J Η 14 14]. The horizontal axis indicates the absolute value of the motor drive amount (rotation amount -19-200800459 (17)), and the vertical axis indicates the degree of convergence. Next, it is explained that when the target output 値 of the output signal 値 of the first optical axis position detector 1 値 approaches the maximum output ,, when the target 値 (maximum 値) is s ,, the pair AS (the ΔSlsi/sd-i ) Ownership function. Here, 31 is the 输出 of the output signal at time t!. The reason for using the attribution function for a s is 'based on the following two points. First, the first point will be described. The lightning Φ output from the laser source is a Gaussian beam. In the nature of the Gaussian beam, the differential of the direction of the dynamics of the intensity near the center of the beam is small. Moreover, the differential enthalpy of the direction of the strength at the center of the beam sufficiently away from the center of the beam is also small. In other words, when the alignment of the incident angle of the laser light to the half mirror 8 (the beam splitter 44) is substantially correct and the alignment is largely deviated, the laser light that is carried out in the optical path adjusting unit 4 is in any case. The effect of the optical path adjustment is to form the same size, and the effect is small. In other words, in the optical path adjusting unit 4, the first optical axis position detector 1 is detected by the first optical axis position detector 1 in the direction of the change surface of the total reflection mirrors 22 and 24 in order to adjust the optical path of the laser light. The proportion of change in received light intensity is small to the same extent. That is, when the alignment is largely deviated, it should be set such that the absolute 値 of the rotation angle of the motor can be increased. However, if only the above-mentioned rules 丨丨~14 are used for the fuzzy inference, the rotation angle of the motor will be small. The amplitude is calculated. Therefore, by setting a new rule for the AS's attribution function, the magnitude of the rotation angle of the motor can be normalized. However, even if the new rule is not set, the optical system adjustment can be performed. It is only the smaller the angle of rotation of the motor being calculated, so the time to adjust the optical system to the optimum state -20- 200800459 (18) will take a long time (more steps of control). Next, the second point will be described. The optical path adjustment function can improve the durability of the noise by the setting of the above-mentioned new rule, and there is no problem even if some noise is mixed in the received light intensity detected by the first optical axis position detector 10. . However, if only a new rule is not set in the rule 1 1 to 1 4, if the noise of the received light intensity detected by the first optical axis position detector 10 is mixed with noise, then S '値It will be particularly anomalous. • A large flaw is formed in the ground. Sometimes the angle of rotation of the motor will be improperly calculated as a large flaw, and there is a possibility that it cannot be properly controlled. Therefore, if the new rule shown below is set in advance, the above-described possibility can be eliminated even if a situation in which noise is mixed in the received light intensity is generated. Therefore, the attribution function of the AS as the basis of the fuzzy inference can be defined in such a manner that the following fuzzy rules (new rules) can be defined. Rule 2 1 : If the signal of the received light intensity detected by the first optical axis position detector 1 〇 is much smaller than the target 値 (maximum 値) s ( (the 値 of the AS is negative, it is absolutely large) , the motor has a large rotation angle. Rule 22: The signal of the received light intensity detected by the first optical axis position photodetector 1 大致 is approximately the same as the target 値 (maximum 値) sG (the △ of the ΔS is negative, which is absolutely small) Then, the rotation angle of the motor is small. Rule 23: If the signal of the received light intensity detected by the first optical axis position detector 10 reaches the target 値 (maximum 値) so or above (the 値 ratio of the AS is larger), the motor The rotation angle is 0. -21 - 200800459 (19) The above new rules are visually explained with reference to FIG. (A1) to (A3) shown in Fig. 12 are the front parts of the respective rules 21 to 23 indicating the above-described fuzzy rule. In (A1) to (A3), the horizontal axis represents AS and the vertical axis represents the degree of convergence (the range of 値 from 0 to 1). On the other hand, (B1) to (B3) shown in Fig. 12 are the latter parts of the respective rules 21 to 23 indicating the above-described fuzzy rule. The horizontal axis represents the absolute enthalpy of the amount of driving (rotation amount) of the motor, and the vertical axis represents the degree of merging. φ The method of calculating the driving amount (rotation amount) of the motor by fuzzy inference, where the center of gravity method can be synthesized using min-max. When the received light intensity is detected by the first optical axis position detector, S ′ and Δ8 can be obtained based on the 値. Now, suppose S' and AS's 値 indicate that they are seeking 3 and AS! Figure 13 is a step for explaining the integration according to rules 1 1 to 14 . In this Fig. 13, the attribution function corresponding to the rules 1 1 to 1 4 is the same as the attribution function shown in Fig. 12 . Since AS^ASi, in the figure indicating the front part corresponding to the attribution function of the rules #11 to 14 shown in Fig. 13, the position of the horizontal axis of the upper display AS is indicated by the vertical dotted line. As can be seen from the figure, since _ in the above rule 13 and rule 14, the suitability of the front part is 0, the rear part is also 0. In the above rules 11 and 12, since the suitability of the front part is not 〇, the heading process of the home function of the post part is performed corresponding to the suitability. As a result, the fuzzy inference of rules 11 to 14 is performed, and the logical sum (integral 1) of the subsequent part as the integration 1 in Fig. 13 is obtained as the result of the above. In addition, the function of displaying the logical sum of the subsequent parts as the integration 1 is obtained by synthesizing the belonging function of the header part of the rule 11-200800459 (20) 11 and the rule 12 to obtain the 归属 〇 Figure 14 is a diagram for explaining the integration according to rules 21 to 23. In this figure, the attribution function corresponding to the rules 2 1 to 23 is the same as the attribution function shown in Fig. 13 . > Since AS^ASi, in the diagram showing the front part corresponding to the attribution function of the rules .21 to 23 shown in Fig. 14, the vertical axis of the AS is shown by the vertical dotted line The location of the AS! As can be seen from the figure, since the suitability of the above rule 2 1 is 0, the rear part is also 0. In the above rules! In I 22 and Rule 23, since the suitability of the front part is not 〇, the heading process of the belonging function of the post part is performed corresponding to the suitability. As a result, the fuzzy inference of rules 21 to 23 is performed, and the logical sum (integration 2) of the subsequent part is shown as the result of integration 2 in Fig. 14 as the result of the above. In addition, the function of displaying the logical sum of the subsequent parts as the integration 2 is the same as the above-described integration 1, and the attribution function after the header processing of the rules #2 1 and the rule 2 3 is performed by combining. Seek. Next, for rules 11 to 14 (hereinafter also referred to as "the first rule." series), it is important to pay attention to rules 21 to 2 3 (hereinafter also referred to as "the second rule series"), or to pay equal attention to the Processing of the additional importance of 1 and 2 series. The results obtained by integrating the above-mentioned integration 1 and integration 2 (the composite attribution function obtained as the logical sum of the post-parts as shown in Fig. 13 and Fig. 14 as integrated and unified 2, respectively) In order to form r times and (1-r) times, the importance of each function is added, and as shown in Figs. 15(A) to (D), the two are integrated. -23- 200800459 (21) Here, r is a real number 値 of the range of 〇~1. For example, selecting r-1, corresponding to only taking the first rule series, ignores the second rule series. Also, selecting r = 0.5 means that the i-th rule series and the second rule series are treated equally. Also, selecting r = 〇 corresponds to only taking the second rule series and ignoring the first rule series. Fig. 15 (A) to (D) are steps for explaining the integration of the logical φ of the integration function 3 as the assignment function of the integration 丨 and the integration 2 in Fig. 13 and Fig. 14 described above. Fig. 15 (A) is a sketch shape of a composite belonging function obtained as the integration i, and Fig. 15 (B) is a sketch shape of a composite belonging function obtained as the integration 2. Fig. 15 (C) is a case where the synthetic attribution function obtained as the integration i is formed by r times, and the synthetic attribution function obtained as the integration 2 is formed (Ι-r ) times and then integrated. The approximate shape of the function. Fig. 15 (D) is a program for explaining the 重 of the combined center of gravity of the assignment function given in Fig. 15 (c), and using the 重 of the combined center of gravity as the driving amount (rotation angle) of the motor. In Fig. 15 (D) φ, the 値 of the horizontal axis indicated by the Μ arrow on the horizontal axis is the position of the resultant center of gravity obtained from the attribution function shown in Fig. i 5 (c), which shows the motor Rotation angle. In other words, the optical path of the laser beam can be adjusted by performing the above-described fuzzy inference, and the angle of rotation of the motor driven to change the angle of the reflecting surface of the total reflection mirror of the optical path adjusting portion 4 can be obtained. In the above description, the rules of rules 1 1 to 14 of the first rule series or the rules of rules 2 1 to 2 3 of the second rule series are equally processed, but it is also possible to attach importance to the rules. The degree is attached to the weight. In the case of this -24 - (22) 200800459, the attribution function of each rule corresponding to the rules 11 to 14 of the first rule series or the rules 2 1 to 23 of the second rule series is multiplied by the above r The parameters can then be integrated. Further, in the above fuzzy inference, the min-max synthetic centroid method is used to obtain the rotation angle of the motor, but it is not limited to this method, and other conventional methods such as the algebraic product-additional gravity method such as the gravity method may be used. . As long as it is based on experience, etc., the most suitable method for the laser processing apparatus that controls the fuzzy inference can be used. Next, in Tables 1 and 2, the parameters of the first rule series and the second rule series used in the above fuzzy inference are summarized in a table. It can be clearly seen from the parameters shown in Tables 1 and 2 that no particularly complicated fuzzy rules are set. Even so, the alignment of the optical system of the laser processing apparatus can be determined simply by performing the control based on the above fuzzy inference.
〔表1〕 有關第1光軸位置檢測器的時間之微分値與 驅動量的絕對値及旋轉反轉參數的關係 規則編號 感測器微分値 驅動量的絕對値 旋轉反轉5 11 LP LP + 1 12 SP SP + 1 13 ZE ZE + 1 14 NE SP -1 -25- 200800459 (23) 〔表2〕 對目標輸出的比AS與驅動量的絕對値的關係[Table 1] The relationship between the differential 値 of the first optical axis position detector and the absolute 値 and rotation reversal parameters of the driving amount. The numbered sensor differential 値 driving amount of absolute 値 rotation inversion 5 11 LP LP + 1 12 SP SP + 1 13 ZE ZE + 1 14 NE SP -1 -25- 200800459 (23) [Table 2] The relationship between the ratio AS of the target output and the absolute 値 of the driving amount
規則號碼 △ S 驅動量的絕對値 21 NL LP 22 NS SP 23 ZE ZE 此表1及表2的内容是分別爲圖11及圖12的歸屬函 數所表現者及數學上同値的内容。在此，該表1及表2所 示的參數意義如下。LP :較大的正値、SP :較小的正値 、ZE : 0、NE :負的値、NS :負的較小値、NL ··絕對値 較大的負値。 由以上説明可知，在本發明之雷射加工裝置的光學系 的調整步驟中，不需要所謂原點恢復動作。這是因爲作爲 上述模糊推論的根據使用的値對第1光軸位置檢測器的輸 出信號的時間差分値及目標値（極大値 )S〇而言，只是以AS = (s "3(0-1所賦予的AS。亦即，爲 了求取Y及AS的値，無論對哪個皆是不必進行所謂的原 點恢復動作而求取的値。其結果，即使基於何種原因（例 如間隙（backlash)等），光路調整部4未按照來自光軸 控制部5的控制信號正確地調整，還是能夠再度對光路調 整部4傳送控制信號，完成符合最適條件的對準。 又，本發明之雷射加工裝置的光學系的調整步驟中， 如上述，針對藉由第1光軸位置檢測器10所測定的受光 強度之一個資訊，調整値算出手段26可在光路調整部4 -26- 200800459 (24) 中實行的複數個調整處算出所分別對應的光路調整値，藉 此實現安定之光軸位置的固定。 另外，若在未利用模糊推論之下實現光路控制，則對 準作業中必須設置錯誤發生處理（程序）或暴走防止處理 " (程序）。供以使該等錯誤發生處理或暴走防止處理實行 ， 的程式量必須與上述模糊推論處理用的處理同等或以上。 而且，裝置的機構設計上亦必須準備限幅器開關（limiter φ switch )等暴走防止用的手段。暴走防止用的手段，特別 是在構成雷射裝置上重要，假設發生暴走，則會導致被加 工物損傷等重大的結果。 在上述實施例中所開示的模糊推論程式是按照非常單 純的算法來作成。爲了以單純的算法爲基礎，程式的性格 上是形成難以發生雷射加工裝置的暴走之構造。亦即，藉 由利用模糊推論，可使程式單純化，因爲使用模糊推論， 所以能夠以單純的算法來進行複雜的作業。 φ 又，雖是針對s’及AS的兩種類來進行判斷，但這將 有助於抑止上述暴走狀態的發生。若只藉由S 5或AS的其 中一方的判斷來控制對準作業，則會因混入控制信號的雜 訊等，暴走狀態顯現的危險性變大。在進行S’及.AS兩種 ’ 類的判斷時，只要使顯現暴走狀態的要因不發生於S’及 AS的雙方，裝置的暴走狀態便不會顯現。藉此，針對S， 及AS的兩種類來進行判斷，可顯著地縮小暴走狀態顯現 的機率。 如以上説明可知，此發明的雷射加工裝置當作控制系 -27- 200800459 (25) 統全體來看時，照樣形成暴走狀態難以發生的構造。 如以上所述’本發明的雷射加工裝置可藉由模糊控制 來調整因自雷射光源輸出的雷射光的光軸搖晃所引起之光 軸的角度誤差，可將安定的雷射光的光軸的位置予以固定 【圖式簡單說明】 • 圖1是表示具有毫微單位的精度之本發明的雷射加工 裝置1的全體構成槪略圖。 圖2是在第1光軸位置檢測器中配列成矩陣狀的受光 畫素中雷射光的照射點僅以1畫素作爲基準畫素來完全照 射的實施例。 . 圖3是在第1光軸位置檢測器中配列成矩陣狀的受光 畫素中雷射光的照射點均等地照射於形成4象限感測器的 4個畫素的一部份的實施例。 • 圖4是有關第1光軸位置檢測器1 〇的受光元件的配 置之變形實施例。 圖5是利用第1光軸位置檢測器及第2光軸位置檢測 器來檢測出雷射光的光軸的位置之變形實施例。 . 圖6是圖4的變形實施例。 圖7是在圖1的集光透鏡9與第！光軸位置檢測器 之間配置以支持台71所支持的針孔72之狀態擴大圖 〇 圖8是供以說明本發明之雷射加工裝置的光路調整機 -28- 200800459 (26) 能的槪略方塊構成。 圖9是供以說明受光強度變化。 圖1 〇是表示根據模糊推論之光路調整步驟的流程圖 〇 圖11是表示對S5的歸屬函數。 圖12是表示對AS的歸屬函數。 圖13是供以說明根據規則11〜14之統合化的步驟。 肇 圖14是供以說明根據規則21〜23之統合化的步驟。 圖15是供以說明求取統合化3作爲統合化1及統合 化2的歸屬函數的邏輯和的圖。 【主要元件符號說明】 1:本發明的雷射加工裝置 2 :雷射光源 3 :防振台 # 4 ··光路調整部 5 :光軸控制部 6 :可動台 7 :平台 8 :半透鏡 9 :集光透鏡 9 ’ ·集光透鏡 I 0 :第1光軸位置檢測器 II :被加工物 -29- 200800459 (27) 1 2 :可動台位置檢測器 1 3 :全反射鏡 1 5 :第2光軸位置撿測器 16 :第2半透鏡 — 18 :第2全反射鏡 • 20 :雷射光的照射點 21 : CCD畫素 φ 22 :全反射鏡 24 :全反射鏡 2 6 :算出手段 28 :馬達控制部 3 〇 :雷射光的照射點 A :畫素間距離 S 1〜S4 :畫素之雷射光的照射面積 40 : 4個畫素（4象限）的配列面 φ 4 1 :雷射光的反射球 42 :針孔 5 〇 ·反射鏡支持台 7 1 ·針孔支持台 * 72 :針孔 -30-Rule number △ S Absolute amount of drive 値 21 NL LP 22 NS SP 23 ZE ZE The contents of Tables 1 and 2 are the representations of the attribution functions of Figures 11 and 12 and their mathematical counterparts. Here, the meanings of the parameters shown in Tables 1 and 2 are as follows. LP: Large positive 値, SP: Small positive 値, ZE: 0, NE: Negative 値, NS: Negative smaller 値, NL ·· Absolute 较大 Larger negative 値. As apparent from the above description, in the optical system adjustment step of the laser processing apparatus of the present invention, the so-called origin recovery operation is not required. This is because the time difference 値 of the output signal of the first optical axis position detector and the target 値 (maximum 値) S〇 used as the fuzzy inference described above are only AS = (s " 3 (0 The AS given by -1, that is, in order to obtain the Y of Y and AS, it is not necessary to perform the so-called origin recovery operation, and the result is obtained, for example, for any reason (for example, gap ( The optical path adjusting unit 4 is not properly adjusted in accordance with the control signal from the optical axis control unit 5, and the control signal can be transmitted to the optical path adjusting unit 4 again to complete the alignment in accordance with the optimum conditions. In the adjustment step of the optical system of the injection processing apparatus, as described above, the adjustment calculation means 26 can be in the optical path adjustment section 4-26-200800459 for one piece of information on the received light intensity measured by the first optical axis position detector 10. 24) The plurality of adjustments performed in the calculation calculate the respective optical path adjustments 値, thereby realizing the fixation of the position of the optical axis of the stability. In addition, if the optical path control is implemented without using the fuzzy inference, the alignment operation is performed. It is necessary to set error occurrence processing (program) or runaway prevention processing " (program). The amount of program for which such error occurrence processing or runaway prevention processing is to be performed must be equal to or higher than the processing for fuzzy inference processing described above. The mechanism design of the device must also prepare means for preventing runaway such as limiter φ switch. The means for preventing runaway is especially important in constructing a laser device, assuming that a runaway occurs, the workpiece is processed. Significant results such as damage. The fuzzy inference program developed in the above embodiment is created according to a very simple algorithm. In order to be based on a simple algorithm, the character of the program is a structure that forms a runaway that is difficult to occur in a laser processing apparatus. That is, by using fuzzy inference, the program can be simplistic, and since fuzzy inference is used, complicated operations can be performed with a simple algorithm. φ Also, although it is determined for two classes of s' and AS, But this will help to suppress the above-mentioned violent state. If only by one of S 5 or AS In order to control the alignment operation, the risk of the violent state is increased due to the noise of the control signal, etc. When the S' and .AS types are judged, the cause of the violent state does not occur. In both of S' and AS, the state of the device will not appear. Therefore, judging the two types of S and AS can significantly reduce the probability of the state of the runaway state. As can be seen from the above description, the invention is When the laser processing apparatus is used as the control system -27- 200800459 (25), it is still possible to form a structure in which the violent state is difficult to occur. As described above, the laser processing apparatus of the present invention can adjust the cause by fuzzy control. The angular error of the optical axis caused by the shaking of the optical axis of the laser light output from the laser source can fix the position of the optical axis of the stabilized laser light. [Simplified illustration] Fig. 1 shows the unit with nanometer The overall configuration of the laser processing apparatus 1 of the present invention is accurate. Fig. 2 shows an embodiment in which the irradiation spot of the laser light arranged in a matrix in the first optical axis position detector is completely irradiated with only one pixel as the reference pixel. Fig. 3 is a view showing an embodiment in which the irradiation points of the laser light arranged in a matrix in the first optical axis position detector are uniformly irradiated to a part of the four pixels forming the four-quadrant sensor. Fig. 4 is a modified embodiment of the arrangement of the light receiving elements of the first optical axis position detector 1 。. Fig. 5 shows a modified embodiment in which the position of the optical axis of the laser light is detected by the first optical axis position detector and the second optical axis position detector. Figure 6 is a modified embodiment of Figure 4. Figure 7 is the collection lens 9 of Figure 1 and the first! The state in which the pinholes 72 supported by the support table 71 are arranged between the optical axis position detectors is enlarged. FIG. 8 is an optical path adjuster for explaining the laser processing apparatus of the present invention -28-200800459 (26) A little square composition. Fig. 9 is a view for explaining changes in received light intensity. Fig. 1 is a flowchart showing an optical path adjustment step based on fuzzy inference. Fig. 11 is a view showing a attribution function to S5. Figure 12 is a diagram showing the assignment function to the AS. Figure 13 is a diagram for explaining the integration according to rules 11 to 14.肇 Figure 14 is a diagram for explaining the integration according to rules 21 to 23. Fig. 15 is a diagram for explaining the logical sum of the integration function 3 as the attribution function of the integration 1 and the integration 2. [Description of main component symbols] 1: Laser processing apparatus 2 of the present invention: Laser light source 3: Vibration prevention table #4 · Optical path adjustment unit 5: Optical axis control unit 6: Movable table 7: Platform 8: Half lens 9 : collecting lens 9 ' · collecting lens I 0 : first optical axis position detector II : workpiece -29 - 200800459 (27) 1 2 : movable table position detector 1 3 : total reflection mirror 1 5 : 2 Optical axis position detector 16: 2nd half lens - 18: 2nd total reflection mirror • 20: Laser light irradiation point 21 : CCD pixel φ 22 : Total reflection mirror 24 : Total reflection mirror 2 6 : Calculation means 28 : Motor control unit 3 〇: Irradiation point of laser light A: Distance between pixels S 1 to S4 : Irradiation area of laser light of pixel 40 : Arrangement surface of 4 pixels (4 quadrants) φ 4 1 : Ray Reflective ball 42 for light: pinhole 5 〇·mirror support table 7 1 ·pin hole support table * 72 : pinhole -30-